International Financing for Development Co-operation On the linkages between foreign capital inflows and development in poor countries

Marco Zupi

ABSTRACT I posited in this paper that debt instruments should be used to provide aid. The capital market remains a veritable source for economic development. Other financial instruments like loans and equity were preferred over grants. This paper also diagnosed bilateral and multilateral aid systems. It was observed that both systems are good depending on the nature of aid. The text established a case against the extension of aids to wealthy nations. Through extensive research, poor governance, bribery and corruption were x-rayed among aid recipients; hence, the doctrine of transparency, openness, accountability, equity and prudence were suggested. Development would be impossible if poor countries refuse to play their own part; thus, reasons to examine these roles thoroughly. The world is a global village. Hoarding of information must be discouraged. Countries’ open data and technological transformation would to a large extent boost aid administration.

The paper examines some of the fundamental issues of dependency and underdevelopment in our contemporary international economic and political relations. The paper also analyses the paradox of foreign debt, which is supposed to be debt for development and financing development projects, but used as an instrument to perpetuate third world domination through financial imperialism. By adopting the secondary methodology, the paper discovers that, both advanced societies of the world and leaders of the third world are developing underdevelopment in the third world. The paper concludes that, third world indebtedness only exacerbates third world dependency on the economies of the advanced industrialized societies in this twenty first century. The paper recommends humanitarian aid to the third world against strategic aid, among other things.

This paper criticizes the current International Development Association (IDA) aid allocation and debt sustainability framework on the grounds of their over-reliance on the country policy and institutional assessment (CPIA) as the guiding criterion. It argues that CPIA-centred allocation of aid fails to introduce an incentives structure supportive of a genuine donor-recipient partnership, conducive to development. Further, it claims that the

Roskilde University International Financing for Development Co-operation On the linkages between foreign capital inflows and development in poor countries Zupi, Marco Publication date: 2002 Citation for published version (APA): Zupi, M. (2002). International Financing for Development Co-operation: On the linkages between foreign capital inflows and development in poor countries. Roskilde Universitet. FS & P Ph.D. afhandlinger No. 34 General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain. • You may freely distribute the URL identifying the publication in the public portal. Take down policy If you believe that this document breaches copyright please contact rucforsk@ruc.dk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 02. Dec. 2021 International Financing for Development Co-operation On the linkages between foreign capital inflows and development in poor countries Marco Zupi (PhD candidate) Roskilde University Centre, Denmark, 2002 The PhD program “Society, Business and Globalisation” Introduction 1 I. Old and New Paradigms of international Finance for Development 1. Economic Literature on Foreign Finance for Development. An Analysis of Alternative Paradigms 12 2. Foreign Debt and Development 36 3. Stock, Burden and Sustainability of Foreign Debt 66 II. Debt, Foreign Capital Inflows and Development Crisis in Sub-Saharan African Countries 4. Debt Crisis in Sub-Saharan Africa 102 5. Review of econometric literature on the linkages between foreign capital inflows and development 138 6. Exploring the complex relationship between foreign capital inflows and development in Africa Appendix 1 - Data Results Appendix 2 - A case-study of foreign capital flows inter-links: Italian development aid and debt relief 181 216 224 III. Current Solutions to Debt Crisis and Holistic Proposals on Global Finance for Development 7. The HIPC Initiative: its rationale and Implementation 241 8. External critiques to the HIPC Initiative 274 9 Current Proposals of International Civil Society on Development, Poverty Reduction and Foreign Finance 303 Conclusion 341 To AnnAngelo & the presence of Nadine Angioline «At the end of every seven years, you must cancel debts....Be careful not to harbor this wicked thought: ‘The seventh year, the year for canceling debts, is near,’ so that you do not show ill will toward your needy brother and give him nothing» Deuteronomy 15:1,9 «There is no practice more dangerous than that of borrowing money» George Washington, 1791 «The biggest single mispection about the foreign aid program is that we send money abroad. We don’t. Foreign aid consists of American equipment, raw materials, expert services, and food. 93 percent of aid funds are spent directly in the US to pay for these things» William S. Gaud, Department of State Bulletin, 1968 «The problems of many heavily indebted developing countries are a cause of economic and political concern and can be a threat to world stability» Declaration of G7, 1988 «If I were the president of a Third World nation… I would be far more frightened by a well-dressed gentleman bringing loans from the IMF or Citibank than by a bearded guerrilla muttering threats of revolution» Lewis Lapham, Imperial Masquerade, 1990 «The debate is not about whether some countries need some debt reduction, but how deep this reduction should be. Is there a level which is sustainable?… How long will it take to get this reduction of debt in place, and what safeguards should there be?» Andrew Rogerson, Jubilee 2000 campaign, 1999 «Drop the debt!» Bono and Jovanotti, G-8 Genoa Summit, 2001 Introduction 1. The Limitations of Scientific Paradigms Thomas Kuhn applied the term «paradigm» to define that «constellation of values, beliefs and perceptions of empirical reality, which, together with a body of theory based upon the foregoing, is used by a group of scientists, and by applying a distinctive methodology, to interpret the nature of some aspect of the universe we inhabit».1 Thus, the scientists use a schematic view of the universe, which is object of their study, an intellectual framework, a gestalt providing the modus operandi to approach their work. Governed by a particular vision of the nature of the problems, they can interpret and explain facts as well as solve problems. The inability of a vision to solve the problems means a paradigm failure. It can be due to some limitations of the paradigm itself, pressed by the challenge of external factors – technological and socio-economic changes, for example –. Then, a paradigm failure can directly influence the emergence of new paradigms; but there is no linear evolution of paradigms. In the social sciences, as well as in the physical sciences, it is inevitable that some paradigms prove politically more acceptable than others, and it does matter. As Northrop and Kurt Gödel demonstrated 2 , physics and mathematics can not be considered very «objective» disciplines, even less – Barrington Moore jr. wrote – the social sciences3 . And as Schumpeter said: “Economic laws are much less stable than are the laws of any physical science… they work out differently in different institutional conditions, and… the neglect of this fact has been responsible for many aberrations” 4 . The «objective» potential of a paradigm to interpret the facts is not what determines the success of this paradigm. We should distinguish between the ontological nature (referred to the object of the scientific knowledge, without considering its relationship with the observer: the real and final nature of «matter») and the epistemological nature (referred to the interrelation of the scientist’s mind and «matter» in the processes of perception and knowledge) of a paradigm. We could say that Albert Einstein, with his general and restricted theory of relativity, changed the ontological theory of physics, through a radical shift to a new vision of a symmetrical link between space-time and matter 5 . On the contrary, quantum mechanics, and particularly the uncertainty principle of Werner Heisenberg6 , changed also the epistemological theory of scientific knowledge 7 . Nevertheless, just following Gödel and Heisenberg, it is important to stress that the content of social theorising always reflects the values and purposes of scientists, even when it seems to be a very objective description of the reality of facts. 1 See Kuhn, T. (1962), The Structure of Scientific Revolutions, University of Chicago Press, Chicago. See Gödel, K. (1962), On Formally Undecidable Propositions, Basic Books, New York. The original German version was published in 1931. 3 See Barrington Moore jr. (1966), Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World, Beacon Press, Boston. 4 This idea has been adopted in modern comparative studies of economic performance. For instance, see Bruno and Sachs argue that general economic theories wrongly consider one and the same basic model as applicable to all theories. See J. A. Schumpeter (1954), History of Economic Analysis, Oxford University Press, New York, p. 34. 5 Einstein conceived, or took from Newton, the hypothesis that light travels in curved, not straight lines; deduced from it the conclusion that a star appearing to be in a certain position in the heavens is really a little to one side of that position. Matter curves space-time, and this curve determines the movement of matter. No geometry is inherent to the space. See Schlipp, P. A. [ed.] (1949), Albert Einstein, Philosopher Scientist, The Library of Living Philosophers, Evanston. 6 The irreversibility, stochasticity and non-locality implications of quantum theory, as well as the recognized failure of predictability in Newtonian dynamics, derive from the important quantum discontinuity. It can be linked to the instabilities, bifurcations and fluctuations in chemical systems as well as to the entropic interpretation of the universe. See Heisenberg, W. (1972), Physics and Beyond. Encounters and Conversations, Harper Torchbooks, New York. 7 See Prigogine, I. and Stengers, I. (1988), Entre le Temps et l’Éternité, Librairie Arthème Fayard, Paris. 2 1 2. New Paradigms of international governance and the changing World-System Thomas Kuhn, with his brilliant work published in 1962, provided an excellent description for those who want to think about how old paradigms are replaced by new ones in the long run. Referred to the natural science, he said that scientific discovery «commences with the awareness of anomaly. It then continues with an exploration of the area of anomaly. And it closes only when the paradigm theory has been adjusted so that the anomalous has become the expected»8 . Nowadays, the paradigms we use to understand the changing World-System trends are strongly stressed by events that discredit theories and predictions. After the end of the Cold War and US hegemony era (1945-1990), can the current North-South history be exhaustively explained? Political science, economics, sociology, law are the pivotal disciplines we use to produce «scientific» theories which very easily fail to understand and predict trends. Yet, these paradigms remain our main reading of real world. Nowadays they become more and more important as we are daily bombarded by information and we risk of making no sense of it. Inevitably, whether we know it or not, we continue to use theories to understand, explain and link facts as well as to decide what to do. It is particularly relevant nowadays, as we are passing through a post-bipolarism crisis, and periods of change like this create opportunities and risks to be decoded9 . Referring to the Third World position, Wallerstein says it risks to become far more difficult, passing from a time of hopes and struggles to a time of troubles and desperation10 . Using the human development concept introduced by the United Nations Development Program11 , the opportunities of (political, economic, cultural and social) choice become more and more limited, and uncertainty and social exclusion risk to be the prevailing phenomena in developing countries. The problem of exploitation must be modified or extended, in the era of globalisation, considering marginalisation process a new main dimension of between and within-country inequality. In particular talking about developing countries, States, societies and citizens are subject to a global economic regime and to push to economic policies, not necessarily respondent to local needs, aspirations and market characteristics 12 . How well does current international strategy to face external debt problems of poor countries create the opportunities for promoting sustainable development processes and equitable market arrangements to serve local needs? And do new paradigms of international systems offer adequate reading of it? The need for more adequate theoretical framework to understand changes over the world economy is evident. Social interest is a concept that must be used extensively in the discussion of external debt problems, as debt involves – directly and indirectly – a wide range of social actors, debtors as well as lenders with their own legitimate interests. A concept to be reconsidered is «class», because the important changes of the organisation and division of labour and the very specificity of developing countries’ context should imply a reconsideration of the concept of class as it was categorised by Marx. The defeat of the USSR, the political fall of the Communism and the collapse of Marxism-Leninism as an ideological force, although the dominant USSR version of Marxism, through its Nomenclature, was never the only one, can produce, as a consequence, the questioning of one of the 8 Kuhn, T. (1962), ibid. Crises amplify the need to change and push to new ideas, but it also risks to discourage and create confusion. Neither more nor less what the ancient Greek demonstrates through its literature: the Antigon, by Soflokes is a good example. 10 Wallerstein, I. (1995), After liberalism, The New Press, New York. 11 UNDP (1990), Human Development Report 1990, Oxford University Press, New York. 12 Amoroso, B. (1998), «Pattern of industrialization in the Mediterranean Region. The Small Medium Size Enterprises», in FEMISE, Les enjeux du Partenariat Euro-Méditerranée, FEMISE, Marseille and Amoroso, B. (1999), Economie Mediterranee. Sistemi produttivi locali tradizionali e di nuova formazione. Cooperazione Sud-Sud e Nord-Sud. V Rapporto sul Mediterrraneo, CNEL-Università di Roskilde, Roma. 9 2 most important concepts of the Marxian theory: the concept of class struggle, which was largely used by a lot of studies on Africa 13 . A flourishing of aggregations on ethnic (or religious) ground looms on the horizon, lacking of evident social features and taking up a defensive position of their inter-classist, particularistic, localistic identities. In such a way, the concept of class could be reconsidered in a ethnic-nationalist context, in spite of the ambiguous and reactionary features that result from a conflict caused by ethnic components: all the liberating potentialities of nationalism finish to be absorbed by the perversity of xenophobia and intolerance, losing sight of the real mechanisms of injustice, discrimination and crisis 14 . The inadequacy of Marxist analysis of nationalist, racial or ethnic struggle has been widely noted, and particularly the thesis that it is always possible to argue that other forms of struggle are masked forms of the class struggle 15 . The Marxian analysis lacks of flexibility and considers ethnicity as a super-structure; in short, the class conflict between the proletariat and the bourgeoisie are primordial, and all other conflicts are epiphenomenal and neglected 16 . This led Marxist analysis deliberately to understate the theoretical importance of any social cleavage other than that of class. Whereas, today social polarisation and marginalisation has got worse and we must search for a more useful definition of social interests: class struggles tend to be conducted under the banner of race and ethnicity (and religion), and the true bourgeoisie and proletariat definitions are too narrow, reflecting the society of the XIX century17 . In Africa, since the 1960s, ethnism and class struggle overlapped to promote the self-determination process of former colonies 18 . When we deeply investigate the problem of external debt of poor countries, other forms of conflicts than the proletariat and the bourgeoisie struggle seem to prevail. However, as Karl Polanyi argued, capitalism remains itself, that is an anomaly since it embodies a system wherein social relations are defined by economic relations; whereas in previous economic systems economic interactions followed from social relations 19 . An anomaly with, at the other hand, very few counter-systems. This general rule must be taken in account, as external debt problems seem to be a main characteristic of contemporary global capitalism. In any case, an even larger proportion of the world’s population falls into the «lower stratum»: postcolonialism and North-South relationship have masked and changed the nature of dominating and dominated classes. A map of numerous and fragmented constituencies (local, national and international ones; from official, private and civil society’s sectors) try to impose their conflicting interests having different political voices and economic powers. From this perspective, therefore, it is evident the fallacy of any axiom on the likelihood of automatic positive-sum games for the global problems, where all the players are the winners 20 . 13 See Arrighi,G. (1974), Sviluppo economico e sovrastruttura in Africa, Einaudi, Torino; Arrighi, G. (1983), "Labour supplies in historical perspective: a study of the proletarianization of the African peasantry in Rhodesia", in Essays, Arrighi & Saul ed., New York; Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London; Daka,V. and G. N. Mudenda (1980), "Class Formation and Class Struggle", in Southern Africa Universities Social Science Conference, Morija Printing Works, Lesotho. 14 Silvestri, G. (1991), «Editorial», in Indipendenza, Monthly bulletin, year VI, May, Rome. 15 Wallerstein, I (1991a) «Class Conflict in the Capitalist World-Economy», in Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London; Wallerstein, I (1991b) «Social Conflict in PostIndependence Black Africa: The concepts of race and Status-group reconsidered», in Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London. 16 Balibar, E. (1991), «From Class Struggle to Classless Struggle?», in Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London. 17 Wallerstein, I. (1995), ibid. 18 Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London. 19 Polanyi, K. (1944), The Great Transformation, Rinehart & Co., New York. 20 The variable-sum game refers to a situation in which two or more players have an option that permits them to cooperate for a greater mutual payoff than either would gain alone in making a choice that did not permit or involve 3 From this point of view and perspective, we can analyse debt crisis in Africa. Nowadays, SubSaharan Africa is facing a debt crisis of major international significance. We argue that debt relief can be considered a sort of «game», representing the complexity of effective governance of globalisation. Jubilee 2000 campaign has been one of the most successful international initiatives of the last decades. Because of the support of very important religious figures – such as the Pope and Dalai Lama – and a direct involvement of famous pop stars - as Bob Geldof and Bono in UK, or Jovanotti in Italy – everyone knows what is about. The international political community itself has been hugely pressed by this non-governmental campaign spread all over the countries. In fact, this campaign, originated within civil society organisations, has induced a more serious political concern on the problem of external debt of poor countries. Consequently, several governments of industrialised countries and international financial institutions have committed themselves to pursue an immediate debt relief strategy. In some cases, as happened in Italy, this movement has determined new national Laws, focused on the bilateral implementation of debt relief initiative. At multilateral level, the «Heavily Indebted Poor Countries» (HIPC) initiative was launched in the 1996 by the World Bank and the IMF and it has been enhanced in the 1999. This initiative intends to reduce the external debt of the world's poorest, most heavily indebted countries to sustainable levels, through debt cancellation. From this perspective, debt relief seems to demonstrate the emergence of international political consensus and capacity to implement a win-win strategy, which will satisfy debtor countries as well as most of the constituencies based in developed countries. But apart from the rhetoric widely spread all over the world on debt cancellation, we should remind that when we talk about developing countries, States, societies and citizens are subject to a global economic regime and to economic policies, not necessarily respondent to local needs, aspirations and market characteristics. Political motivations have been by far the more important for aid and loans, especially for the major industrialised countries. Creditors and donors give resources because it is in their political, strategic, or economic self-interest to do so. Some assistance may be motivated by humanitarian desires, but there is no historical evidence to suggest that over longer periods of time, rich nations assist others without expecting some corresponding benefits (not necessarily financial) in return. Political motivations are the real determinants of decisions. When the US decided to write-off certain debt owed by Egypt, it followed a more fundamental switch in Egypt’s regional policy, as Egypt recognised Israel and became a US partner in a long-term regional settlement and in the Gulf war against Saddam Hussein. Or, more recently, as the US-led Operation “Enduring Freedom” against Afghanistan’s Taliban regime got under way, the United States needed crucial intelligence and logistics support from Pakistan. In return Pakistan’ President General Pervez Musharraf expected economic assistance and a revival of close ties with Washington. And in fact the US Congress waived sanctions against Pakistan for an initial period of two years and Britain and the United States have come forward with economic packages, which fall far short of wiping out Pakistan’s $36 billion outstanding debt. In contemporary world, there are no permanent friends and enemies in international relations, and therefore countries base their ties on shared interests, and debt cancellation can simply derive from temporary political agreements. Nevertheless, rhetoric prevails, when politicians say that a final solution to external debt problem of poor countries is currently implemented, and it strikes people as being expensive for lender countries to assist poor countries, as far as guided by their moral responsibility, and given the existence of technical problems to provide more debt relief. And the economic rationale has been given lip service as the overriding motivation. At this regard, our purpose is to investigate the cooperation. This contrast with the zero-sum game (i.e. chess) in which cooperation in not rational since each choice with a positive payoff for one player has exactly the same negative payoff for the other player: the focus is upon the distribution of the existing «pie», rather than upon the creation of new pie. See von Neumann, J. and O. Morgenstern (1947), Theory of Games and Economic Behavior, Princeton University Press, Princeton NJ, 2nd edition. 4 economic aspects and rationale of external debt crisis in order to unmask the abuse of technical excuses, which often cover the lack of political will 21 , and to demonstrate the complexity of problems and interests in the age of globalisation. We have a map of multiple interests and preferences (political, strategic, economic, environmental,…), each of one having different time horizons (long-term or short-term horizons). The Multiple Players of "Debt Relief" Game Lenders (good and bad) Multilateral II.OO. (=I.F.I.) the World Bank, IMF and RDBs Borrowers (good and bad) HIPC (small debtors, official exposure) Regional official lenders EU The Least developed countries Bilateral official lenders G7 Other members of the Paris Club Other countries Low income countries Private lenders Commercial Banks (London Club) Private Enterprises Export Credit Agencies Middle income countries (big debtors, private exposure) Civil society (indirect stakeholders) Savers, consumers, NGOs, poor... Civil society (indirect stakeholders) Savers, consumers, NGOs, poor... The relationship between all the involved interests is thus of crucial importance. However, its significance is not often fully understood nor its developmental potential explored. In this case the basic question remains “is a positive-sum solution possible and likely in the debt relief game?” In sum, there are limits to the liberal belief that the pursuit of self-interest in a free, competitive economy achieves the greatest good for the greatest number in international no less than in the national society. And, if liberals think too little about powers, realists focus excessively on its relevance for most powerful states. Any outcome, like what is said by those who subscribe to complexity theory, must be judged in terms of specific results for high or low-empowered constituencies, representing different, legitimate social interests. International politics offers no quick panacea for these myriad social interests, even though the liberal school perspective has tended to elide power consideration and its adherents have not thought very hard about how everyday market imperfections and failures, born of conflicts between different preferences, block the achievement of the liberal dream22 . Particularly in international economics and debt crises management, the role of conflict and co-operation have been treated as largely unproblematic. Considering all these elements, we share Barrington Moore jr.’s ideas 23 : impartial and truthful analyses seem pieces of criticism, denunciations rather than «objective» statements; to embrace the cause of the victims of historical processes and to be sceptical regarding the winners’ pretension is the basic safeguard against myths. 21 Any argument that the money is not available is not acceptable: in just one week, during December 1997, donors managed to pledge US$57 billion to bail out South Korea (Japan pledged US$10 billion, the US pledged US$5 billion). 22 Durfeu, M. And J. M. Rosenau (1996), «Playing catch-up: international relations theory and poverty», in Millennium: Journal of International Studies, Vol. 25, No. 2. 23 Barrington Moore jr. (1968), Critica della Tolleranza Pura, Einaudi, Torino. 5 Debt crisis in the 1970s, 1980s and 1990s has been one of the main factors inducing a radical shift in development strategies of all developing countries and a severe reduction in their concrete opportunities to search for alternatives to orthodox measures. Old and new forms of debt management can be considered satisfactory or not, depending on what social interests we are referring to. Debt crisis is clearly one of incipient signs that globalisation and the institutional architecture for its governance 24 are not providing positive results to developing countries and poor. Analysts are increasingly citing Karl Polanyi’s 1944 classic, The Great Transformation, as a basis for thinking on the possibility to have again in the twenty-first century the experience of «double movement» described by Polanyi as a (first) movement of the twentieth century induced by fervour for wide-open capitalism without government control which led to economic crises, followed by a (second) movement founded on authoritarian governments that disciplined all society, culminated in the fascism and produced the onset of world war. What prevented, during those tragic moments, the economic and political constituencies and powers from taking the step necessary to prevent a global crisis? Is the debt crisis management demonstrating that current international governance of globalisation is more aware of the profound negative consequences debt is causing on poor people or simply the illusion of a real control over the dramatic problems of this period? This research considers international finance and debt relief as a part of coherent strategy to reduce poverty and promoting growth factors, particularly for Sub-Saharan Africa (SSA). 3. Objective of this work This study has three aims. The first is to provide a systematic analysis of the economic rationale of external debt for development of poor countries. This analysis combines the heritage of economic theory literature and the more recent debate on finance for development. Surprisingly, such a combination does not already exist in one book, and this work aims, albeit partially, to fill this gap by covering a lot of major issues related to debt of poor countries. The second aim is to analyse debt crisis in SSA. The problem of debt can be judged in principle at several levels. This work tries to assess the complex issue from different perspectives: political, economic, social and historical dimensions are presented, at aggregated level as well as at national level, using both data description, inferential statistics and econometric methods. A final aim of this study is to review the extent, the effectiveness and the future prospects for current debt relief initiatives and also to explore the vision, expressed as a set of recommendations, of international civil society, which has been seriously involved in the search for solutions of debt crisis. This work presents an original synthesis of a lot of documentation and discussions of different Non governmental organisations (NGOs) and networks, all of them working on finance for development issue. Civil society feels a growing need to take part in the construction of international policies through co-operation; the social movements concerned with peace, human rights, solidarity, the environment and development have established the organisational foundations of which a broad, heterogeneous and variegated “grassroots globalisation” movement has started building. Our summary unquestionably proves the complexity of the discussion, the importance of the contents (rather than slogans) and the maturity of the proposals within the multifaceted nongovernmental world. This part is an expression of the capacity of the numerous organisations of civil society, apart from a lot of different and legitimate positions and expectations, to engage in proposals on the content, which is a reflection of the skill and know-how accumulated with regard to the external debt issue, within a coherent comprehensive approach to financing for development. In fact, we have to remind that, when we talk of Sub-Saharan African countries, the concept of external debt is not very separate from development aid. 24 The globalized political structures: United Nations system, Bretton Woods institutions, OECD, G8. Liberal theory allows for the resort of macroeconomic coordination at the international level, as the G8, as a means of enhancing the smooth operation of market, but it is precise at the international institutional level that effective solutions to debtors crises are not guaranteed. 6 Moreover, for a number of years “aid was calculated as the sum of official and private capital flows, although now these two items are listed separately” 25 , and aid is “a term which had been ambiguously used and sometimes actually applied to the entire flow, including private investment and export credits” 26 . And, “Although the private international capital market revived in the late 1950’s, for LDCs capital flows remained almost entirely official. This remained the rule in the 1960’s, with the notable exceptions of a few dramatically successful countries” 27 . In the case of African countries, it is still true. That’s why, to analyse the relationship between debt and development in Africa as well as current debt relief initiatives is a way to analyse development co-operation, within the main context of finance for development. This is the final result of our work. 4. Approach and methodology of this work In theoretical terms, our reflections try to link some pair issues, which have been traditionally considered as opposite terms in a dualistic approach. The vision lying beneath this traditional approach assumed a broader opposition between aid and debt (and trade), between domestic and international determinants of development, between economic growth and social development. It is, in our view, partly fallacious to see these pairs as internally contradictory and more so the longer the run we consider. Our purpose is to demonstrate how these pairs, in the long run, can mutually interact and strengthen each other, given the characteristics of the debt relief game. The Set of Pairs to Be Combined in an Olistic Approach to Solve the Debt Relief Game Issues (A) Issues (B) Analysis of the macro-economic performance and prospects: GDP growth, saving and investment growth, the impact of stabilisation and structural adjustment programs, economic policies - budget, taxation Analysis of the sources of an endogenous growth: the strategic nature of investment, development of infrastructure, expenditure on education, positive externalities of investment, industrial development policy and agriculture, State-market equilibrium Analysis of objectives of national development strategies Analysis of the interests of creditors Analysis of the impact on social development: poverty assessment, health and education situation Analysis of the exogenous factors: regionalisation and globalisation, international trade and financial markets, export diversification and import substitution Analysis of international development co-operation policies Analysis of the interests of debtors In linking these issues and the complex map of multiple interests involved in debt relief and poverty reduction strategies, we hope this could be of interest to national policy makers and to the international community, to establish the priorities for action an to offer a means for broadening theoretical lenses to encompass debt relief. In approaching the debt relief issue, our aim is to provide an integrated debt relief-development strategy, taking in account all the major constraints that, both at theoretical and operational level, can affect any strategy of debt relief, and are used to justify the limitations of debt relief for SubSaharan Africa. 25 M. P. Todaro (1989), Economic Development in the Third World, Longman, New York, p.481. W. T. Newlyn (1977), The Financing of Economic Development, Clarendon Press, Oxford, p. 98. 27 A. O. Krueger (1987), “Debt, Capital Flows, and LDC Growth”, American Economic Review, Vol.77 N. 2, May, p. 159. 26 7 Problems to be Considered in Solving the Debt Relief Game - Theoretical problems Moral Hazard Information Asymmetry Free riders Fungibility of aid International Reputation Operational Constraints Political (will) Financial (resources) Technical (equilibrium and steady-state path) Tension: Conditionalities vs. Ownership Future Lending - From a methodological point of view, the study of debt relief is an end in itself (debt reduction) and a means to another end (sustainable human development). It is based upon the purpose of producing a systematic and cogent case for analysing external debt problems of poor countries. This research results from wide consultation with international NGOs, the exchange of questionnaires, information and documentation. This study cites and derives from many background studies prepared on thematic issues as well as analyses of experiences in countries, international technical reports and position papers by international organisations, who generously shared their materials. This study tries to provide an original synthesis of the main economic analyses and policy recommendations on external debt problems of poor countries provided by academic literature and experts, too. The Paris Office of the World Bank, in the person of the external counsellor Susana Esteban Barrocal, kindly provided data and information on a number of debt issues and invited me to participate to seminars on global finance. The Italian Ministry of Foreign Affairs, through its Statistics Office, General Directorate for Development Co-operation, provided data on bilateral issues. As co-ordinator of the Genoa Non Governmental Initiative, promoted by the Italian Presidency of the G8, and as co-ordinator of the research activities to monitor the implementation of debt relief initiatives, co-financed by the European Commission, DG VIII, during the year 2001, I had the opportunity to contact and discuss the main issues of debt relief with researchers and experts. In 2000 CeSPI informal group on debt relief was a great opportunity to discuss debt issues with some researchers coming from different Italian institutions and represented a great opportunity of discussion. Thus, this work is mainly based on reviewing academic research papers, official documentation from international organisations - UN agencies and international financial institutions -, internal memoranda and interviews with experts. This study is based on direct communication with officials from international organisations and bilateral agencies. This study has benefited from numerous discussions and criticism. Prof. Bruno Amoroso was my supervisor and I have greatly benefited from many discussions with him. I was, furthermore, very stimulated by the questions that were asked and the remarks that were made in the discussions following my lectures and seminars at Roskilde University Centre and La Sapienza University. It is important to record that this study has been completed from lenders’ capitals (Brussels, Copenhagen, London, Paris, Rome, Washington D.C.). No visits were made for the purposes of this work to SSA countries. This will inevitably create some bias to the evidence presented, since the reality of debt crisis is basically its impact on real live in poor countries. 5. Structure of this work Based on this preliminary care, this work is divided, in terms of structure, into three parts. The complexity of external debt problem in SSA is the basic mainstreaming, spread all over the three parts. They have to be considered a progressive way of approaching the issue of debt relief. The analysis departs, in Part I, from a review of theoretical foundations of international debate on debt relief: the unconscious long history of economic thoughts – chapter 1 -, the ambiguous linkages between development and debt – chapter 2 -, and the difficulties to define, in general terms, burden and sustainability of debt – chapter 3 -. Then, it passes, in Part II, through the history 8 of debt crisis in SSA countries – chapter 4 -, and it investigates the empirical evidence of the problems – chapters 5 and 6 -. The final Part III analyses peculiarities, solutions and problems of current debt relief initiative – chapters 7 and 8 – and provides an unquestionable proof of civil society’s capacity for proposals on debt relief for the future – chapter 9 -. These three parts, and all the chapters within them, are directly linked together; every chapter must be considered a partial verification of the general idea. Beyond prevailing rhetoric of international initiatives, which affirm to make external debt sustainable (as the current multilateral initiative says), differences among countries are important, the nature and composition of debt are important, specific history and context do matter, the interests and motivation and bargaining power of different lenders and debtors are crucial. There is no “scientific” reason to define a general threshold of debt sustainability or to identify 41 countries eligible to debt relief initiatives (as the current HIPC initiative does), rather than concrete, legitimate political will. And it is important to link debt relief initiatives to aid policies, trade regimes, financial systems and national economic and social policies. In detail, in part I, chapter 1 tries to provide an original review of the development economic theories that emerged at different historical phases, in different contexts, with different purposes. The originality of this review is its focus on the relationship between foreign finance and theory of development process. Even though development economics emerged after the II World War to confront the causes of poverty in developing countries and to define strategies for economic progress, nevertheless the economists brought with them the legacy of mainstream economics – appropriate or not – within the new sub-discipline. Chapter 2 explores and analyses alternative models used in theoretical and empirical studies to capture the links between debt and development, in order to provide an original synthesis of different points of view. In particular, it emphasises the importance of fiscal components of external debt problem, the so called three-gaps model, which is considered a peculiarity of SSA case. But it also stresses the complexity of factors that interact, as demonstrated by the fungibility argument and the Dutch disease. The complexity and the presence of counteracting effects seem to be the real nature of development and debt linkages, when we analyse in detail economic literature. Schematic and simple relations must be abandoned in favour of case-by-case analysis. Chapter 3 provides a review and critique of the literature on the sustainability of foreign debt, conducting the discussion in economic and financial as well as political terms, and how this relates to the question of debt relief. Part II is an analysis of SSA debt crisis. Chapter 4 presents the crisis exploded in the 1980s, which invested the least developed countries of Sub-Saharan Africa, describing the measures implemented to solve it. The chapter tries to emphasise the characteristics of African problems, which stress the importance of contextual aspects (SSA is different from Latin America), within the same thematic issue of debt crisis. This chapter tries to define the profile of current debt problem, within the context of total external capital inflows. Chapters 5 and 6 are devoted to an econometric analysis of debt problems. In economics and other fields, statistical models that allow for more complex relationships than can be inferred using only cross-sectional data, are taken in account as a complementary empirical test. Chapter 5 is a review of the econometric analysis applied to foreign debt problems of developing countries. It provides a synthesis of the main approaches, which have been used in econometrics to analyse the relationship between external finance (debt and aid) and development: a complex map of different approaches and set of variables, which provide competing results. Chapter 6 uses ideographic technique, multi-variate analysis and pooled time series analysis, correcting an initial specification of the relationship to be estimated, basically related to a linear regression form, through different pooled models, in order to provide the most adequate model needed for investigating the links between external debt and development in SSA. Through the help of ideograph technique, the chapter provides some information on the dynamic structure of the relationship between the variables linked to foreign debt problem in SSA. The empirical analysis 9 examines whether or not debt relief, in fact, interacts with development process in SSA. It examines whether debt crisis reduced the resources available to developing countries, particularly those directed towards sustainable human development, inducing a change in the expenditure patterns of recipient governments. Part III presents the current political debate on the need of linking debt relief, development and poverty reduction strategies, animated by NGO’s networks (Third Sector organisations, interest groups, trade unions, academics, research centres and local administrations). Chapter 7 addresses the «Heavily Indebted Poor Countries» initiative launched during the 1990s by the World Bank and the IMF. This chapter analyses, from inside the World Bank and the IMF, the HIPC rational, describing how this initiative intends to reduce the external debt of the world's poorest, most heavily indebted countries to sustainable levels and what conditionalities it implies in terms of critical social and economic reforms; it analyses how much the HIPC initiative has accomplished over the little more than four years it has been in place, drawing a comprehensive review of the HIPC Initiative, including updated cost estimates. Chapter 8 provides a critical and comprehensive review from outside the World Bank and the IMF of the key problems of the HIPC Initiative: it begins with some brief comments on the rational for debt relief and analyses then some key issues related to the HIPC Initiative’s goal to achieve debt sustainability and the weakness of its long-term implications. The aim of this chapter is to demonstrate how the external critical comments have contributed to correct the nature and implementation of the HIPC Initiative, thus confirming the importance of dialogue between governmental and non-governmental constituencies to improve the quality and performance of policies. Nevertheless, the distance between positions remains clear, as it is clear that civil society organisations constitute an innovation on the scene of international relations, new global actors in the globalised world. Chapter 9 does represent an original synthesis of a lot of separate political requests that can be reduced at coherent and comprehensive approach. This is an important political counter-part to the economic analyses presented in the previous parts. It is not necessarily based on the theoretical assumptions proposed in the economic literature 28 , but it represents an effective lobbying campaign to be taken in account, which should be seriously assessed, through empirical analyses on their arguments. This chapter cites and derives from many background studies, which have been recently prepared on thematic issues as well as analyses of experiences in countries, international technical reports and position papers by NGOs and international organisations. The international civil society has become an increasingly powerful influence on international relations; its political contribution to international finance and development co-operation has to be considered. A concluding chapter brings the strands of the argument together, in order to comment the outcomes in terms of the multiple interests and preferences involved in a «debt relief game» and to suggest some measures needed to fully implement and strengthen the implementation of initiatives of debt relief, treating them as a component of a coherent package that must constitute an holistic approach to international development co-operation. 28 In the political debate, many of the theoretical arguments are considered just that, theoretical argument. 10 Bibliography B. Amoroso (1998), «Pattern of industrialization in the Mediterranean Region. The Small Medium Size Enterprises», in FEMISE, Les enjeux du Partenariat Euro-Méditerranée, FEMISE, Marseille. B. Amoroso (1999), Economie Mediterranee. Sistemi produttivi locali tradizionali e di nuova formazione. Cooperazione Sud-Sud e Nord-Sud. V Rapporto sul Mediterrraneo, CNEL-Università di Roskilde, Roma. G. Arrighi (1983), "Labour supplies in historical perspective: a study of the proletarianization of the African peasantry in Rhodesia", in Essays, Arrighi & Saul ed., New York. G. Arrighi (1974), Sviluppo economico e sovrastruttura in Africa, Einaudi, Torino. E. Balibar (1991), «From Class Struggle to Classless Struggle?», in Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London. E. Balibar and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London. B. Barrington Moore jr. (1968), Critica della Tolleranza Pura, Einaudi, Torino. B. Barrington Moore jr. (1966), Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World, Beacon Press, Boston. V. Daka and G. N. Mudenda (1980), "Class Formation and Class Struggle", in Southern Africa Universities Social Science Conference, Morija Printing Works, Lesotho. M. Durfeu And J. M. Rosenau (1996), «Playing catch-up: international relations theory and poverty», in Millennium: Journal of International Studies, Vol. 25, No. 2. K. Gödel (1962), On Formally Undecidable Propositions, Basic Books, New York. The original German version was published in 1931. W. Heisenberg (1972), Physics and Beyond. Encounters and Conversations, Harper Torchbooks, New York. A. O. Krueger (1987), “Debt, Capital Flows, and LDC Growth”, American Economic Review, Vol.77 N. 2, May. T. Kuhn, (1962), The Structure of Scientific Revolutions, University of Chicago Press, Chicago. W. T. Newlyn (1977), The Financing of Economic Development, Clarendon Press, Oxford. K. Polanyi, (1944), The Great Transformation, Rinehart & Co., New York. I. Prigogine, and I. Stengers (1988), Entre le Temps et l’Éternité, Librairie Arthème Fayard, Paris. P. A. Schlipp, [ed.] (1949), Albert Einstein, Philosopher Scientist, The Library of Living Philosophers, Evanston. J. A. Schumpeter (1954), History of Economic Analysis, Oxford University Press, New York. G. Silvestri, (1991), «Editorial», in Indipendenza, Monthly bulletin, year VI, May, Rome. M. P. Todaro (1989), Economic Development in the Third World, Longman, New York. UNDP (1990), Human Development Report 1990, Oxford University Press, New York. J. von Neumann and O. Morgenstern (1947), Theory of Games and Economic Behaviour, Princeton University Press, Princeton NJ, 2nd edition. I. Wallerstein (1991a) «Class Conflict in the Capitalist World-Economy», in E. Balibar, and I. Wallerstein, Race, Nation, Class: Ambiguous Identities, Verso, London. I Wallerstein, (1991b) «Social Conflict in Post-Independence Black Africa: The concepts of race and Status-group reconsidered», in Balibar, E. and I. Wallerstein (1991), Race, Nation, Class: Ambiguous Identities, Verso, London. I. Wallerstein (1995), After liberalism, The New Press, New York. 11 1. Economic Literature on Foreign Finance for Development. An Analysis of Alternative Paradigms 1. The pre-Classical Heritage: the mercantilism Referring to the pre-classical literature on economic growth that could be considered remarkable in terms of the foreign finance issue, we should first mention the age of mercantilism. During the XV century, when international trade increased together with development of shipping companies, merchants, bankers and cities, Spain achieved the international supremacy through the Invincible armada and huge inflows of gold and silver from its Latin American colonies. At the same time, the letter of exchange and bank notes were introduced. These last five centuries have been important in terms of long historical processes of transformation (Polanyi’s framework), world-systems and world-economies (Braudel’s and Wallerstein’s approaches), hegemonies (in the Gramscian sense)1 . The age of mercantilism was the historical context of the fundamental political conflict and the doctrine which it gave birth was the early process of State formation, in favour of a market-oriented economy. Mercantilism can be considered the political economy of Statebuilding2 and the birth of a theory of capitalism based on international trade and foreign finance as the engine of development. For mercantilists – John Hales and Thomas Mun in the United Kingdom, Antoine de Montchrestien and Barthélemy Laffemas in France, Ortiz in Spain, Giovanni Botero and Antonio Serra in Italy -, wealth is essentially profit, that is produced by merchants and manufacturers and that produces, through accumulation, new profits. Any surplus of workers supply and the abundance of capital can facilitate credit and finance for trade and manufacturing. This virtuous circle corresponds to the interests of the State: the abundance of capital and workers is the end of the State (and means for merchants), increase of trade is means of the State (and it is end for merchants). Trade is crucial and it is based on the existence of a big amount of money, assuming that circulating money is good money. Even though the presence of bad money (coins adulterated in their metal content)3 and the huge inflows of gold and silver provoked an inflationary pressure in Europe during the XVII century, nevertheless mercantilism continued to support the idea of abundant money, as it implies less expensive credit opportunities. As indicated by Thomas Culpeper 4 in 1621, if the interest rate for credit lines is 10 percent, then all the 1 The theory of hegemonic stability asserts that an open world economy requires a dominant global power for its smooth functioning. See Kindleberger, C. (1973), The World in Depression, University of California Press, Berkeley. 2 See Hettne, B. (1995), International Political Economy. Understanding Global Disorder, Zed Books, London. 3 Linked to this phenomenon was the so called Thomas Gresham’s law: when two coins are equal in debt-paying value but unequal in intrinsic value, the one having the lesser intrinsic value tends to remain in circulation and the other to be hoarded or exported as bullion. 4 The mercantilists were «Pamphleteers» rather than a school of thought. They had no systematic, comprehensive, consistent treatise, no leader, common method, or theory. Each «mercantilist» sought advantage for a specific, trade, merchant, joint-stock company or social group. Thomas Culpeper (1578-1662) was one of the major representatives of «bullionists», who emphasised gold and silver as «wealth» and considered crucial the acquisition of gold and silver through «favourable balance of trade», colonies or conquest (colonies to supply resources or gold). See R. Larry Reynolds (1998), Mercantilism: An Outline, Boise State University, mimeo. 12 investment options having an expected rate of return lower than 10 percent are abandoned. When money and financing are abundant, then trade and private profits increase as well as State enhances its power. Money (gold and silver) is important, thus to increase reserves is important; if you can’t produce it, you have to import it from abroad through trade. Contrary to Machiavelli who said that a well-organised government should require a rich State and poor citizens 5 , history demonstrated that bourgeois merchants and European States developed together and the new idea of a close link between trade interests of nations, political success of sovereign powers and military results prevailed. This implied a shift from the State as the end of human life to the wealth as a value itself. Society must develop, and development is economic growth and wealth; thus, both the means and the ends are economic ones. Man is understood as «economic man». If Machiavelli’s Prince appears to be addressed to the powerful, to the palazzo; then a specific economic theory – the Mercantilism – developed to support the new riches and the domination of societies by the market principle. We can consider this as the theoretical and historical pre-condition of what Paolo Sylos Labini calls the first stage of modern capitalism (being referred to the XVII and XVIII century) 6 . As the market system penetrated all spheres of human activities, social structures were eroded and redistribution had to be guaranteed in order to let people satisfy their social needs. Polanyi described these changes as «double movement», the first part being the expansion and deepening of market exchange and the second part being the social reaction originating mechanisms of economic integration as the modern welfare system7 . Antoine de Montchrestien, who published the first treatise on political economy in 1616 8 , recapitulated the basic ideas of mercantilism. Private profits are based on international trade which is enhanced by the abundance of silver and gold, that is money; the abundance of gold and silver keeps interest rate low and, when domestic endowment of gold is not adequate, then international flows are needed. Thus, if development is economic growth, trade is an engine of growth requiring the existence and circulation of a big amount of money; then foreign finance is directly involved in the process of development. In this connection, it is also relevant to note that mercantilists have sought to recommend a balance of trade surplus, which implied an interventionist State for regulating trade, to compensate the huge amount of gold and silver inflow. 2. The counter-mercantilism If the foundations of international capitalistic trade were laid during the XV and XVI centuries, in 1588 the Spanish Invincible armada was destroyed and Spain lost its supremacy. One century later, at the beginning of the XVIII century, the success of capitalism was rapid, supremacy passed to England, which promoted financial speculation9 . 5 See Machiavelli, N. (1518), Discorsi sopra la prima deca di Tito Livio, 1954 edition, Ricciardi, Milano. 6 See Sylos Labini, P. (2000), Sottosviluppo. Una strategia di riforme, Editori Laterza, Bari. 7 See Polanyi, K. (1944), The Great Transformation. The Political and Economic Origins of Our Time, Beacon, Boston. 8 See de Montchrestien A. (1616), Traité de l’’économie politique, Paris. 9 Financial speculation means the purchase or sale of financial assets for the sole purpose of making a capital gain. An important component of this phenomenon was the existence of the South Seas Company, created in 1771, which managed public debt. 13 In 1767, James Stuart, in contrast to the mercantilists, observed that if international trade becomes permanent, the consequence is a financial inflow, which can not be considered automatically as a positive phenomenon. It provokes an increase of domestic demand of consumption, which induces: (a) increase of imports, (b) higher cost of manpower, (c) demand of luxury goods, which are produced abroad, creating dependence on foreign production and altering domestic equilibrium 10 . In the meantime, David Hume, one of the founders of liberal theory, criticised another aspect of the mercantilism: the abundance of money is not the cause of reduction in interest rate, which depends on the amount of profits from trade and manufacturing. Purchasing power of money is inversely proportional to quantity: the abundance of gold and silver is unimportant to a State, as its only effect in the long run is the increase of prices, which determines a reduction in exports. Thus trade accounts are in balance again, stopping the inflow of gold 11 . Prior to the founding of the classical school, another economic school, in the XVIII century, analysed the process of economic growth in France, contrasting the mercantilist vision of development. The physiocrats criticised economic policy in France, which discriminated against agriculture, but above all studied the role of capital in economy. Economic activity as a whole is based on the existence of capital and of expenditure, that is advances through capital. Capitalism is based on a permanent circulation of capital: incomes generated through production are translated into expenditure, thus capital is regenerated through sale, permitting an expansion of output also. Thus, development of capital investment is the basic condition of growth12 . Concerning the role of finance for development, the importance of capital was reaffirmed. All these theories paved the way for the emergence of a new theory concerning the causes of growth, including also a new elaboration of the opportunities and constraints to foreign finance. 3. The Classical Heritage. The age of Smith and Marx Adam Smith, the father of the classical school of liberalism, published the Wealth of Nations in 1776 and observed that growth in output and productivity rely upon investment in capital, which can occur if and only if there is a prior increase in savings. Smith did not consider foreign finance, and said that only merchants and manufacturers, who have sufficient income and motivation to save, can achieve increased savings. Smith recommended eliminating all the barriers to free-trade and competition: any intervention to direct capital is an artificial diversion from its natural usage, which spontaneously follows the more productive opportunities. The promotion of infant industries through the use of tariffs to cut imports alters – in Smith’s opinion - the mechanism of market: the correct way is to buy less expensive goods abroad and concentrating domestic production into products linked to the national endowment of factors, used to buy those goods abroad. An automatism-oriented perspective seems to prevail in Smith’s approach: we should avoid to reduce national income, that is the path of capital accumulation, through interventionist actions of State and remember that available capital can be always devoted to domestic productive usage. When the stage of national development has become particularly 10 See Steuart, J. (1767), An Inquiry into the principles of Political Economy, London. 11 See Hume, D. (1752), Writings on Economics, 1955 new edition, Madison, US. 12 See Quesnay, F. (1758), Oeuvres, 1888 Ed. Oncken, Paris. 14 advanced, then a surplus of capital may naturally occur, and it leads to foreign flows of capital to promote industrial growth abroad, making the usage of capital more productive. In the third chapter of his Wealth of Nations, Adam Smith describes the natural evolution of capital. At the beginning, it is generated by agriculture and it is invested there. Then, as capital profitability in agriculture decreases, it is devoted to manufacture. In the end when industry itself is saturated with capital international markets become the convenient place to allocate capital in surplus 13 . During a period of great optimism on the future of domestic and international economy, which prevailed in England and France 14 , a minority of scholars expressed serious doubts on the benefits of liberalism and mercantilism. Most of them lived in Germany, at that time a backward country compared to England and France. Johann Gottlieb Fichte, who lived in Berlin in the earlier XIX century, introduced the concept of the «closed trading State». International trade increases poverty in backward countries, as people migrate abroad, lands remain uncultivated or are sold to foreign enterprises who are interested in speculation, State receives money from selling manpower abroad as well as from a sort of subsidy given by industrialised States. The only alternative for backward countries to promote development is to interrupt all the international relations 15 , because foreign trade and finance become a way to impoverish poor countries. In 1800, Fichte published a book, which used a lot of arguments developed also by Hamilton in the United States and List in Germany, to advocate protection as a means of promoting the economic development in the backward countries and to be cautious on foreign finance. But these arguments have received a very scarce attention by the mainstream of economic theory16 . Concerning the importance of foreign finance, Thomas Robert Malthus, who was not so relevant in terms of theory of value and income distribution, identified the crucial role of effective demand in the determination of development. It is not sufficient to have domestic or foreign capital in order to promote development, the prerequisite is the effective demand, that is the existence of purchasing power to buy what is produced through the usage of capital. It is not true that supply creates its demand, capitalism needs a preventive enlargement of demand. It is not true that foreign finance automatically generates development, it is important the enabling environment of the system17 . This idea, which did not represent the core of international financial relations in the past, has become one of the milestones of current international financing institutions. Surveillance and lending are closely interlinked, the availability of conditional financing can strengthen the leverage of financing institutions and provide the incentive for borrower countries to follow sound policies to create an enabling environment. Coming back to the XIX century, based on the English classical economy, particularly on David Ricardo, Karl Marx tried to demonstrate that current economy is an historical mode of production, inherently based on the exploitation of manpower by capital. In his dualistic vision, capitalist class control production, whereas the working-class is condemned to be 13 See Smith A. (1776), Wealth of Nations, Pelican Books, London (1974 ed.). 14 At the end of the XVIII century, England could base its development on the industrial revolution and its innovations: in 1765, Hargreaves introduced spinning jenny; in 1771, Arkwright used the Water-Frame, a primitive power spinning machine driven by waterpower. 15 An ante litteram idea of «déconnexion», re-written in the 1970s by the neo-Marxian Samir Amin. Amin, S. (1985), La Déconnexion – Pour Sortir du système mondial, Éditions La Découverte, Paris. 16 See Léon X. (1958), Fichte et son temps, vol. II, Paris. 17 See Malthus, T. (1798), An Essay on the Principle of Population, Ward Lock, London (1890 ed.). 15 proletarised. In capitalism there are two classes, with opposing interests. The core of capitalism is the relationship between capital and labour, and there is no room for a peaceful shift in the distribution of power. However, some internal contradictions of the system occur, as profits require cheap labour and, simultaneously, rising output requires an expanding demand to be absorbed. Capitalism has an inherent tendency to periodic crises of overproduction and under-consumption. The objective of the capitalist class is to accumulate wealth in the form of exchange value, not to produce use values. The capitalistic exchange relationship is the opposite to the precapitalist modes of production (based on the following sequence: production of commodities, exchanged for money, which is transformed into other commodities: C M C). The capitalistic production process begins with a stock of financial capital, that is exchange value, with which can buy factors in order to produce commodities, which are sold to realise an augmented exchange value (M C M’). Production generates the surplus value (i. e. the difference between M and M’), the extraction of which is based on the rate of exploitation of labour 18 . Marx was marginally interested on the colonial States, but he encouraged international flows of foreign finance in order to destroy stagnant pre-capitalist modes of production, establish a capitalist mode, develop the productive forces and prepare the way for a transition to a socialist workers’ State and, then, to a communist society. When capital in surplus goes abroad, it creates the conditions in which the rate of profit tends to rise again, contrasting a natural decline in the rate of profit due to capital intensification. As in Smith’s case, Marx’s thesis was that foreign markets represent an emergency market to prevent system from crises of overproduction and declining rate of profit, thus accepting the basic idea that the amount of saving and investment solely depend on the average rate of profit, an idea earlier formulated by David Ricardo and based on the «loi des débouchés» of Jean-Baptiste Say19 . Following Malthus, Marx severely criticised Say and Ricardo, because they did not consider the risk of overproduction and did not understand the usage of foreign finance to balance the system. In 1917, Vladimir Il’ic Ul’janov Lenin published a work to update the Marxian theory. Capital has accumulated faster than the outlets for profitable investment, thus banks and private companies used imperialism to keep the rate of profit high, the colonies became outlets for surplus finance capital, as well as source of raw materials and new markets 20 . In Lenin’s analysis, imperialist capitalism differs from the endogenous capitalism described by Marx. The imperialist capital has suppressed the emergence of local one. Foreign investment of equity and loan capital in infrastructure, plantations and mines have produced profit and interest, preventing from the emergence of local capitalism, as well as crowding out local manufacturers. Lenin, much more than Marx, paved the way for the neo-Marxist theory of development. 18 See Marx, K. (1887), Capital, Vol. I, Lawrence and Wishart, London (1970 ed.). 19 Supply creates its demand, the production as a whole is always sold, and thus saving is automatically translated into investment, without being associated with enlargement of markets. The automatic mechanism of market is perfect: through a decrease in the rate of profit, the risk of capital in excess is always avoided. 20 See Lenin, V. (1917), Imperialism, the Highest Stage of Capitalism, Progress Publishers, Moscow (1975 ed.). 16 4. The neo-classical Heritage From the end of the XVIII century, most economists devoted their studies to the problem of the efficient allocation of resources. It was a net shift of the mainstream economics from the long-run perspective to firms and consumers behaviours and the role of free markets. William Stanley Jevons, Alfred Marshall, and John Hicks in England, Kenneth Joseph Arrow and Gérard Debreu in the USA, Carl Menger and Eugen von Böhm-Bawerk in Austria, Léon Walras in France, Vilfredo Pareto in Italy have been very significant theorists in the neoclassical perspective. The neo-classical theory is based on a short-run, efficiency-maximisation of partial and general equilibrium approach, which is linked to the theory of comparative advantages. The neo-classical consequences concerning development and foreign finance are that if there is a positive propensity to save and to invest in excess of the capital replacement, then sustained growth is possible. The rate of investment will be determined by the social rate of time preference (which, given the rate of interest, determines the supply of savings) and the marginal productivity of capital (which, given the rate of interest, determines the producers’ propensity to invest). In terms of foreign finance, as price distortion leads to a distorted allocation of resources – reducing efficiency -, any form of subsidy (like soft loans) are market distortions and they must be removed. Technical progress is the real source of longrun income growth, as the increase in saving itself can have a positive effect on the growth rate only temporarily (it lowers the price of capital). In 1933, a Swedish economist, Bertil Ohlin, introduced the so-called Heckscher-Ohlin theorem, a neo-classical interpretation of Ricardo’s law of comparative costs. It is true that there is an advantage enjoyed by a country in the cost ratio of one commodity to another in comparison with the ratio of costs of these same commodities elsewhere. However it should not be explained in terms of «real» costs (linked to the labour content, in Ricardo’s opinion), but in terms of relative prices of factors, as in neo-classical approach. If a country has a huge amount of capital, compared to other factors endowment, then the rate of interest of capital is domestically lowered, and the country can opt for investing abroad its capital or for importing other factors, assuming a world of perfect mobility of factors. Neo-classical theory adopted, with the Ricardian law of comparative cost, also the classical theory of automatic equilibrium of the balance of payments. There are some basic assumptions that are implicit in the neo-classical theory of capital accumulation: (1) the growth of production is due to investment, which is financed by savings in advance; (2) those who save become the owners of new capital goods, and their ownership is represented by the interests, which they receive for the savings in advance; (3) the entrepreneur and the banker receive a payment for their job, as free competition eliminate any profit. These assumptions are linked to the idea of perfectly competitive markets, profit maximisation by firms, utility maximisation by consumers, an infinite range of production technologies. These assumptions generate the proposition that free market equilibrium prices are the correct prices for the purpose of guiding resource allocation. Thus, relative factor prices determine factor income shares, whereas any government intervention to determine income distribution is a distortion. Hence the key policy recommendation is to remove all market distortions, in order to improve efficiency and welfare, let international capital be free to move all over the world, without distortions or rigidities. 17 All these assumption are removed by Joseph Schumpeter. 5. The Schumpeterian and Keynesian Heritage In 1911, Joseph Schumpeter introduced a clear distinction between economic growth and development. As briefly described the classical economists based growth on saving and investment, considering growth as additional expansion of production and market. This is exactly growth, added Schumpeter, whereas development is a revolutionary process, which transforms the conditions of production, markets and institutions. Development implies innovation, and it must be studied in terms of the dynamics of long-run economic change. The accumulation of capital in an incremental way that can not produce economic development, if it is not embodied in the mobilisation of existing factors. Capital does matter, if it interacts with new combination of factors of production and new institutions. The capitalist, that is the banker, who creates finance capital, requires the existence of innovative entrepreneurs, who create a new combination of factors. The basic condition to promote development is not the mobilisation of savings to finance the accumulation of productive capital, but the innovative action of entrepreneurs in mobilising credit towards innovative combinations of factors 21 . The classical and neo-classical economists said that savings are crucial to finance development; Schumpeter rather stressed the importance of credit creation. The neo-classical theory said that the rate of interest is the price of savings; Schumpeter rather said that the rate of interest is paid to the bank, not to the saver. The mechanism of capital accumulation is completely changed: innovation (linked to entrepreneurs and remunerated by profit) and finance (linked to banks and remunerated by rate of interest, which rests on part of profit) are the bulk, whereas the savers are excluded. In the context of international development, we could derive the basic assumption that foreign finance can not be de-linked by the crucial role of investment capacity and entrepreneurial culture. A manifest truth which has been widely forgotten during the last thirty years as we will try to demonstrate in part III. During the 1930s, John Maynard Keynes, as well as Michal Kalecki, emphasised the importance of effective demand and the fact that, during recession phases, deficit spending22 can be very useful. Keynes based his theory on the comparative static, and he analysed the macro-economic equilibrium aspects of the market mechanism, taking into account the problem of unemployment. Even though his studies were not focused on development of developing countries and on foreign finance, the Keynesian heritage has been influential into the core mainstream of economics and, directly or not, also into development studies. His only major work in what would now be called «development economics» was Indian Currency Reform, based on a plan worked out at the Indian Office 23 . Neo-classical development economists have neglected the importance of Keynes. But all of us should recognise that the importance of macro-economic policy (rather than the priority to micro-economic efficiency), the crucial role of investment and aggregate demand as the engine of growth, the positive role of State, the need of international institutional reforms (both trade and finance), clearly linked to the Keynesian heritage, remain on the agenda of international relations. 21 See Schumpeter, J. (1911), The Theory of Economic Development, Oxford, (1961 ed.). 22 The creation of a government budget deficit for the purpose of influencing economic activity, producing new demand. 23 H. G. Johnson (1977), «Keynes and the Developing World», in R. Skidelsky (ed.), The End of the Keynesian Era, MacMillan, London. 18 6. The Harrod-Domar model The birth of modern growth theory, in the work of Roy Forbes Harrod and Evsey David Domar, has been one of the by-products of Keynes’s General Theory that have been more directly linked to development economics. A common purpose of their independent analyses was to give a dynamic dimension to Keynesian approach, transforming a short-term dimension (where the implications of the level of investment and saving for the stock of capital are ignored: the stock of capital is an exogenous variable and investment is considered as influencing the effective demand) into a long-period approach. In this perspective, investment expenditure influences the stock of capital. If Schumpeter criticised the idea that an economy can sustain a steady rate of growth for an indefinite period (the idea of a steady growth state), Harrod and Domar considered the conditions to realise continuous growth with full employment 24 . Roy Forbes Harrod 25 , assumed the Keynesian savings function, where savings (=S) is a function of output (=Y). Savings is the sum of spending power which people plan to withhold from consumption and can be made available to finance new capital formation. At any time t, S is a constant proportional function of output: S(t) = s Y(t); 0<s<1 (1.1) where s is the average (and marginal) propensity to save. In order to ensure a perfect balance between demand and supply at aggregate level, ex-ante identity between investment (=I) and savings (=S) is required at any time t: I(t) = S(t) (1.2) Being the Keynesian condition for equilibrium in the goods market. Supposing that the investment function is governed by the acceleration principle 26 , we have: I(t) = vY(t+1) – Y(t) (1.3) 27 Where Y(t) is the output, Y(t+1) - Y(t) is the instantaneous rate of change of output that is expected with respect to time, and v is the desired incremental capital-output ratio (ICOR), representing the value of the capital needed to produce a given output divided by the value of that output, that is the acceleration coefficient, also assumed to be constant. If the expectations have to be realised, then the expected Y(t+1) must be equal to the real Y(t+1). Remembering that: Y(t+1) – Y(t) = ∆Y (1.4) In this case, we are treating time as a discrete variable, and ∆Y represents the change in Y between two successive periods. Otherwise, we could treat time as a continuous variable, thinking in terms of point of time, and use the derivative with respect to time (=dY). We get: 24 See G. Hacche (1979), The Theory of Economic Growth, London, MacMillan. 25 See R. F. Harrod (1939), «An Essay in Dynamic Theory», Economic Journal, vol. 49 and R. F. Harrod (1948), Towards a Dynamic Economics, London, MacMillan. 26 An increase or decrease in income induces a corresponding change in investment; thus the acceleration coefficient is the ratio of increase or decrease in investment to an increase or decrease in income. 27 Y/Y and K/K are the rate of change in the proportion to the level, that is the proportional rate of growth of Y and K. 19 s Y(t) = v ∆Y (1.5) which implies s/v = ∆Y/Y(t) = G(t) (1.6) This is the crucial equation of Harrod’s dynamic theory, combining the acceleration principle and the multiplier theory28 . Given (1.2) the proportional rate of growth (=G) of output (=Y2 ) is: G(t) = [I(t)/Y(t)]/[I(t)/Y(t)] = [S(t)/Y(t)]/[K(t)/Y(t)] (1.7) where K is the volume of capital (both fixed capital and stock). The identity: G(t) = [S(t)/Y(t)]/[K(t)/Y(t)] (1.8) is the Harrodian identity: the growth rate of output can be expressed as the ratio of the proportion of income saved to the ICOR. G(t) is an equilibrium growth rate, and the identity assumes that the proportion of income saved (s Y(t)), matches the average propensity to save of the community, and the ICOR satisfies investment demand. Starting from an equilibrium, and given constant and positive s and v, there is a unique growth path of output (and investment and saving), along which all the core variables of the system (that is Y, I, S) grow at the constant rate s/v. Assuming that the labour force (=L) is growing at a constant exogenous rate (λ) and that labour productivity, that is output per head (y), is increasing at a constant exogenous rate (τ), then the necessary and sufficient conditions to maintain full employment is full employment at the starting point and output growth at a rate equal to λ + τ. Thus, labour market equilibrium requires that growth rate which allows full employment (so-called natural growth rate, n, that is the maximum rate of growth allowed by population growth and labour-saving technical innovation) to be attained and maintained. The unique equilibrium growth path is when growth is both warranted (actual investment and saving = desired investment and saving = G w) and natural (full employment): G(t) = Gw = s/v = n = λ + τ (1.9) Obviously, in the real world, there is no automatic mechanism whereby warranted growth and natural growth would tend to coincide, as well as the constancy of s and v are simplifying assumptions. A growth path along which the capital-output ratio (K/Y) is constant is sometimes referred to as a balanced growth path. A particular balanced growth path is a path along which each economic aggregate grows at a proportional rate which is constant over time (=G), except the growth rate of the labour force. This growth path is the steady-state growth path. The basic contribution of Harrod dynamic theory of development remains the fact that, starting from the multiplier and accelerator principles, the warranted growth rate is the way which guarantees that supply adequately responds to demand expansion. In such a way, the engine of development is the dynamism of the demand side, rather than the supply side.This is a typical Keynesian vision of development. 28 The multiplier is the ratio of the income held to result from an addition to investment to the amount of such addition. 20 Evsey David Domar29 adopted the same proportional saving function as Harrod (1.1) and the equilibrium condition (1.2). He added a proportional relationship between the rate of increase of productive capacity (full employment output = Y) and investment: Y =σI (1.10) σ I is the aggregate net investment multiplied by the output-capital ratio. For savinginvestment equilibrium and full employment, income and investment must grow at a constant rate which is the product of the parameters s and σ. Y /Y = I / I = sσ (1.11) The equilibrium growth rate, sσ, is similar to the Harrod’s warranted growth rate s/v. In fact, σ embodies the influences of labour-force growth and technical progress in productive potential, and the idea of optimal ICOR. σ refers to an increase in capacity which accompanies investment; its magnitude depends on technological progress. Since σ is simply the inverse of the capital-output ratio, Domar’s equation is the same as Harrod’s. Domar emphasised potential growth dependence on capital accumulation, whereas Harrod emphasised the interaction of the multiplier and the accelerator principles to show the lack of any automatic tendency for warranted growth to conform with full employment. Adopting the Keynesian multiplier, without introducing any theory of what determines investment, Domar’s model shows what the path of investment is required to be. As in the Harrod model, Domar model implication is that the automatic maintenance of equilibrium steady-state growth, with full employment, is unlikely to be the outcome of a free-market capitalist economy, extending Keynes’s analysis to a dynamic economy. If actual growth rate, equal to the warranted rate of growth, is below full employment, then steady growth may occur, but any divergence of actual growth rate from the warranted one will lead the economy away from the steady state growth path. Cyclical swings are inevitable unless government policy intervenes. Nevertheless, the savings rate, together with the capital-output ratio, is considered as one of the determinants of growth. The Harrod-Domar growth formula has been widely adopted and used for calculating target rates of investment in economic planning and for estimating foreign aid requirements. The savings constraint is considered dominant as well as in Western development paradigm of Modernisation. And foreign capital inflows make it possible a poor economy to escape from a stagnation vicious circle 30 . 7. Western development paradigm of Modernisation 31 During the 1940s and the 1950s, the paradigm of the expanding capitalist nucleus was the dominant approach, contrasting with the alternative structuralism approach, spread in Latin America. Both of them rejected the neo-classical paradigm (in particular, perfectly competitive markets, perfect divisibility of factors and products and the absence of significant technological or pecuniary externalities) and the static theory of comparative advantages as a basis for analysing the development process. This growing interest among the economists in 29 E. D. Domar (1946), «Capital Expansion, Rate of Growth, and Employment», Econometrica, vol. 14 and E. D. Domar (1947), «Expansion and Employment», American Economic Review, vol. 37. 30 W. T. Newlyn (1977), The Financing of Economic Development, Clarendon Press, Oxford. 31 A general review that traces the theoretical history, since the 1940s, of those who analysed economic development is Hunt, D. (1989), Economic Theories of Development: An Analysis of Competing Programs, Harvester Wheatsheaf, New York. A review of development co-operation theories and policies, referred to the same period, is Zupi, M. (ed.) (1997), Teorie dello sviluppo e nuove forme di cooperazione, Molisv Ed., Roma. 21 the Western countries depended on the purpose of containing the expansion of communism and preserving the international stability. In 1943, Paul Narcis Rosenstein-Rodan, a Polish economist who lived in UK and served as a World Bank and United Nations’ consultant, proposed a strategy for the post-war economic development of Western Europe32 . Following Adam Smith, he emphasised the importance of expanding markets as a stimulus to growth, and he stressed the need to base the industrial development strategy on specialisation and integration into world economy, which permits the mobilisation of international capital to fund part of the development effort with loans. These loans need to be repaid from export revenues. State planning and organisation of a large-scale investment programme would be needed to help mobilise the finance for a programme of industrialisation, and State intervention is requested to guarantee abroad loans and to promote exports in order to permit loans repayment. Rosenstein-Rodan emphasised market failure (the small size of the domestic market and the inability of individual firms to internalise the value of the external economies that they generate are important constraints) and the need for State interventionism. In the late 1940s and 1950s the successful implementation of the Marshall Plan for economic reconstruction in Europe generated confidence in the role of foreign finance and economic aid, and had a significant influence on the approach to development. The dominant constraint to development were seen as internal rather than external: the low rate of saving out of national income, and the need to rise GNP so that the benefits of growth would trickle down to the population. The importance of the role of savings and capital accumulation derived directly from the classical growth theory and from Harrod’s and Domar’s model. This model was not interpreted as the search for the equilibrium growth rate given existing savings propensities, but as a way to raising the savings rate in order to warrant higher growth. Nurkse 33 supported the Rosenstein-Rodan’s idea to encourage the mobilisation of both domestic and foreign finance to support the investment programme. What was considered important was that the government should ensure effective use of foreign aid, through curtailing the growth of domestic consumption. Foreign aid is not able to raise automatically investment, because of the fungibility of foreign resources. Harvey Leibenstein 34 stressed the importance of foreign aid to escape from the low level equilibrium trap, where people have a high marginal propensity to consume stimulated by the demonstration Duesenberry effect – emphasised by Nurkse -. Finance is a way to break out the trap into cumulative growth if it really promotes productive investment. Greater emphasis to autonomous capital accumulation in the private sector was given by complementary contributions of Lewis and Rostow. William Arthur Lewis 35 assumed that monetary expansion can promote growth in a country with surplus labour, when: (1) credit is provided to private capitalists or to finance government capital formation, and (2) the country can contain the pressure on the foreign balance, while ensuring a significant increase in investment. Thus, the opening up of the Lewis model is clearly compatible with the encouragement of capital import. 32 P. Rosenstein-Rodan (1943), «Problems of Industrialisation of Eastern and South-Eastern Europe», Economic Journal, June-September. 33 R. Nurkse (1952), «Some international aspects of the problem of economic development», American Economic Review, May. 34 H. Leibenstein (1957), Economic Backwardness and Economic Growth, Wiley, London. 35 W. A. Lewis (1954), «Economic development with unlimited supplies of labour», Manchester School, May. 22 The historical evidence of the inevitability of capitalism was summarised by the influential work of Walt Withman Rostow36 . Like Lewis, for Rostow the crucial factor needed to takeoff an economy into sustained growth path is a significant increase in the share of savings and investment in national income. In both Lewis and Rostow’s works a prevalent emphasis was given to the role of capital accumulation in economic development, interpreted as a cumulative process. The central assumption was that the capitalist class has the highest propensity to save in underdeveloped countries; in order to raise the overall savings rate it is needed to increase the capitalists’ share of national income. Moreover, if capital is scarce, then the important thing is to maximise its productivity. Based on the emphasis of the Lewis model on the savings constraint and on the complementary work of Rostow, from the mid-1950s simple savings gap models for foreign aid requirements have emerged. External capital requirements were the difference between total capital requirements and domestic savings. These ideas were discussed as the country’s absorptive capacity issue, that is the capacity to absorb effectively an increase in investment made possible by financial aid and foreign loans. Rostow himself said that foreign capital should be made available only when it can be effectively used. Millikan and Rostow37 attempted to estimate foreign aid requirements by considering both absorptive capacity and the savings-investment gap. Rosenstein-Rodan38 mentioned three indices to measure the absorptive capacity, in order to project the foreign aid requirements: (i) the rate by which a country has succeeded in increasing the volume of investment in the past five years; (ii) an estimate of the extent to which a country has succeeded in raising the deviation between the marginal and average propensity to save over the past five years; (iii) a judgement on the country’s overall administrative organisation39 . 8. The Structuralist paradigm of development After the Great Depression in the 1929, in Argentina exchange controls were introduces, as well as import controls to protect the balance of payments and debt repayment. These measures were needed to face the changes in trading conditions: as world prices of primary products declined relative to manufactured goods, Argentina’s import capacity fell. In Argentina, Brazil and Chile there was a rapid growth of national industry in response of the shortage of imports. In that period, it was widely accepted that industrialisation was the key to economic development and that it could not be promoted by concentrating economy on primary exports in exchange for manufactured imports, but by a dynamic cumulative process. The Latin American experience demonstrated the possibility to promote industrialisation through management of imports: the introduction of political measures to change the composition of imports from consumer to capital goods. In 1948, the Economic Commission for Latin America (ECLA) was created, and its director became the Argentinean economist Raoul Prebish. ECLA economists rejected traditional trade theory as a basis for national economic policies and developed a new body of theory. It was based on the need of imposing improved conditions of international trade for primary products and supporting initial periods of 36 W. Rostow (1960), The Stages of Economic Growth: A Non-Communist Manifesto, Cambridge, Cambridge University Press. 37 R. Mikesell (1968), The Economics of Foreign Aid, Weidenfeld and Nicolson, London. 38 P. Rosenstein-Rodan (1961), «International aid for underdeveloped countries», Review of Economics and Statistics, Vol. XLIII, No. 2, May. 39 The so-called gap-models are discussed in chapter 2. 23 import-substituting industrialisation national policies inducing structural transformation, focused on the development of a diversified domestic industrial sector 40 . “Imports are specified as a linear function of output, implying that fixed amount of imports are required in production (that is, there is a constant marginal propensity to import). This assumption, together with that of exogenous export growth, make this a structuralist model” 41 . With respect to international finance, Karl Gunnar Myrdal42 emphasised the inter-connection between internal and international causes of under-development. A small group of countries, dominating science, technology, finance and industrial production, a experiencing a path of cumulative development, whereas the majority of countries, with low levels of output and savings, low levels of health and education, are condemned to stagnation or declining per capita incomes, enlarging international inequality. There is a tendency of finance capital to flow into areas where cost structures and market prospects seem attractive, and there is the presence of developing countries’ governments to protect their own interests and export enclaves to prevent international relations from disseminating among people all the potential benefits derived from foreign finance. Thus, national poverty is associated with rising intra as well international inequality. Foreign aid is not sufficient to change the structural inadequacy of system and it risks exacerbating, through the repayment mechanism, the underdevelopment. Another economist who stressed the structural component of development was Hla Myint 43 . He said that a combination of market forces having different power, social institutions and prejudice act to prevent poor people from improving their status; developing countries are borrowers of capital who face monopolistic markets of creditors. The main effort should be concentrated on the search for developing counter-powers to counteract the existing unequal distribution of market power. Also Albert Otto Hirschman developed a theoretical work which was fully consistent with structuralist thinking, even though he did not regard himself as a structuralist, and the ECLA economist did not considered him as a member of the structuralist school. Hirschman44 attacked the balanced growth theories – developed by the theory of modernisation –which identified the causes of underdevelopment in terms of the lack of key factors, be it finance or skilled labour. He proposed a strategy of unbalanced growth, guided by major resources bottlenecks as revealed in the market, he advocated the use of large-scale capital-intensive techniques of production (which do not require high organisational and managerial resources) and favoured foreign capital for its ability to identify successful priority sectors and regions. The inducements to invest in the industrial sector can be maximised if economies follow an investment path in which each stage of investment generates, through backward and forward linkages to other branches of the economy, cumulative inducements to invest in these branches also. Concerned with the criticism to the Western development paradigm of modernisation, the structuralism comment stressed that such a perspective which over-optimistically invoked 40 See R. Prebish (1962), «The Economic Development of Latin America and its Principal Problems», Economic Bulletin of Latin America, Vol. VII, No. 1, February and C. Furtado (1961), Desenvolvimento y Subdesenvolvimento, Editora Fundo de Cultura, Rio de Janeiro. 41 H. White, “The Macroeconomic Impact of Development Aid”, in The Journal of Development Studies, vol.28 n°2, January 1992, p.170. 42 G. Myrdal (1957), Economic Theory and Underdeveloped Countries, Duckworth. 43 H. Myint (1954), «An interpretation of economic backwardness», Oxford Economic Papers, June. 44 A. Hirschman (1958), The Strategy of Economic Development, Yale, Yale University Press. 24 international finance support to development, over-emphasised the supply-side constraint to growth (the rate of capital accumulation, determined by the supply of savings) and ignored the possibility of a domestic demand constraint to capitalist accumulation. It also ignored the possibility of a foreign exchange constraint. Because of the growing technological gap between industrialised and underdeveloped countries, combined with the low income elasticity of demand for primary products in industrialised countries and high income and low price elasticities of demand for imports in underdeveloped countries, it should be stressed the importance of the serious balance of payments difficulties of developing countries, facing a secular deterioration of their terms of trade. A further structuralism criticism emphasised the importance of market imperfections and inequity context which could lead international loans to a growing stock of under-unutilised new capital stock, provisions to rent-seekers, without any real match to the needs of the population and the economy. The structuralism school of thought rejected neo-classical and monetarist45 solutions to the problem of balance of payments disequilibria (domestic deflation and devaluation of the currency, in order to reduce the demand for foreign exchange and increase foreign currency earnings) and recommended policy measures for both developing and developed countries: Developing countries should: (i) cut non-essential imports; (ii) expand primary and manufactured exports, (iii) orient investment on import-substituting sectors, (iv) introduce import control as a counter-cyclical policy, (v) create regional integration. Developed countries should: (i) reduce protectionism against developing countries, (ii) orient foreign direct investment to developing countries, (iii) introduce compensatory finance schemes to assist primary goods exporters to meet their foreign payments obligations, (iv) increase lending to developing countries, (v) introduce systems of preferences for manufactured exports from developing countries. In order to project the foreign aid requirements of underdeveloped countries, being the primary role of foreign aid to help relieve balance of payments pressures, Chenery and Strout 46 employed the first index proposed by Rosenstein-Rodan to measure the absorptive capacity (the rate by which a country has succeeded in increasing the volume of investment in the past five years). They assumed that during the first part of the take-off stage, foreign aid serves to plug a savings gap, provided that marginal propensity to save is not high enough and the incremental capital/output ratio is not low enough to sustain the desired rate of growth domestically, without recourse to foreign finance. In their model Chenery and Strout used the absorptive capacity, investment-savings gap approach and, in the structuralism tradition, foreign exchange gap approach47 . In this context of two-gap model, foreign aid can 45 The monetarism is a school of thought in economics that places money at the centre of macroeconomic policy. Based on the quantity theory of money, it relates the price level to the quantity of money in the economy. It claims that monetary factors are a major influence on the economy and that, in particular, government expansion of the money supply will tend to generate inflation rather than employment. This view is associated with Milton Friedman of Chicago University and has been recently strengthened by the success of the so-called theory of rational expectations, proposed by Robert Lucas jr., Thomas Sargent and Neil Wallace. The monetarism is coupled with neo-classical school in supporting the neo-liberal vision of free market forces, without any restriction on supply or government intervention that can affect demand, supply, or price. In development economics, the monetarism is linked to the story of the Pinochet regime’s economists, known as the «Chicago Boys», who followed their training as economists at the University of Chicago and, after their return to Chile, took advantage of the opportunity afforded them by the 1973 military coup to launch the first radical free market strategy implemented in a developing country. See J. G. Valdés (1995), Pinochet’s Economists. The Chicago School in Chile, Cambridge University Press, New York. 46 H. Chenery and A. Strout (1966), «Foreign assistance and economic development», American Economic Review, Vol. LVI, No. 4, Part 1, September. 47 See chapter 3. 25 help to finance essential imports. For eliminating the foreign exchange gap the two crucial parameters are the rate of growth of export and the marginal propensity to import: if the former rises and the latter falls, then the gap is eliminated48 . Economic development consists not only of raising per capita income through international finance, but also in structural transformation, so that economies acquire the endogenous capacity to sustain development. The central issue is therefore whether international finance can in turn generate an internal demand to induce a process of sustained investment to expand domestic market. 9. The neo-Marxist and dependency approaches In the 1960s, against the structuralism and the optimistic modernisation perspectives, which put emphasis on capital accumulation and focused on domestic problems of capital scarcity and inadequate institutions (which impede savings and investment), radical left criticised the mainstream orthodoxy. They explored Marxist principles and Lenin’s innovations on the class modes of appropriation and use of the actual economic surplus. The neo-Marxist critique of classical structuralism theory has elements in common with the dependency-oriented auto-critique of some of the structuralists themselves. But there is a clear distinction in the analytical concepts and in the logic of argument: the neo-Marxists moved towards the analytical framework of the global development of the capitalist system49 . Paul Baran50 considered underdevelopment as the process of continuing extraction of surplus from the underdeveloped countries and its transference to the centres of world capitalism, through monopsonistic trade and the perpetuation of mass pauperisation. Dominant classes in underdeveloped countries failed to use surplus for productive accumulation within the domestic economy. Andre Gunder Frank 51 adopted Baran’s vision and emphasised surplus appropriation mechanism through trade and the alliance of the domestic comprador bourgeoisie and foreign capitalists. Rather than considering existing economic structures as the causes of underdevelopment – as in the structuralism approach -, the neo-Marxists stressed the importance of class control over the surplus as the main factor inducing underdevelopment. International finance is not useful, if the masses don’t replace the existing ruling class alliances in the periphery as well in the centre of the world and take control of the economic surplus. If underdeveloped economies remain dependent on the world metropolitan economies for access to finance, as a result the latter continues to determine the underdevelopment in the periphery. Samir Amin 52 concluded that the only way to full autonomous development is via socialist revolution or de-linking the underdeveloped countries from the western economies. 48 According to Joshi, apparent foreign exchange constraints actually are a savings constraint: if populations consumed less of domestic output and exported more, they would have more foreign exchange. Two-gap models have contributed to this exaggeration. See V. Joshi (1970), «Saving and foreign exchange constraints», in P. Streeten (ed.), Unfashionable Economics, Weidenfeld and Nicolson, London. 49 Further development of the neo-Marxist paradigm was the analysis of the functioning of the world capitalist system as a whole, of which the periphery in an integral part. Concerning the theoretical debate on international finance, this areas of studies are not relevant. Nevertheless the work of Wallerstein should be mentioned. See I. Wallerstein (1979), The Capitalist World Economy, Cambridge, Cambridge University Press. 50 P. Baran (1957), The Political Economy of Growth, Monthly Review Press, New York. 51 A. G. Frank (1978), Dependent Accumulation and Underdevelopment, MacMillan, London. 52 S. Amin (1973), Le Développement inégal, Les Editions Minuit, Paris. 26 In the 1970s, dependency analyses were carried out in Latin America, focused on a range of unequal and dependent relationship between the countries of the periphery and the industrialised countries. If the structuralists, who favoured macro-economic balance, emphasising the limits of the monetarist approach and the output and welfare costs associated with orthodox stabilisation, otherwise dependency theorists favoured state interventions to correct market failures, and considered the monetarist approach to stabilisation unnecessary and the risks connected to deficit financing exaggerated. 10. A neo-classical revival towards the Washington Consensus In the 1950s most of the economists supported the idea that, in the short-run, governments of developing countries had to try to mobilise large quantities of national and international capital for public development expenditure. Only a few economists followed the neo-classical approach, as Peter Bauer and Basil Yamey53 who said that government should always concentrate upon removing the impediments to private saving and investment, through the maintenance of law and order, political stability, adequate monetary conditions, without interfering with the operation of market forces, but supporting them to operate more efficiently. The focus of neo-classical approach was efficiency in resource allocation and its critique to the programmes of import-substituting industrialisation strategies and to any form of subsidised international financial flow, which altered the true indication of the relative opportunity cost of resources (given by a correct structure of prices reflecting the relative scarcity of capital). Foreign finance is an extraordinary fuel of growth engine, but private flows must be the bulk, without any distortion induced by over-regulations. If during the 1950s and 1960s a lot of neo-classicals had accepted a crucial role of governments of developing countries in the first stages of their national development, from the 1970s the neoclassical critique to the public sector involvement in economy improved dramatically. From a chronological point of view, there have been three important elements which concurred to strengthened this idea: (i) in the 1970s, the perception that transition from pre-capitalist States to capitalism had been achieved in developing countries and private sector could become an autonomous and mature actor to be linked to international private sector, reducing the role of public sector; (ii) in the 1980s, debt crisis reduced de facto the State bureaucracy capacity of resistance to a radical transition to a new State, much more market-oriented; (iii) in the 1990s, the end of communism in the Eastern block eliminated the need of strong (authoritarian) governments in developing countries, used by the Western countries to prevent the world from the risk of widened opposition to capitalism, and this post-bipolarised context created the opportunities for a «light» public sector. In the 1970s, at micro-economic level, neo-classical critique was complemented by the literature on cost-benefit analysis, designed to facilitate greater efficiency in investment decisions. In the 1980s, as it will be illustrated in chapter 4, foreign debt crisis of developing countries represented a main factor, which facilitated a radical shift in the theory and practice of development policies. “Until the 1980s the predominant concerns of development economists were sectoral (...). During the 80s the predominant concern shifted to macroeconomic issues. This shift was not because the sectoral problems had been solved, but because they were 53 P. Bauer and B. Yamey (1957), The Economics of Underdeveloped Countries, Cambridge University Press, Cambridge. 27 suddenly swamped by even more serious macroeconomic shocks” 54 . And “The series of external shocks at the beginning of the 1980s (...) forced macroeconomic issues back into the heart of the development debate” 55 . A paradox occurred: if during the previous decades different paradigms stressed the inadequacy of neo-classical approach to face the challenge of development and encouraged the usage of foreign finance and debt, then the debt itself paved the way for a renewed vigour of the neo-classical revival. This revival was reinforced by the application from developing countries to the IMF and WB assistance with stabilisation plans and structural adjustment programmes. These programmes emphasised control of the money supply and removal of price distortions and the freeing of markets from public sector interventionism. In 1981, the World Bank published an influential report on Africa 56 , which emphasised the importance of domestic correct pricing policies and reduced government intervention in African economies, rather than searching for the New international economic order (NIEO), proposed by the Group of 77 countries, the most important aggregation of developing countries. Foreign finance is an issue on which the neo-classicals expressed a wide range of opinions. The strongest opposition to aid, linked to a rigorous philosophy of laissez faire, was expressed by Bauer 57 in the late 1960s. Aid, based on subsidised fund (grants and highly concessional soft loans), prevents necessary policy reforms in developing countries and encourages the view that developed countries must carry the moral responsibility for the underdeveloped countries. Lal58 sustained the importance of technical assistance, whereas the financial aid is likely to be counter-productive. But most of the neo-classicals argue that financial aid has a positive role to play, and in order to ensure this potential it is needed to adhere to cost-benefit analysis methodology and, rather than supporting specific projects and programmes, aid and loans should support stabilisation and structural adjustment programmes. It means to provide funds not to all developing countries, but only those who accept these conditionalities aimed at creating a performing macro-economic structure. Later, in 1998, the World Bank published a report entitled “Assessing Aid”, as the result of an extensive investigation into the effectiveness of development co-operation59 . The empirical analysis of the report shows that development cooperation helps to stimulate economic growth and reduce poverty, but only in recipient countries with good economic management (or good governance), that is good economic policies and strong institutions. The main conclusion of the report is therefore that development co-operation should be allocated by “selecting” recipient countries according to 54 D. Bevan, P. Collier, J.W. Gunning (1991), “The Macroeconomics of External Shocks”, in V.N. Balasubramanyam and S. Lall (editors), Current Issues in development economics, MacMillan, London, p. 91. 55 H. White (1992), “The Macroeconomic Impact of Development Aid”, The Journal of Development Studies, vol. 28 N. 2, January, pp.168. 56 World Bank (1981), Accelerated Development in Sub-Saharan Africa: An Agenda for Action, World Bank, Washington D.C. 57 P. T. Bauer (1971), Dissent on Development, Weidenfeld and Nicolson, London. 58 D. Lal (1983), The Poverty of Development Economics, Institute of Economic Affairs, Washington D.C., Hobart Paperback 16. 59 World Bank (1998), Assessing Aid: What Works, What Doesn’t, And Why, The World Bank, Washington D.C. 28 their policy environment. This report has led to heated debates, both among academics and policy makers, about the future of development co-operation policies 60 . The monetarist macro-economic approach which underpins the IMF views of the causes of the balance of payments instability and domestic inflation has its foundations in the neoclassical model of economic systems, based on flexible prices, high competition, absence of rigidities. When the growth rate of the money supply exceeds the growth rate of output and incomes, then the demand for goods and services will grow faster, leading to pressure on domestic prices (inflation) and on the balance of payments. The solution is to control the growth of the money supply in order to reduce the growth of the domestic demand. A strategic measure to control money supply is to reduce the public sector deficit, as the government’s need to finance expenditure programmes that exceed tax revenue is the main source of the pressure. To reduce imports and increase exports, using the devaluation of the domestic currency, is important in order to restore the balance of payments equilibrium. An increase in the domestic interest rates can reduce capital outflows and increase savings. Price and exchange rate rigidities, monopolies, taxes, subsidies and trade restrictions are all distortions to be removed, as they reduce short-run efficiency61 . When the stabilisation and structural adjustment programmes faced failures, during the 1980s, neo-classicals ascribed them to attempts to introduce large changes too fast, creating political tensions and uncertainty. The failure of the programmes was due to the lack of implementation of the complete package of reforms. Thus, by introducing some safety-net components, to reduce the negative effects on the poor, Neo-classicals have continued to support their theoretical model, debating on the optimal timing of liberalisation programmes, which only implicitly recognised the importance of structural rigidities and the need for medium and long-term reforms. Today, the influential role of the WB and the IMF makes it harder for governments dependent on these institutions (and their loans) to embark on economic strategies that depart from neo-classical orthodoxy. Partly as a consequence of the conditionality imposed by the WB and the IMF, but also because of widespread domestic dissatisfaction with the structuralism set of import-substitution policies, outward-oriented policies have become the norm among developing countries 62 . A remarkable convergence of views among policymakers and academics has developed on the virtues of the market-oriented model, the neo-liberal approach, which has been called the Washington Consensus 63 . A main limitation of this consensus, as Joseph Stiglitz noted 64 , is that it seeks to make economies adjust to the 60 Chapter 5 analyses the econometric and methodology debate linked to this paper. See also N. Hermes and R. Lensink (eds.) (2001), “Changing the conditions for development aid. A new paradigm?”, The Journal of Development Studies, Special Issue, Vol. 37, N. 6, August. 61 M. Kahn and M. Knight (1981), «Stabilisation programs in developing countries: a formal framework», IMF Staff Papers, Washington D.C. 62 D. Rodrik (1999), The New Global Economy and Developing Countries: Making Openness Work, ODC Policy Essay No. 24, Washington D.C. 63 The origin of this concept lies back in 1989, when John Williamson was expressing his support for the Brady Plan in front of a US Congressional Committee. In his original presentation, Williamson identified ten specific policy prescriptions: (1) Fiscal discipline, (2)Public expenditure redirection, (3) Tax reform, (4) Financial liberalisation, (5) Trade liberalisation, (6) Exchange rates unification, (7) Privatisation, (8) Deregulation, (9) Promotion of FDI, (10) Defence of property rights. Williamson refuses to consider the Washington consensus as a neo-liberal, conservative manifesto. See J. Williamson (1996), “Lowest Common Denominator or Neoliberal Manifesto? The Polemics of the Washington Consensus”, R. M. Auty and J. Toye (eds.), Challenging the Orthodoxies, MacMillan Press Ltd., London. 64 J. E. Stiglitz (1998), More Instruments and Broader Goals: Moving toward the Post-Washington Consensus, WIDER Annual Lectures 2, Helsinki, January. 29 distorted world market rather than to challenge the global wealth and power imbalances, and it paves the way for severe criticism against this approach. 11. The alternative perspective of sustainable human development approach Based on the structuralism and dependency theorists’ perspectives, and moving towards an alternative holistic vision of development, wide criticism mounted since the 1980s against the Washington Consensus approach, which adopted measures that hit the poor hardest, attack the symptoms but not the causes of instability, do not guarantee significant improvement in the balance of payments. Global inequality has risen during the last years, not only because of an increase in betweencountry inequality, but also because of greater within-country inequality. Distribution does matter for poverty reduction65 . Successful poverty reduction strategies depend not only on favourable changes in average GDP growth, but also on favourable changes in income inequality. New poverty means social exclusion. The World Development Report 1990 projected that the total number of poor would have fallen from 1,125 million (1985) to 825 million if developing countries had performed good macro-economic policies (projected figure, 2000). The World Bank assumed a neutral effect of the implementation of the structural adjustment programmes. But, in practice, the Washington Consensus’ measures mean: inflation control by means of large rises in interest rates and budget cut; demand management through a reduction in money supply, wage repression, cuts in public expenditures and revenue raising measures; trade liberalisation inducing increased wage inequality; deregulation of the labour market, through greater wage flexibility, reduced regulation, erosion of minimum wages, lower unionisation; financial liberalisation; erosion of the redistributive role of the State. The reality of past 20 years has been a huge increase in inequality, as demonstrated by recent trends in income distribution, poverty data set and econometric studies. The World Development Report 2000/2001 clearly indicates that the World Bank target has been missed by a large margin: poor have increased to 1,250 million (2000). Through the Washington Consensus’ assumptions, the structural adjustment programs have negatively affected developing countries in a distributionally manner, and inequality trends are likely to depress growth and reduce the poverty alleviation effect of national and international policies, whereas debt burden increased. Since the late 1960s some economists questioned the assumption that an increase in inequality was inevitable in the early stages of economic growth, to be followed later by a tendency towards greater equality. Their unwillingness to accept the inevitability of a tradeoff between growth and equity66 meant to put the problem of continuing widespread poverty 65 G. A. Cornia (2000), Inequality and Poverty in the Era of Liberalisation and Globalisation, UNU/WIDER, Helsinki. 66 In the 1950s, Simon Kuznets suggested that "as economic development occurs, income inequality first increases and after some 'turning point,' starts declining." This idea is now known as the Kuznets U-Curve, which details the changing relationship between economic development and income disparity. In essence, Kuznets found that the distribution of income becomes more unequal at early stages of development but it eventually moves back toward greater equality as economic growth continues. Simon Kuznets originally came to this conclusion by plotting data about the history of economic growth in developed countries. The Kuznets Curve has since been used to predict social inequality in developing countries and to guide policy implementation. The common explanation for this trend is that, initially, only the wealthy have the capital to invest in factors of development and therefore disproportionately benefit from the resulting economic growth. After some time, a delayed trickle down effect eventually occurs and the poorer populations receive the secondary benefits of the economic growth, thereby reducing income inequality. See Kuznets, S. (1951), «LongTerm Changes in the National Income of the United States of America since 1870», in Income and Wealth of the United States: Trends and Structure, International Association for Research in Income and Wealth, Income 30 at the core of the agenda. Academic institutions, branches of the UN and NGOs shared this vision of development. Dudley Seers 67 noted that development should be reinterpreted to take account not only of growth but poverty, income distribution and employment. The 1972 ILO Kenya report68 identified the informal sector as an important source of output growth, employment and productivity gains for the poor. In 1974, a team from the World Bank and the Institute of Development Studies (IDS) at Sussex, guided by Hollis Burnley Chenery69 synthesised and provided more robust analytical foundations to the redistribution with growth approach, endorsing a new enlarged definition of development, assuming different savings rates for rich and poor and, consequently, accepting the idea of a trade-off between the rate of growth of GDP and greater equality, in the short-term. In 1976, the ILO 70 proposed that all countries should give priority to the meeting of the basic needs of all members of their population by the year 2000, as the policies of redistribution with growth might not be sufficient to guarantee an increase in welfare for the 800 million people living in absolute poverty. Basic needs were defined to include the minimal consumption requirements needed for a healthy population, minimal standards of access to public social services, access to employment opportunities to achieve a target minimum income, and the right to participate in decisions affecting livelihood of the people. From the «Basic needs first» strategy of Paul Streeten71 in the 1970s, to the capabilities and entitlement approach of Amartya Sen in the 1980s 72 , and to the Human development approach of the UNDP in the 1990s, all these perspective shared the vision of investing on the human resources and public service provision, emphasising the need also to raise directly the incomes and the «power» of the poor. In this context, foreign exchange savings can fruitfully increase the supply of savings to finance essential imports and can contribute to empower people if, and only if, it serves their interests. These approaches stress the importance of non-economic factors in the development process. It does mean that development is a multi-dimensional process, where political interests are the basic elements to be considered in a long-term perspective, which implies a sustainable nature of the process. International finance and aid can be useful components of development if they are oriented to funding public services, institutional development (research and development, small-scale credit provision) and compensation for land reform. This approach of foreign finance implies a significant attention to be devoted to the time horizon of loans: if long-term strategies need and Wealth, Series II, Bowes & Bowes, Cambridge; Kuznets, S. (1956), «Quantitative Aspects of the Economic Growth of Nations», Economic Development and Cultural Change, University of Chicago Press, No. I, October. 67 D. Seers (1972), «What are we trying to measure ?», Journal of Development Studies, Vol. 8, No. 3, April. 68 ILO (1972), Employment, Incomes and Equality: A Strategy for Increasing Productive Employment in Kenya, ILO, Geneva. 69 H. Chenery (1974), Redistribution with Growth, Oxford, Oxford University Press. 70 ILO (1976), Employment, Growth and Basic Needs: A One-World Problem, ILO, Geneva. 71 P. Streeten (1981), «Development ideas in historical perspective», Development Perspectives, MacMillan, London. 72 Sen, A. (1981), Poverty and Famines: An Essay on Entitlement and Deprivation, Clarendon Press, Oxford; Sen, A. (1990), The Political Economy of Hunger, in 3 volumes (jointly edited with J. Dreze), Clarendon Press, Oxford; Sen, A. (1997), On Economic Inequality, Extended Edition, Clarendon Press, Oxford. 31 to be supported, then loans must be long-term oriented in their repayment schedule. This point is particularly important relatively to the question of debt sustainability73 . International finance must be strictly linked to national poverty reduction strategies and to trade policies and development co-operation, in a coherent framework. It does imply an holistic approach to development. This approach animated the international political debate on the presence of debt relief as a main component of the G8 agenda at the 2001 Genoa Summit and can represent one of the most interesting and original ways to define a new approach to development co-operation and debt relief74 . 73 Time does matter, in every case. For those theoreticians being part of the mainstream economics, attention turned to the objective of the maximisation of short-run efficiency in resource allocation and in management; whereas for those economists who were interested in macro-economic dynamics, the analysis of long-term development path and strategies was the main area of investigation. Clearly, specific theoretical models, concepts, analytical techniques and consequently strategies of intervention developed in a short-term context can not be adapted to the purposes of a long-term framework, and vice-versa. 74 This debate and its recommendations involving NGOs, academics, research centres, trade unions, international organisation are a useful corollary to the current review of competitive theoretical paradigms and they are discussed in chapter 9. 32 Bibliography Amin, S. (1973), Le Développement inégal, Les Editions Minuit, Paris. Amin, S. (1985), La Déconnexion – Pour Sortir du système mondial, Éditions La Découverte, Paris. Baran, P. (1957), The Political Economy of Growth, Monthly Review Press, New York. Bauer, P. and Yamey, B. (1957), The Economics of Underdeveloped Countries, Cambridge, Cambridge University Press. Bauer, P. T. (1971), Dissent on Development, Weidenfeld and Nicolson, London. D. Bevan, P. Collier, J.W. Gunning (1991), “The Macroeconomics of External Shocks”, in V.N. Balasubramanyam & S. Lall (editors), Current Issues in development economics, MacMillan, London. Chenery, H. (1974), Redistribution with Growth, Oxford, Oxford University Press. Chenery, H. and Strout, A. (1966), «Foreign assistance and economic development», American Economic Review, Vol. LVI, No. 4, Part 1, September. Cornia, G. A. (2000), Inequality and Poverty in the Era of Liberalisation and Globalisation, UNU/WIDER, Helsinki. de Montchrestien A. (1616), Traité de l’’économie politique, Paris. Domar, E. D. (1946), «Capital Expansion, Rate of Growth, and Employment», Econometrica, vol. 14. Domar, E. D. (1947), «Expansion and Employment», American Economic Review, vol. 37. Frank, A. G. (1978), Dependent Accumulation and Underdevelopment, MacMillan, London. Furtado, C. (1961), Desenvolvimento y Subdesenvolvimento, Editora Fundo de Cultura, Rio de Janeiro. H. G. Johnson (1977), «Keynes and the Developing World», in R. Skidelsky (ed.), The End of the Keynesian Era, MacMillan, London. Hacche, G. (1979), The Theory of Economic Growth, London, MacMillan. N. Hermes and R. Lensink (eds.) (2001), “Changing the conditions for development aid. A new paradigm?”, The Journal of Development Studies, Special Issue, Vol. 37, N. 6, August. Hettne, B. (1995), International Political Economy. Understanding Global Disorder, Zed Books, London. Hirschman, A. (1958), The Strategy of Economic Development, Yale, Yale University Press. Hume, D. (1752), Writings on Economics, 1955 new edition, Madison, US. Hunt, D. (1989), Economic Theories of Development: An Analysis of Competing Programs, Harvester Wheatsheaf, New York. ILO (1972), Employment, Incomes and Equality: A Sttrategy for Increasing Productive Employment in Kenya, ILO, Geneva. ILO (1976), Employment, Growth and Basic Needs: A One-World Problem, ILO, Geneva. Joshi, V. (1970), «Saving and foreign exchange constraints», in P. Streeten (ed.), Unfashionable Economics, Weidenfeld and Nicolson, London. Kahn, M. and Knight, M. (1981), «Stabilisation programs in developing countries: a formal framework», IMF Staff Papers, Washington D.C. Kindleberger, C. (1973), The World in Depression, University of Califormia Press, Berkeley. Kuznets, S. (1951), «Long-Term Changes in the National Income of the United States of America since 1870», in Income and Wealth of the United States: Trends and Structure, 33 International Association for Research in Income and Wealth, Income and Wealth, Series II, Bowes & Bowes, Cambridge Kuznets, S. (1956), «Quantitative Aspects of the Economic Growth of Nations», Economic Development and Cultural Change, University of Chicago Press, No. I, October. Lal, D. (1983), The Poverty of Development Economics, Institute of Economic Affairs, Washington D.C., Hobart Paperback 16. Larry Reynolds, R. (1998), Mercantilism: An Outline, Boise State University, mimeo. Leibenstein, H. (1957), Economic Backwardness and Economic Growth, Wiley, London. Lenin, V. (1917), Imperialism, the Highest Stage of Capitalism, Progress Publishers, Moscow (1975 ed.). Léon X. (1958), Fichte et son temps, vol. II, Paris. Lewis, W. A. (1954), «Economic development with unlimited supplies of labour», Manchester School, May. Machiavelli, N. (1518), Discorsi sopra la prima deca di Tito Livio, 1954 edition, Ricciardi, Milano. Malthus, T. (1798), An Essay on the Principle of Population, Ward Lock, London (1890 ed.) Marx, K. (1887), Capital, Vol. I, Lawrence and Wishart, London (1970 ed.). Mikesell, R. (1968), The Economics of Foreign Aid, Weidenfeld and Nicolson, London. Myint, H. (1954), «An interpretation of economic backwardness», Oxford Economic Papers, June. Myrdal, G. (1957), Economic Theory and Underdeveloped Countries, Duckworth. W. T. Newlyn (1977), The Financing of Economic Development, Clarendon Press, Oxford. Nurkse, R. (1952), «Some international aspects of the problem of economic development», American Economic Review, May. Polanyi, K. (1944), The Great Transformation. The Political and Economic Origins of Our Time, Beacon, Boston. Prebish, R. (1962), «The Economic Development of Latin America and its Principal Problems», Economic Bulletin of Latin America, Vol. VII, No. 1, February Quesnay, F. (1758), Oeuvres, 1888 Ed. Oncken, Paris. Harrod, R. F. (1939), «An Essay in Dynamic Theory», Economic Journal, vol. 49. Harrod, R. F. (1948), Towards a Dynamic Economics, London, MacMillan. Rodrik, D. (1999), The New Global Econoomy and Developing Countries: Making Openess Work, ODC Policy Essay No. 24, Washington D.C. Rosenstein-Rodan, P. (1943), «Problems of Industrialisation of Eastern and South-Eastern Europe», Economic Journal, June-September. Rosenstein-Rodan, P. (1961), «International aid for underdeveloped countries», Review of Economics and Statistics, Vol. XLIII, No. 2, May. Rostow, W. (1960), The Stages of Economic Growth: A Non-Communist Manifesto, Cambridge, Cambridge University Press. Schumpeter, J. (1911), The Theory of Economic Development, Oxford, (1961 ed.). Seers, D. (1972), «What are we trying to measure ?», Journal of Development Studies, Vol. 8, No. 3, April. Sen, A. (1981), Poverty and Famines: An Essay on Entitlement and Deprivation, Clarendon Press, Oxford. 34 Sen, A. (1990), The Political Economy of Hunger, in 3 volumes (jointly edited with J. Dreze), Clarendon Press, Oxford. Sen, A. (1997), On Economic Inequality, Extended Edition, Clarendon Press, Oxford. Smith A. (1776), Wealth of Nations, Pelican Books, London (1974 ed.). Steuart, J. (1767), An Inquiry into the principles of Political Economy, London. Stiglitz, J. E. (1998), More Instruments and Broader Goals: Moving toward the PostWashington Consensus, WIDER Annual Lectures 2, Helsinki, January. Streeten, P. (1981), «Development ideas in historical perspective», Development Perspectives, MacMillan, London. Sylos Labini, P. (2000), Sottosviluppo. Una strategia di riforme, Editori Laterza, Bari. Wallerstein, I. (1979), The Capitalist World Economy, Cambridge, Cambridge University Press. J. G. Valdés (1995), Pinochet’s Economists. The Chicago School in Chile, Cambridge University Press, New York. H. White (1992), “The Macroeconomic Impact of Development Aid”, The Journal of Development Studies, vol. 28, N. 2, January. J. Williamson (1996), “Lowest Common Denominator or Neoliberal Manifesto? The Polemics of the Washington Consensus”, R. M. Auty and J. Toye (eds.), Challenging the Orthodoxies, MacMillan Press Ltd., London. World Bank (1981), Accelerated Development in Sub-Saharan Africa: An Agenda for Action, World Bank, Washington D.C. World Bank (1998), Assessing Aid: What Works, What Doesn’t, And Why, The World Bank, Washington D.C. Zupi, M. (ed.) (1997), Teorie dello sviluppo e nuove forme di cooperazione, Molisv Ed., Roma. 35 2. Foreign debt and development 1. Introduction How much does foreign debt matter for development process? From a very general point of view it is important that external debt does not fall in a spiral. Since the surplus or deficit on the current account represents the change in a country’s net foreign assets, one approach to judging whether a current account deficit will become a problem is to assess the country’s external debt situation. This approach, which evaluates intertemporal solvency, investigates the country’s ability to repay its external debt. For a country to be intertemporally solvent, the present discounted value of future trade surpluses must equal the present value of its foreign debt. That is, intertemporal solvency requires that, in the long-run, the country’s external debts will be paid. In this case, a country can remain technically solvent even while running large account deficits. Another approach is to consider the implications of a path of external imbalance for the size of external indebtedness relative to GDP. The “sustainable” path of an economy’s external debt depends on two factors: (i) trade imbalances; and (ii) a debt dynamics term equal to the difference between the world interest rate and the domestic growth rate of the debtor country. If the world interest rate exceeds the domestic growth rate, then the only sustainable path along which the current account (excluding net factor payments) is constant is when the trade balance is in surplus. In order to keep the stock of external debt from rising without limit, the debtor country must generate a surplus on the net trade account in order to be able to service its debt commitments. The more rapid GDP growth, the larger is the sustainable level of external debt. But the real picture is much more complex. The nature of external borrowing is another important factor, given the critical relationship between the current account and the capital account. Inflows that are not permanent can potentially destabilise the domestic economy when they arrive and when they depart. Short-term capital inflows are likely to be more volatile than long-term inflows; and variable interest rates (compared with a fixed coupon) are all likely to increase the risk of an external crisis. The composition of a country’s external liabilities can affect its ability to manage its external position. Swings in international exchange rates and interest rates can be potentially devastating, especially if a large proportion of foreign debt is classed as short-term. The currency composition and maturity profile of a country’s external debt can contribute as much to vulnerability to external shocks as does the total volume of debt. A country’s ability to sustain a current account deficit is influenced not only by the size and nature of existing liabilities and the stock of foreign assets, but also by the extent of the debt-servicing burden. Degree of openness does matter, too. A country’s propensity to import and export will clearly influence the sustainability of its external position. The production of tradable goods provides an invaluable source of foreign exchange as these inflows can be used to service and reduce the size of the country’s external indebtedness. As any sustained structural deterioration of the trade balance would make a current account deficit even less sustainable, then a real exchange rate appreciation may be adverse or not. And strongly related to movements in the real exchange rate are a country’s terms of trade: a strengthening of the real exchange rate, or negative shocks in foreign demand, for example, would put downward pressure on the domestic country’s terms of trade. This might subsequently impact on the sustainability of a current account deficit. However, countries with larger export sectors are able to service a given external debt more readily. The more reliant an economy is on international trade, the less likely it is to default on its external debt commitments due to the threat of default reprisals. However, the more open an economy is, the more vulnerable it will be to external shocks such as terms of trade fluctuations or falls in foreign demand. 36 There is also a clear link between the sustainability of the current account and deviations of consumption, investment or government expenditure: high levels of domestic investment that are not met by domestic savings weaken the current account balance. In this situation, individuals acquire foreign debt in an attempt to finance investment opportunities while avoiding sharp drops in their consumption. Higher levels of investment may suggest higher future growth rates, which might enhance intertemporal solvency. However, higher savings and investment levels do not necessarily translate into higher output levels. Investment projects may be allocated inefficiently. Under these circumstances, high levels of public investment may not enhance external sustainability. The picture is very complex: positive productivity shocks can cause investment to rise, as it raises the expected path of future output directly. This tends to worsen the current account balance as domestic residents borrow abroad to finance the additional capital accumulation. The productivity increase may, however, also increase saving, which tends to offset the impact of increased investment on the current account balance, with uncertainty on the net effect on the current account balance. Thus, we have mentioned at least three categories of influential factors: (i) external debt and its composition imbalances, (ii) external trade imbalances, (iii) domestic factors – savings and investment - and public budget spending1 . Referring to the last point, the net effect of fiscal deficits on the current account and on the overall balance of payment remains an empirical issue, but econometric modelling is highly limited in its ability to predict or forecast a pending external crisis. The interrelations between saving-investment gap, deficit spending, trade imbalances and external debt must be more adequately investigated. 2. Fiscal deficit spending and development A research approach to current accounts has focused on analysing current account determinants and the response of current accounts to different types of shocks 2 . Chinn and Prasad (2000) use a large panel that includes both industrial and developing countries to examine the medium-term determinants of current accounts. They find that, in developing countries, current accounts are positively correlated with government budget balances, measures of financial deepening and initial stocks of net foreign assets3 . Most African external debt represents either direct loans to governments (and government owned firms) or publicly guaranteed loans to the private sector (i.e. loans that the government agrees to be responsible for in case the private non-payment). The interaction between external payment problems and government budget problems is a consequence of the government responsibility for paying the nation's debt. 1 A. North (1999), “Current account sustainability: Evidence from South Africa”, in LSE-Crefsa Quarterly Review, N. 1. 2 For related work examining the dynamics of the trade balance in response to different types of shocks, see E. Prasad and J. Gable (1998), "International Evidence on the Determinants of Trade Dynamics," IMF Staff Papers, Vol. 45, No. 3; and E. Prasad (1999), "International Trade and the Business Cycle," The Economic Journal, October. 3 M. Chinn and E. Prasad (2000), "Medium-Term Determinants of Current Accounts in Industrial and Developing Countries: An Empirical Exploration," IMF Working Paper 00/46. For related work on estimating equilibrium measures of current accounts and exchange rates, see P. Isard and H. Faruqee (eds.) (1998), "Exchange Rate Assessment: Extensions of the Macroeconomic Balance Approach," IMF Occasional Paper No. 167. 37 And this clearly reflects the specific context of SSA countries. Given the small share of government expenditure (both current and capital ones)4 that can be financed through government revenue (mainly taxes), due to the administrative, political and social constraints, SSA countries need sources of deficit financing. In theory, budget deficit (= government expenditure [=G] - government revenue [=T]) can be financed by borrowing from the private sector, from the monetary system or from abroad. The first option, that is based on domestic saving, is limited by the fact that African private capital markets are small, the risk of government default prevents bond buyers from being particularly interested in lending to the governments, the real rates of return are very low due to regulation over interest rates, exchange rate overvaluation makes foreign currency assets more convenient. Moreover, in SSA there is evidence of a particular form of “forced” private lending represented by deferral of government payments to suppliers and employees. The second option, being money printing and monetary expansion, tends to increase inflationary pressures, as the mechanism of seignorage makes it possibly to expand money supply without inflation up to a certain point, then the real value of money falls and it implies an “inflation tax” to be paid. Thus, the third option, external borrowing, was the most available solution to SSA countries, particularly when the availability of abundant and cheap international credit, in the late 1970s, was a very attractive opportunity. Differently from other regions, debt burden problem of Sub-Saharan African countries is essentially a public sector problem, governments have dramatic problems in collecting the needed revenue for paying debt service and debt service payments risk to be a drain on budget resources, which are subtracted to investment purposes. It seems to suggest a strong relationship between external debt and fiscal problems in SSA, and possibly a fiscal origin of the external debt itself, due to the need of financing an expansionary fiscal policy through international borrowing5 . These relations imply the need of focusing on the relationship between domestic debt, debt capacity and development, rather than exclusively on the Balance of Payments dimension of debt crisis. Put simply, we can argue that an increasing external public debt in relation to tax revenues can represent a disincentive and resource constraint to investment, thus leading to negative development implications. An emphasis on the domestic fiscal dimension of external debt, in the SubSaharan African countries, requires starting from some considerations on the relationship between domestic debt and development. Most governments have traditionally found it necessary to borrow to finance part or all of their expenditures for development programmes. The assumption is that as the economy develops, the increased government revenue derived from taxes on growing incomes as well as from government-owned productive investments will enable the government to pay off these debts, both external and internal. The excess of government expenditure over government revenue (the so-called primary, or noninterest, fiscal balance), which must be financed either by domestic or foreign borrowing or by printing money (that is issuing monetary liabilities) is the so-called «deficit spending». Keynesians have advocated that governments should run budget deficits during recessions in order to stimulate aggregate demand. The expansionistic effect of deficit spending on real economy is due to the multiplier effect of deficit used to finance productive investments. An 4 Particularly when SSA countries, not only financed the big push of development and war expenditure, but also decided, in order to maintain their political legitimacy and to solve potential conflicts, to tax too little and to spend too much, favouring some “sensible” interest groups. 5 P. Hjertholm (1997), An Inquiry Into the Fiscal Dimension of External Debt: The Case of Sub-Saharan Africa, PhD dissertation, Institute of Economics, University of Copenhagen. 38 increase in investment will raise national income by an amount equal to its monetary value, but in addition it will have a wider positive feedback effect by stimulating other parts of the economy, thus creating new jobs and additional demand for goods. From this perspective, the aim of developing countries is to attract much more debt rather than reducing it. Quite the opposite, monetarists and new classical macroeconomists argue that budget deficits simply stimulate inflation and crowd out private investment. This idea is based on the neutrality of budget deficit, as in the inter-temporal horizon any immediate expansion of real economy due to debt is not true in the long run. If current expenditures increase because of debt, in the long run it implies increasing taxation to insure the pay-back debt. Thus, the net present value of expenditure must always be equal to the present value of income. In empirical terms, it is not always the same the effect of deficit spending on the economy. It depends on the nature and composition of the given deficit spending. It can be due to a reduction of tax revenue or to an increase of public expenditure. If deficit spending is aimed at increasing public expenditure, its effect will depend on what kind of expenditure is involved. The positive effect, measured in terms of increasing aggregate demand, will be higher and direct, if deficit spending will be devoted to finance the purchase of goods and services or public investment. The effect will be lower and indirect if deficit is due to finance transfers (such as pensions and subsidies). The same effect occurs when deficit responds to a reduction of direct taxation, which determines an increase of disposable income. A reduction of indirect taxation determines a reduction in the price of goods and services, thus directly benefiting demand from households and firms. Obviously, there is no automatic passage from higher aggregate demand to higher production and to higher real national income. Any supply constraint can prevent aggregate demand from inducing positive effects on income. The way of financing deficit spending can itself determine the effects on development. As indicated, the supply of money in order to finance deficit can induce inflation pressure, which reduces the purchasing power of money, and produce a vicious circle of higher deficit spending 6 . Hyperinflation (defined as an inflation rate of 50 percent per month or higher) is characterised by large fiscal imbalances and lack of credibility in the policymakers’ ability to stabilise. Thus fiscal imbalances and pressures to monetise budget deficits may lead the economy to a stable inflation equilibrium or high inflation trap. Otherwise, the government can finance deficit through fixed-interest securities issued as short- or long-term bonds. Nowadays it has become common for governments of industrialised countries to borrow funds from their own citizens and institutions such as banks, insurance companies and investors. The interest payments are raised from taxes on the community and paid back to members of the same community. These are therefore transfer payments. The impact of these transfer payments on income distribution depends on which group pays the taxes as determined by the structure of the taxes paid and which groups and institutions hold the public debt. The securities issued by governments feel the effect of a globalised financial market: it can induce inflow of foreign investments and affect the foreign exchange. 6 The conventional fiscal deficit can be very sensitive to inflation, through the effect of inflation on nominal interest payments on the public debt. Moreover, inflation can have a sizeable effect on the conventional deficit when the domestic public debt (and, consequently, the interest rate bill) is high. The inflation-induced part of the nominal interest bill is a sort of amortisation payment that compensate holders of government debt for the erosion of the real value of the debt stock, maintaining real debt at the previous level. See V. Tanzi, M. I. Blejer and M. O. Teijero (1993), «Effects of Inflation on Measurement of Fiscal Deficits: Conventional versus Operational Measures», M. I. Blejer and A. Cheasty (eds.), How to Measure the Fiscal Deficit: Analytical and Methodological Issues, IMF, Washington D.C. 39 In general terms, rising debt is not necessarily a bad thing. Long-term economic growth will be faster if spare savings are channelled into productive investment rather than sitting idle under the mattress. It is natural that, as financial systems have become more efficient in channelling funds from those with surplus income to those who need to borrow, a developing country’s ratio of public sector debt to GDP increases. The problem is that, within this trend, a sudden, excessive surge in debt can have negative consequences. The sensible level of debt for an economy depends on interest rates, expected future growth in income, and the variability of growth. Families or industries with a more stable cash flow can afford to take on bigger debts than highly cyclical businesses. Only if the business cycle really has been made more stable and there is little risk of recession, firms and households can safely borrow more. The same counts for countries: a country with declining terms of trade can derive high problems from debt. Short-term debt must be used for investment being productive in the short-term. But even if the optimal level of debt is higher, a sudden surge in borrowing can leave households, forms and countries horribly vulnerable to shocks, such as higher interest rates, a fall in asset prices or an economic slowdown. One of the three is possible, indeed likely, in Sub-Saharan African countries. High debt levels are unlikely by themselves to trigger an economic collapse, but they do threaten to amplify any downturn, turning it into a deeper recession if debtors are forced suddenly to cut their spending in order to service or reduce their debts. There are three major contending interpretations on the linkages between public deficit and savings: 1. The conventional view asserts that a fall in government saving (due to a tax cut or deficit spending) tends to raise consumption and reduce saving by people who care only about the present, by shifting the burden to future, thus reducing national saving. 2. The Keynesian view asserts that higher temporary government dissaving will raise consumption and income, in the presence of under-utilised production capacity, proportionally to the inverse of the marginal propensity to save, as predicted by the multiplier effect. Higher income will raise private saving, which can be – but non necessarily is – large enough to offset the initial decline in government saving. 3. The Ricardian view asserts that, as the individuals are rational and far-sighted, they assume that a permanent rise in government spending today will be paid later, thus they will increase saving by an equivalent amount (the so-called Ricardian equivalence) and a rise in the budget deficit will have no effect on the national saving rate. The Ricardian idea that the timing of taxes does not influence household consumption behaviour implies that the choice between debt financing and tax financing of fiscal deficits is irrelevant. And a large ratio of public debt to output may have an adverse effect on private investment through various channels. • Resources used to service debt may crowd out government investment, which could be complementary to private investment, thus reducing national private investment; • A high debt-to-output ratio (with a high debt service ratio) may induce domestic agents to transfer funds abroad because of the uncertainty and fear of future tax liabilities to service the debt, thus raising the domestic cost of capital goods and lowering national private investment; 40 • A high debt-to-output ratio can discourage foreign direct investment because of the fear that government may impose restrictions on external payment obligations, thus reducing the national domestic investment, which are complementary to FDI; • The risk of nominal depreciation affects firms that hold stock of foreign-currency liabilities, because of the reduction of their net worth induced by the raising debt burden due to depreciation, thus domestic banks tend to tighten credit restrictions and depress investment. Given these premises, very high level of public debt are considered dangerous, because of three different traps: 1. The «crowding-out effect», determined by the fact that high public borrowing pushes up interest rates and crowds out alternative productive private sector investment. 2. The «inflationary pressure», due to the government attitude to print money in order to reduce the real value of public debt. This phenomenon was deeply analysed in developed countries because of the risk in the 1960s and 1970s of inflationary spiral i.e. linkages between cost-push factors and demand-pull ones: higher prices induce higher wages translated in higher costs which determine much higher prices and so on- 7 . 3. The «debt trap», due to rising interest payments which increase the government’s budget deficit. Thus, even though the government is reducing public spending, total debt continues to grow in a spiralling of its own. If interest rates exceed nominal GDP growth (real growth plus inflation) and there is a primary budget deficit (government spending other than interest payments minus tax revenues), then the country is enveloped in a indefinitely growing spiral of debt-to-GDP ratio, and balancing the budget would not be enough to stop it. Particular consequences derive from the decision to finance deficit spending through international savings, which is external debt: the commitments of one government to other governments or banks or to other financial institutions abroad. In this context, which seems to be the most adequate for SSA countries, the analysis of fiscal deficit can be translated into the analysis of external debt. However, we have to remind that “...the relationship between higher public savings and growth is not clear cut. In the face of such ambiguity it can only be concluded that examining how aid inflows affect government behaviour will not necessarily provide any concrete answers to the nature of the aid-growth relationship” 8 . 7 Barro focused on the positive relationship between expected inflation and budget deficits. Higher inflation expectations imply higher real interest rates and therefore higher debt service costs. See R. J. Barro (1979), On the determination of the public debt, J. Polit. Econ. 87. This phenomenon is linked to the Ricardian hypothesis on debt neutrality (in order to finance a war, a government can choose new debt or new taxes, but the difference is only inter-temporal, because current debt will imply new taxes in the future to pay back debt) and rational expectations (i.e. recognising the inter-temporal constraint of budget). In fact, the Ricardian hypothesis has been revived recently by Barro (1988). According to this viewpoint, which has been used by monetarists and new classical macroeconomists, changes in budget deficits cause offsetting changes in private savings through anticipations of changes in future taxation. Therefore, they have no effect on national savings and, consequently, on the current external account. See R. Barro (1988), «The Ricardian Approach to Budget Deficits», NBER, Working Paper, no. 2685. 8 H. White (1992), “The Macroeconomic Impact of Development Aid”, The Journal of Development Studies, vol. 28, N. 2, January, p. 190. On “savings debate”, Keith Griffin stresses the risk that foreign capital is not additional to domestic saving. Whereas, Eprime Eshag says that the basic assumption which is made by Griffin is that the elasticity of supply of labour and goods and services in underdeveloped countries is zero... Once the assumption of zero elasticity of supply is removed, as indeed it has to be in any realistic discussion of the impact of the inflow of foreign capital on savings, the whole picture is radically changed... There is in fact every reason 41 3. From domestic debt to foreign debt Some developed countries, such as Italy 9 , demonstrate how different is the concept of publicsector debt from external debt. The public sector debt is perceived by citizens as an ambiguous fact: when interest rates decrease, it has positive effects on the national accounts, but negative effects on the private investors who hold Treasury bonds. At the same time, citizens are borrowers and lenders in the case of public-sector debt. Quite differently, debt owed to foreign lenders is external debt. For all countries, total payments in international transactions must equal total receipts. When all items, both visible and invisible, have been taken into account, each country must cover its current payments with its current receipts. A surplus in the Balance of payments (BoP) may be inserted as a credit and added to the country’s reserves of foreign currencies accumulated over time. A deficit must be covered: (a) by payments out of accumulated reserves of foreign currency; (b) by borrowing or obtaining investments from another country; (c) by acquiring import from suppliers for credit; (d) by using exports receipts; (e) by obtaining assistance from the IMF in the form of credit; (f) by receiving ODA flows. In general terms, a country could not go on paying for deficits over a prolonged period out of its accumulated reserves of gold and foreign currencies. The balance of payments of a nation is a double-entry accounting statement of its transactions with the rest of the world during a specified time period. Net (of amortisation) inflows are recorded as a plus (that is receipts, credits), outflows are a minus (that is debits or payments). The 5th edition of the IMF BoP manual provides for the separate reporting of financial capital inflows (liabilities) and outflows (assets). Both the capital inflows and outflows of the BoP are reported net of repayments. One advantage of these data is that they distinguish three types of capital flows: foreign direct investment (FDI), portfolio investment and other financial flows - comprised primarily of bank loans 10 . This enables to study, for example, whether FDI is «different» from loan inflows. A second advantage of this source is that the consistent accounting framework enables us to relate capital inflows to other components of the balance of payments: in particular, the current account, reserve accumulation, capital outflows and reserve accumulation. to expect domestic savings will increase on the assumption that some domestic resources, which would otherwise have remained unemployed, are used in conjunction with foreign resources”. See K. Griffin (1970), “Foreign Capital, Domestic Savings and Economic Development”, Bulletin of the Oxford University Institute of Economics and Statistics, Vol. 32, N. 2, May, pp.99-112; K. Griffin and J. Enos (1970), “Foreign Assistance: Objectives and Consequences”, Economic Development and Cultural Change, Vol. 18, pp.313-327; E. Eshag (1971), “Comment”, Bulletin of the Oxford University Institute of Economics and Statistics, Vol. 33, N. 2, May. 9 In 2000, Italy had the largest public-sector debt in Europe, worth 110 percent of GDP. Italy is bound by European Union rules to balance its budget in the medium term: Italy pledged in 2000 to cut 2001 deficit to 0.8 percent of GDP, and to stop borrowing more money than it earns by 2003, whereas its debt should be reduced to 100 percent by 2003. 10 Usually, the three types of capital inflows (FDI, portfolio, and loans) are not significantly correlated with one another over time or across countries. That is, there is little tendency for countries with large amounts of portfolio capital or loans to receive correspondingly large amounts of FDI. 42 Tab. 1 - Summary of international transactions Line 1 2 1+2 3 4 5 6 1+ …+6 7 8 9 10 11 12 7+…+12 13 14 Item Exports of goods and services Imports of goods and services Merchandise balance of trade Government grants Investment income, net (bonds, stocks,...) Debt-service payments Remittances, pensions, other transfers Current account balance FDI and portfolio assets, net Short term loans, net Long term loans, net Official assets, net Official reserve assets, net Allocations of special drawing rights Capital and financial account balance Changes in reserves, net Statistical discrepancy (errors-omissions) (effects on the Balance of Payments) (+) (-) (-) (-) (-) (+/-) (-/+) (+/-) (+/-) (-/+) (-/+) (+/-) Foreign aid (bilateral and multilateral) is largely classified as a transfer in the BoP and excluded from capital inflows, but some forms of concessionary finance are included. Exceptional finance transactions and IMF credits are included with total reserve accumulation (RES). Such transactions could be classified as capital inflows; but because they are frequently associated with crises, debt forgiveness and debt restructuring, they are negatively correlated with other inflows. Errors and omissions are frequently associated with capital flight. The following identity provides a simplified overview of the different types of transactions in the BoP accounts: 0 = (CA + KA) + FINI + FINO + ERR + RES (2.1) where: CA = Current account balance, KA = Capital account balance, FINI = Financial inflows, FINO = Financial outflows, ERR = Errors and omissions, and RES = Reserves and related items. As usual, sources of foreign exchange, such as financial inflows and exports, are denoted as positive (credits) while uses of foreign exchange, such as financial outflows and imports, are negative (debits). Thus, we have a current account, a financial account and a reserve account. What used to be referred to as the capital account now corresponds most closely to the sum of the financial account and the reserve account. Within this simplified framework, inflows of financial capital can either be set aside in reserves, used to finance current account deficits, or be offset by financial capital outflows. If a country runs a trade deficit by having an excess of imports over exports, a number of offsets may take place: increasing liabilities to foreigners, decreasing claims on foreigners, 43 liquidating the pressures that would result if payments in foreign currencies had to be made immediately. Similarly, a country that is increasing its investments abroad will improve its investment position, but it will be creating minus entries in its current balance of payments statement. But the long-run outlook for a country that is making substantial foreign investment should be favourable, as a result of the future income that may be generated from those investments. From a very broad point of view, a common characteristics of Ldcs is a structural current account deficit, where debt service payments represent a major and growing component – together with imports -, which is not adequately contrasted by the value of exports, flows of the net investment income received from abroad, net private and public remittances and transfers. In the case of the capital account, most Ldcs remain fundamentally dependent on foreign loans by private and public creditors (also in the form of foreign aid), net of amortisation, to adequately contrast the relevant flows of capital flight (to Western bank accounts, real estate ventures, stock and bond purchases) and the resident capital outflow. The international reserve account, or cash account, is the balancing item, along with the errors and omissions item, which reconciles statistical discrepancies, that is lowered in most of Ldcs, because total disbursements on the current and capital accounts exceed total receipts. In fact, Ldcs should balance any purchase of goods and services abroad and re-payments to foreign countries – including debt-service payments (in the current account) and debt amortisation (in the capital account), which represents the gradual liquidation (i.e. its reduction) of the principal of former loans. But during the 1980s, the outflow to repay accumulated debts exceeded the inflow of both public and banking loans, determining a negative net transfer (or a perverse flow of financial flow transfer from poor to rich countries). Moreover, debt-service payments must be made with foreign exchange; thus, they must be covered through export earnings, curtailed imports, or further external debt. From the BoP perspective, international reserves (which consist of gold, a few major foreign currencies and SDR at the IMF) can be drawn on to pay bills and debts. But the poor countries – as the SSA countries are - are likely to have a very limited stock of reserves, thus the overall BoP deficit can inhibit the country’s ability to continue importing needed capital and consumer goods. In these cases, there are some policy options: (1) To seek to improve the balance of current account by promoting export expansion (by supporting potentially competitive sectors) or limiting imports (through import substitution policies, selective tariffs, physical quotas) or both of them (through a currency devaluation or a structural adjustment and stabilisation policies to reduce domestic demand – in order to lower imports – and to avoid inflationary pressures, which contributes to an overvalued exchange rate that discourages exports). (2) To seek to improve the balance of capital account, encouraging FDI and more international borrowing, which implies the necessity of future repayments of principal and interest. (3) To seek to improve the stock of international reserves 11 . A very simple and useful way of describing external debt, clearly referred to the BoP, is through the usage of the concept of the basic transfer 12 . It is the net foreign-exchange inflow 11 Linked to this purpose, the Ldcs as a group are now exerting pressure at international level to agree to the creation of supplementary SDRs, as the present forum for distributing current SDRs gives 75 percent of the total to the 25 industrialised nations. 12 F. Stewart (1985), «The international debt situation and North-South Relations», World Development, N. 13, February. 44 or outflow related to its external debt. It is measured as the difference between the net capital inflow (i.e. the difference between the gross inflow and the amortisation on past debt) and interest payments on the existing accumulated debt. Thus, the basic transfer concept represents the amount of foreign exchange that a country is receiving or losing each year because of international borrowing. FN = dD Where: FN is the net capital inflow D is the total accumulated foreign debt d is the percentage rate of increase of debt BT = dD – rD = (d – r)D where: BT is the basic transfer r is the average rate of interest (and rD is total annual interest payments). If r < d, then BT will be positive and the country will be receiving foreign exchange; if r > d, then BT will be negative and the country will be losing foreign exchange. From these relations, it derives that in the early stages of debt accumulation (when D is relatively small), d is likely to be high and r is low (at least lower than d); exactly the opposite to the situation when the accumulated debt becomes large and determines a selfreinforcing phenomenon, into a downward spiral of negative BT, dwindling foreign reserves, expectations of currency devaluation and substantial flight of capital, BoP deficit. 4. Foreign debt and development In the 1990s, by analysing foreign debt and economic performance in an aggregate context, economic literature has found a fairly robust statistical relationship between high foreign debt burdens and poor economic performance 13 . What has not been adequately studied is that in poor countries fragile budgetary systems may entail difficult fiscal policy tradeoffs. There is a link between the effects of public debt on economic development and the effects of foreign debt on economic development. In fact, possible channels that can transmit negative effects of the fiscal burden of foreign debt on economic performance are: (i) Cash-flow effects stemming from public expenditure crowding-out. Public expenditures could have complementary effects on private investment, but as these expenditures are crowded out by public debt service, poor countries miss out on these complementarity effects, and thereby experience lower levels of private investment and growth than it would be possible. (ii) Cash-flow effects stemming from import compression. In fact, as the ability of poor economies to substitute between imported and domestic capital goods is limited, a reduction in capital goods imports will lead to a decline in investment activity and growth. Moreover, import compression occurs both at the balance of payments level 13 P. Hjertholm, J. Laursen, H. White (2000), Macroeconomic Issues in Foreign Aid, DERG-Institute of Economics, University of Copenhagen. 45 and at the budgetary level (i.e. the effect of public debt service on the import-content of government expenditure). Reductions in the import capacity of the government, as a result of debt service, can reduce government investment activity. (iii) Disincentives associated with a large debt overhang. Investors hope that their expected return will be realised. An «overhang» of public debt can affect this «incentive structure», because the tax disincentives associated with high debt burdens (part of any future increase in output will go to repay foreign creditors: this will be perceived domestically as a «tax» on investment returns, which in turn will discourage investors) 14 and broad measures of macroeconomic instability (there is a lack of the element of certainty guaranteed by macroeconomic stability, which is enabling for investment planning). (iv) As a result of higher debt burdens, uncertainty about international creditworthiness may produce restrictions on access to international capital markets. Moreover, debt can increase the overall fiscal deficit directly by increasing debt service payments 15 . It can lead to exchange rate depreciation, for balance of payments reasons, whereby the fiscal deficit widens because the home currency value of public debt service may increase relatively to public revenue. An increase in the part of the fiscal deficit that is monetized (through higher government credits), can lead to monetary expansion and inflation (exactly the same as in the budget deficit case described above). Exceptional financing, such as payment arrears and rescheduling of debt payments maintains uncertainty about the future debt servicing profile of the public sector, while at the same time disrupting ordinary trade flows. Debt-induced fluctuations in macro indicators as inflation and exchange rates may signal fiscal distress, thus reducing the incentive to invest16 . In more general terms, external debt may have adverse effects for the savings, foreign exchange and fiscal gaps in the longer term, that is negative effects for economic development. These aspects should be part of any serious analysis of debt sustainability, which is described in the next chapter. 5. Foreign debt and two-gap models First of all, as a premise to this paragraph, it must be reminded that political motivations have been by far the more important for official aid and loans, especially for the major industrialised countries 17 . Creditors and donors give resources because it is in their political, strategic, or economic self-interest to do so. Some assistance may be motivated by 14 As such, debt stock reduction may benefit debtors and creditors, and may thus be described as an efficient «market-based» debt reduction. However, it must be added that the general state of the empirical evidence on the link between debt overhang theory and poor investment performance of poor indebted countries appears inconclusive. 15 In budget account this phenomenon is well known: at the beginning there is negative primary deficit (for example, 10 US$ bn), that is a difference between expenditure and revenue, net of interests. In the following year, if there is an interest rate equal to 10 percent, no new primary deficit, no budget surplus (revenues bigger than expenditures), no monetary expansion to cover past deficit, then there will be new debt to cover nonprimary deficit (equal to interest rates: 1 US$ bn), and total debt stock will amount to 11 US$ bn. This exponential spiral will lead to an increase of debt stock equal to 10 percent every year (the following year total debt stock will amount to 12,1 US$ bn), that is the interest rate. The way to face this problem, taking in account the inter-temporal constraint of budget, is to reverse this trend, inducing a positive primary surplus. 16 P. Hjertholm, J. Laursen, H. White (2000), op. cit. 17 It is very impressive the figures of US ODA at the beginning of the 1990s, when US assisted Israel with $176 per poor person, compared to $1,7 to Bangladesh, $2,7 to Tanzania, $3.6 to Mozambique. 46 humanitarian desires, but there is no historical evidence to suggest that over longer periods of time, rich nations assist others without expecting some corresponding benefits (not necessarily financial: it may be military) in return. Nevertheless, the economic rationale has been given lip service as the overriding motivation. And the principal economic argument 18 is that external finance can play a critical role in supplementing domestic resources in order to relieve savings or foreign-exchange bottlenecks. This is the so-called two-gap analysis, which requires specific attention. Theoretical literature has analysed why most of the countries have had to supplement domestic savings with foreign finance. This phenomenon shows the link between nationalincome balance (the problems of internal imbalances in the macro-economy) and balance of payments (external imbalances) analyses. In fact, from an internal imbalance derives an external disequilibrium. In national income analysis, the uses of national income must equal the disposal of national income: (C+I+G+X-M) = (C+S+T) → (I+G) – (S+T) = (M-X) (2.2) where: C = consumption, I = gross investment, X = exports, M = imports, G = government expenditure S = gross savings, T = taxes. From the identity 3.1, it derives that: (I+G) >(S+T) → M>X (2.3) The relationship 2.3 means that if a government is spending more than it is earning, it will imply a resource gap within the economy, and then imports will be greater than exports. The model has two important features: (A) The Harrod-Domar idea that investment requirements to achieve a given growth rate are proportional to the growth rate by a constant known as the Incremental Capital Output Ratio (ICOR)19 . 18 Another important economic rationale is derived from the first one: external finance can facilitate and accelerate the increase of domestic savings as a result of the higher growth rates that it induces. Another rationale assumed to be important (but scarcely considered in the reality by the creditor countries) is the socalled absorptive capacity of the recipient country, that is its ability to use funds wisely and productively. 19 In the neo-classical growth model of Solow with labour-augmenting technical change, the ratio of capital to output will be constant in steady state. Both output per worker and capital per worker will increase at the rate of technical progress. The level of output, not the growth rate of output, will be a function of the investment rate in steady state. One can derive a constant ICOR in steady state - it will be given by the ratio of the investment rate to the sum of population growth and the rate of labour-augmenting technical progress. A high ICOR here could reflect a high investment rate and a low population growth rate - both thought to be desirable - not necessarily low efficiency of investment. The constant ICOR in the Solow steady state does not signify a causal, proportional relationship between investment and growth. An exogenous increase in investment will raise growth temporarily during the transition from one steady state to another. In the endogenous growth models of recent years, the ICOR also does not fare well. A decrease in the ICOR also does not necessarily imply improved quality of investment. It may mean simply a lesser quantity of capital invested relative to other factors. The endogenous growth models stress a multitude of inputs besides physical capital, such as human capital, intermediate «new goods», and organisational capital. The ICOR is constant in the steady state of the endogenous growth model, as in the Solow model. However, the constant ICOR reflects the endogenous steady state response of both growth and investment to the model parameters and to policies like the tax rate. It does not represent a linear causal relationship between physical capital and output, because any increase in physical capital with human capital held constant will run into diminishing returns. See W. Easterly (1999), «The Ghost of Financing Gap. Testing the Growth Model Used in the International Financial Institutions», Journal of Development Economics, No. 60, December. 47 (B) External finance requirements are given by the «Financing Gap» between the investment requirements and the financing available from the sum of private financing and domestic saving. (A) and (B) imply the following testable propositions: (1) aid/loans will go into investment one for one 20 , and (2) there will be a fixed linear relationship between growth and investment in the short run. The constant of proportionality is one over the ICOR. As in the Harrod-Domar model, output depends on the investment rate and on the productivity of that investment. Investment is financed by savings, and in an open economy total savings equal the sum of domestic and foreign savings. This model simply repeats what the Harrod-Domar model stressed. Every economy must save (total national savings = S) a certain proportion of its national income, in order to replace capital goods (buildings, equipment, materials), and to grow. New investment (= I) represent net additions to the capital stock (= K), and it will bring corresponding increases in flow of national output, GNP (= Y). The relationship between K and Y is the capital-output ratio (=K/Y=k). S is some proportion (= s) of Y; I is defined as the change in the capital stock (=∆K=k∆Y). As we assume S=I, thus S=sY=∆K=k∆Y=I and, if sY=k∆Y, then ∆Y/Y (= the rate of change or rate of growth of GNP, i.e., the percentage change in GNP) = s/k. Thus, the rate of growth of GNP is determined jointly by the national savings ratio and the national capitaloutput ratio: the more an economy is able to save – and invest (and not to consume) – out of a given GNP, the greater the growth of that GNP; the higher k is, the lower the rate of GNP growth will be. Moreover, 1/k, i.e. the inverse of k, is simply the output-capital or outputinvestment ratio, that is productivity. Multiplying the rate of new investment (s=I/Y) by its productivity (= 1/k), will give the GNP growth rate. This is what Rostow and others defined the take-off stage: to increase s in order to have a self-sustaining development, which is simply a matter of increasing national savings and investment. From (2.3), in order to fill the domestic resource gap (insufficient domestic resources to permit domestic expenditures), a foreign exchange gap is needed and created (imports exceeding exports). It requires to fill the foreign exchange gap. This can be done through the usage of foreign exchange reserves or through external financing (foreign aid, gross-border sovereign lending by commercial banks, private foreign investment – foreign direct investment and portfolio investment – loans from the World Bank, access to the drawing rights in the IMF). Trade gap, which is based on the further assumption that not all investment goods can be produced domestically, is in addition to the savings gap. If the savings gap is directly linked to the Harrod-Domar model, the foreign exchange gap derives from the assumption that production is constrained by inputs, which are provided by imports. In this case, the foreign exchange gap can be described as a technological gap and the filling of this gap is useful to buy these needed inputs, rather than being complementary to national saving. Hence a certain level of imports is required to attain desired investment (i.e. once again the investment 20 In an endogenous growth model, a lump sum transfer like aid/loans would have no effect on the rate of investment. There is also a moral hazard problem with giving aid/loans on the basis of a «financing gap». Recipient countries will have an incentive to maintain or increase the «financing gap» by low saving (i.e. high consumption) so as to get more external resources. Even if the donor/lender puts savings conditions on the aid/loans, the conditions may not be credible if the recipient perceives the donor as soft-hearted (donor conditionality has indeed turned out to be ineffective in changing recipient behaviour). All of these theoretical perspectives are inconsistent with the one-to-one aid/loan to investment link postulated in the Financing Gap Model. See W. Easterly (1999), ibid. 48 required to achieve the target growth rate). The import bill is financed either from export earnings or foreign capital inflows (e.g. external debt and aid). If exports are not sufficient to cover the whole bill the availability of foreign exchange (forex) to purchase imported capital goods (rather than the supply of domestic savings) may become the binding constraint on growth. Hence, there is binding trade gap, also called the foreign exchange (forex) gap 21 . Through foreign finance it is possible to fill the foreign-exchange gap, which allows the real capital transfer in the form of imports greater than exports. The more domestic investment is high, the more the foreign finance must be large. The amount of needed foreign finance depends on the target level of GNP: at the given level of GNP, foreign finance must cover the balance of trade deficit (=M-X), plus servicing of external debt, outflow of interest, dividends and profits on private foreign investment, capital flight, and the desired build-up of foreign exchange reserves. The net capital inflow is equivalent to the net resource transfer, which finances the resource gap and the foreign exchange gap. In filling the resource gap, foreign capital provides an equivalent increment to the capital stock and, in filling the foreign-exchange gap, makes it possible to complement domestic capital in production with imports (current account deficit). Ldcs face either a shortage of domestic savings to match investment opportunities (= savingsgap, or gap of domestic real resources) or a shortage of foreign exchange to finance needed imports of capital and intermediate goods (foreign-exchange gap). In order to facilitate econometric analysis, but at cost of being unrealistic, these two gaps are assumed to be unequal in magnitude and independent, that is there is no substitutability between savings and foreign exchange. In this case, one of the two gaps is dominant at a given time. If the savings gap dominates, then the country is not using all of its foreign-exchange earnings, because of lack of sufficient productive resources (labour included) to carry out additional investment through imports of additional capital goods from abroad. It means that if the import occurs, then it will lead to inflation and to redirect domestic resources. Savingsgap countries therefore do not need external debt. Most Ldcs are assumed to fall into the second and worst case, when the foreign-exchange gap is dominating. These countries have excess of productive resources (particularly labour) and all used foreign-exchange is used for imports. They have complementary domestic resources, which can not be used for new investment projects because of lack of external finance to import new capital goods. Thus, they do not reach the potential real rate of economic growth. Obviously, such gaps forecasts are quite mechanistic and even if exports (=Ex) and capital inflow (=F=M-Ex) seem substitutable, they have different indirect effects, especially considering that F represents loans that need to be repaid. Given the Harrod-Domar model assumptions, savings and trade limited performance can prevent countries from dynamic growth process. If in a given country exports earnings make it possible to get all the foreign exchange needed to buy imports and foreign capital, then GNP growth could be constrained by the domestic savings gap 22 . Whereas, if the country is 21 The two gaps are combined in the two-gap model. Growth will be constrained by the larger of the two ex ante gaps. If aid is insufficient to fill the larger of these gaps the desired growth rate cannot be attained. That is, the gaps are not additive: aid simultaneously fills both gaps (by paying for imported capital equipment a single aid dollar relaxes both the savings and the forex constraint). 22 The idea of a gap only makes sense given an exogenously determined target growth rate. A distinction is thus made between the ex ante savings gap (the difference between desired investment and domestic savings) and the ex post savings gap (the difference between actual investment and domestic savings). 49 not able to export enough, then the external gap will be the main constraint, and the economy can not reach an higher productive capacity, because it can not augment the capital stock, and it induces reduction to the rate of growth. Within the context of the Harrod-Domar model, this corresponds to the situation in which the warranted growth rate (G w = s/v = saving propensity/capital-output ratio), which satisfies the Keynesian condition for macro equilibrium, is lower than the natural growth rate (n = λ+τ = population growth + progress of technological knowledge), which allows the continuous maintenance of full employment. In this situation there is a growth path which is warranted (actual investment and saving = desired investment and saving), providing the equilibrium between demand and supply in the long run, without fully employment of labour force. The problem of knife-edge balance between warranted and natural rates of growth that was introduced by Harrod and Domar, has attracted a number of comments 23 . Foreign finance creates the situation, which satisfies the condition for steady-state growth that is full employment, contrasting the resource and/or foreign exchange gaps. Chenery and Bruno introduced this two-gap model, based on internal and external balance, in its theoretical aspects 24 in 1962, and it was clearly described by McKinnon25 in 1964, within a neo-classical framework, based on the importance of financial support to growth. Then, in 1966, Chenery and Strout stressed, through an explicit model, the importance of expansion of productive structures, in order to increase domestic investment and production, that is the pre-condition to repay external finance and close the foreign exchange gap, through the rise of exports 26 . Still today economists at the World Bank use some variant of the Financing Gap model to make growth and financing gap projections. According to the Spring 1995 reference guide to the standard World Bank model, «the ICOR and prior investment determine GDP». Country economists make assumptions about ICORs and national saving and calculate the Financing Gap corresponding to a target growth rate. World Bank staff presents the result of this calculation at meetings where aid donors agree upon aid amounts for a specific country. The World Bank is not alone; virtually all international institutions addressing the needs of poor countries stress the short-run necessity of both investment and aid for growth. The International Monetary Fund (IMF) today trains developing country officials to project investment requirements as the «target growth rate times the ICOR» 27 . As the World Bank (1991) noted, «This so-called two gap model of the domestic saving and foreign exchange constraint to growth guided external aid and lending agencies in judging the extra resources that developing countries would need to finance imports and investment». Economists computerised Chenery’s version of the Financing Gap model at the World Bank in 1971, where Chenery was now the chief economic adviser to Bank President Robert 23 Robinson and Kaldor treated saving propensity as a variable depending on the distribution of income, whereas Solow and Swan considered the capital-output ratio as a variable. See A. Sen (ed.) (1970), Growth Economics. Selected Readings, Penguin Books, Harmondsworth. 24 H. B. Chenery and M. Bruno (1962), «Development Alternatives in an open economy: the case of Israel», Economic Journal, No. 72. 25 R. I. McKinnon (1964), «Foreign exchange constraints in economic development and efficient aid allocation», Economic Journal. 26 Chenery and Strout (1966) called their model the Two Gap model. The investment-savings gap was one of the Two Gaps. The other was the trade gap which ex post is equal to the investment gap, but ex ante might be a constraint in an import-constrained economy with fixed prices. IFI staff still occasionally use the trade gap instead of the investment-saving gap, but this is of little consequence since the two are equal. 27 See W. Easterly (1999), op. cit. 50 McNamara. Although Chenery and Strout were vague over what time horizon to use their model, the computerised version of the model had a lag of one year from investment to growth28 . 6. The IMF and the World Bank approaches Given the links between external disequilibrium, fiscal disequilibrium and inflation pressure, it is useful to describe and examine the basic macroeconomic models utilised by the IMF for «financial programming» of stabilisation plans. This approach, called the monetary approach to the balance of payments29 , ensures consistency between the monetary impact of policy changes and the desired balance of payments outcome. It is aimed at closing external disequilibrium and ensuring inflation control, through budget control. The IMF mandate is to finance temporary balance of payments disequilibria; when deficits are not temporary they must be rendered so by corrective policy measures. The basic assumption is that inflation is due to an excess of money supply, caused by deficit spending. The core idea is that austerity reduces demand and, given the production level, prices decrease. Real GDP is taken to be exogenously determined and quite stable, the velocity of money is assumed to be constant, whereas the change in the domestic price level is the endogenous variable. Economy is divided into four sectors: public, private, banking and foreign sector. Public sector neither invests nor produces and it uses receipts for consumption and foreign assets accumulation; private sector has no domestic financial asset, can borrow from commercial banks and accumulates asset-holdings abroad through capital flight. The balance of payments is expressed as the difference between the private sector’s flow demand for imports and the flow of domestic credit. Increases in domestic credit will be offset by decreases in foreign reserves; a reduction in the rate of expansion of domestic credit will improve the balance of payments performance and reduce domestic inflation, with any given rate of inflation. But through changes in the rate of expansion of domestic credit, it is not possible to choose independently the targets for the balance of payments and domestic inflation: in Tinbergen terms, two targets can not be achieved with a single instrument. The use of the exchange rates (the domestic currency price of a unit of foreign currency) as a policy tool provides the way out of this problem. With two instruments at disposal, a government can attain the targeted values for both the balance of payments and the rate of inflation. Any increase in domestic real GDP and in domestic price raise spending on imports, while devaluation will reduce imports if the volume of imports is responsive to relative prices30 . The volume of exports is assumed to depend on the relative price of foreign goods in terms of domestic goods. 28 See W. Easterly (1999), op. cit. 29 See J. J. Polak (1957), Monetary analysis of income formation and payments problems, IMF Staff Papers, Vol. 6, Washington D.C.; W. E. Robichek (1967) "Financial Programming Exercises of the International Monetary Fund in Latin America", Address to a seminar of Brazilian professors of economics, Rio de Janeiro; IMF (1977a), The Monetary Approach to the Balance of Payments. Washington D.C.; IMF (1977b) "Theoretical Aspects of the Design of Fund-Supported Adjustment Programs". IMF Occasional Papers, September; H. G. Johnson (1977) "The Monetary Approach to the Balance of Payments - A nontechnical guide", Journal of International Economics, vol.7; F. Hahn (1977) "The Monetary approach to the Balance of Payments", Journal of International Economics, vol.7. 30 But a common problem for developing countries is that a fall in the value of national currency relative to other currencies makes terms of trade worse, reducing the real value of local resources for consumption, as imports become more expensive, and increasing the cost of debt service, which is expressed in foreign currency. 51 Money demand rises in proportion to the inflation rate and the growth rate of capital stock (or capacity). Higher money demand due to capacity growth is «seignorage»: it is the money that the banking system gets to create without inflationary complications. The IMF approach is focused on inflation and foreign reserves, which is the instruments to solve the problem of balance of payments disequilibria. The foreign exchange problem is assumed to be the core issue, whereas the fiscal problem is not a priority within this model. Foreign exchange and domestic total credit are the basic instruments. A ceiling level of budget deficit is fixed. A limited increase in total domestic credit (instrument 1) should affect domestic equilibrium, while devaluation of foreign exchange (instrument 2) should affect the external equilibrium. But, in this model, the domestic effect induced by the foreign exchange policy is not clear: income rate growth changes derived by this policy can be inversely correlated to price changes. Polak’s basis for the IMF approach to stabilisation policies The basic assumptions are: Y = vMs M = mY Where: Y = GNP Ms = money supply M = imports v = marginal propensity to keep money m = marginal propensity to import These assumptions, based on the quantity theory of money (MV=PT, where V= velocity of circulation; P= price level; T= the number of transactions in the period), imply a constant velocity of circulation of the quantity of money, which is the central pivot of monetarism: for a given number of transactions, the relationship between the quantity of Money (=Ms ) and the price level (=p) is direct; thus, any increase in the money supply will lead to an increase in the price level, i.e. to inflation31 . The changes in the quantity of money and balance of payments are: ∆Ms = B + ∆D B=X–M+K Where: B = changes in the balance of payments D = total domestic credit X = exports K = total capital inflow, net In equilibrium: Y=Y -1 M s =M s-1 B=B-1 ∆Y=∆Y-1 Thus, any increase in domestic credit induces an higher level of imports (in order to keep the monetary and income equilibrium), which reduces foreign reserves, without affecting real production, employment and domestic rate of inflation. 31 «The principal tenet of monetarism is that inflation is at all times and averywhere a monetary phenomenon. Its principal corollary is that only a slow and steady rate of increase in the money supply – one in line with the real growth of the economy – can ensure price stability». G. Macesich (1983), Monetarism, Theory and Policy, Praeger Special Studies, New York, p. 3. 52 For any level of K and X, the increase of domestic credit is dependent on the desired target of balance of payments. Money demand is always stable, linked to income changes: ∆M s = f (∆Y) Any change of credit or quantity of money should be based on inflation and balance of payments target. Assuming a balance of trade deficit in a given Ldc, the IMF approach implies a stabilisation plan to correct the external unbalance. Given the level of real production, domestic inflation and employment, money demand is estimated, on the basis of the reduction of credit, needed to correct the balance of payments. A reduction of fiscal deficit is the way, which can improve trade account, through a reduction of domestic absorption; a reduction of the quantity of money is another way, which operates through a reduction of inflation and imports. The main limitations of this model are: (a) it is based on monetary variables, without taking in account any real variable (production, employment); (b) it is based on the flexibility of nominal wages and on the link between domestic and international inflation and interest rates (at least, a linkage referred to tradable goods prices); (c) it does not consider the possibility of a Keynesian context, where a reduction of domestic credit is not the correct way to reduce the balance of payments disequilibria, because it can produce, in the short-term, a change in production, employment and inflation. Also devaluation can produce positive or negative effects in the short-term, according to the change induced on nominal wages, and it determines domestic inflation pressure, in the case of full employment. This IMF approach can be combined to the World Bank model, which is a variant of the twogap growth model and is defined within the Revised Minimum Standard Model (RMSM). If the IMF approach is based on financial variables, the World Bank focus is on real variables. If the IMF is concerned with temporary balance of payments disequilibria, the World Bank is charged with the financing of growth and development over the medium-term. The basic approach of the World Bank stresses the relationship between savings, foreign capital inflows, investment and growth. The RMSM is an accounting framework linking the national accounts and the balance of payments, focussed on the foreign financing gap and on the need of foreign borrowing. Inflation, in this case, is not considered an endogenous variable within the model. The incremental capital-output ratio, private sector savings rate and exports are assumed as exogenous variables. The growth of real GDP is based on the available level of investment (or the required level, consistent with a desired rate of growth), thus the basic relationship is between changes in output and investment. An increase in output increases both domestic savings according to the given saving rate, and the inflow of foreign saving, according to the marginal propensity to import. In order to raise domestic savings the policy tool available in the World Bank model are public consumption and tax receipts: (a) A reduction in public consumption increases public sector saving while leaving private saving unchanged. (b) An increase in tax revenues raises public sector saving while reducing private saving, but this reduction is smaller than the increase in public sector saving, as the private sector responds to the decline of disposable income by reducing both saving and consumption. 53 The RMSM framework is used, in the context of the two-gap growth models, to determine the growth effects of different levels of foreign financing. For a given growth rate the incremental capital-output ratio determines the required level of investment; if foreign savings is constrained, domestic savings is determined residually. Nothing can ensure in this model that this domestic savings level is consistent with the real level. With the introduction of the exchange rate, the model is determined even with a constraint on foreign inflows: through an exchange rate depreciation, an increase in exports can be induced to produce the foreign exchange needed to purchase the «required» minimum level of imports. The model developed by Moshin Khan, Peter Montiel and Naadem U. Haque 32 in 1990 is an attempt to merge the IMF and the World Bank approach within a common framework, in designing adjustment programs to support lending activities. Another interesting study, presented by Khan and Knight 33 , simulated two different approaches, referred as a «standard» short-term-oriented stabilisation program and an «extended» version. Both of these models are aimed at improving the balance of payments situation; what is different is time horizon. The standard approach is searching for an immediate improvement (a coherent outcome is expected within one year), whereas the extended version is a less intensive, but more prolonged strategy. What the comparison, based on 29 Ldcs, shows is that any reduction in domestic credit and fiscal deficit affects real variables and that different time horizons produce different real costs of adjustment. The shorter is time horizon, more drastic is intervention (strong recession followed by strong improvement of balance of payments) and higher are social costs, in terms of unemployment, and less durable is performance 34 . In the long run, a gradual approach seems to produce an higher income growth rate and/or lower external deficit 35 . 7. A three-gap model Even as its use spread throughout the World Bank and other International Financial Institutions (IFIs), the run of the Financing Gap Model in the academic literature was 32 M. Khan, P. Montiel and N. U. Haque (1990) "Adjustment with Growth: Relating the Analytical Approaches of the World Bank and the IMF", Journal of Development Economics, vol.31. 33 M. Khan, M. D. Knight (1981), «Stabilisation Programs in Developing Countries: A Formal Framework», IMF Staff Papers, Washington D.C. 34 «(...) It took the United States seven years to reduce its fiscal deficit from 6 percent of GNP to below 4 percent. IMF stand-bys, meanwhile, can require lowering a 6 percent fiscal deficit to 3 percent in only one year. Is it logical to ask others to do what one is unwilling to do oneself?». P. Meller (1990), «Comments to "What Washington Means by Policy Reform"», J. Williamson (1990), Latin American Adjustment: How Much Has Happened?, Institute for International Economics, Washington D.C. 35 A. Buira (1983), «IMF Financial Programs and Conditionality», Journal of Development Economics, No. 1. 54 closing36 . A «gap» segment of the literature still exists today in the three-gap model of Bacha 1990 and Taylor 1993, but the spirit of the three-gap model is far from the two-gap model37 . The basic idea is that on growth analysis, it is important not only to look at the efficient allocation of resources, but at the amount of savings (Adam Smith’s approach) and if and how it is translated into productive investment (the Keynesian problem). In the same way, debt crisis can not be reduced to a conventional problem of Balance of Payments, given its structural nature and fiscal implications (which are neglected by the IMF approach). From these main limitations of the two-gaps approach, focused on the abstract idea of savings and operating on the foreign exchange problem, it derived the so-called third gap. As current account deficits become structural, because of the huge amount of foreign debt compared to domestic economy stock and flows, it then implies a fiscal problem, when foreign debt is public or publicly guaranteed, as it is in SSA case. There is a structural gap between fiscal revenue and debt service, which induces a structural constraint on public budget, imposing drastic cut on expenditures and investment. This gap, which can be linked to hyperinflation that induces demonetisation38 and dollarisation39 , represents a main feature of permanent foreign debt crisis and can be modelled as a much more flexible approach. In the three gap models, output is an adjustment mechanism in Keynesian fashion and so it has not a stable relationship to investment. As Taylor 1993 (p.19) puts it, the three gaps model drops «the older gap models’ maintained hypothesis that output is predetermined by capital accumulation» 40 . The amount and maturity of foreign debt repayment obligations in poor countries create dramatic disequilibrium of balance of payments, which should be financed by trade balance surplus. Moreover, given the public nature of foreign debt, foreign disequilibrium is linked to public finance problems, which have exacerbated the problems. The solution of these two 36 The majority view in the literature is based on the 1980s neo-classical critics of development policy: resource allocation is more important than resource quantity. Neo-classical critics emphasised «getting prices right». They pointed out growth failures who had high investment but «wrong» prices. The Financing Gap model had no role for prices in resource allocation and so the neo-classical approach seemed to rule out reliance on the Financing Gap model. The defenders of the use of the Financing Gap model in the IFIs responded that aid and investment were necessary but not sufficient conditions for growth. The idea seemed to be that the Financing Gap model gave the financing requirements for the «necessary» investment in the short-run, while IFI-imposed conditions on getting prices right and other policies would give you the «sufficient» conditions for short-run growth. 37 Over the years a number of other gaps have been proposed, such as the skills gap (lack of technical expertise constrains the level of investment which could be attained.), technology gap, the food gap, the gender gap and the environment gap. 38 Financial component of indebted economies have a role in the vicious circle: because of national macroeconomic crisis, the private sector’s portfolio is much more attracted by foreign assets, inducing an excess of domestic assets supply and increasing demand of international assets. Linked to the problem of tax evasion and elusion, which are increased by the presence of credit rationing in international markets, government is not able to finance its debt and it is obliged to cut expenditure. Taking in account the important flows of capital flights, the net result may be a resource transfer from indebted to foreign countries. 39 Dollarisation (or currency substitution) refers to a situation in which a foreign currency is used, concurrently with the domestic currency, as a unit of account. It may involve a loss of seignorage revenue (i.e. the extraction of real resources by means of base money creation, which reduces the purchasing power of private holdings of cash money balances due to inflation: an inflation tax), because the demand for domestic base money is lower than otherwise, thus leading to increased monetary financing and high inflation, which induces a further reduction of domestic money balances and an inflationary spiral and full dollarisation. 40 See L. Taylor (ed.) (1993), The Rocky Road to Reform: Adjustment, Income Distribution, and Growth in the Developing World, MIT Press, Cambridge and E. L. Bacha (1990), « A Three-Gap Model of Foreign Transfer and the GDP Growth in Developing Countries», Journal of Development Economics, No. 32. 55 transfer problems linked to debt servicing – from State (through budget surplus, the increase of public debt and/or creation of money) to abroad (through foreign debt) – produced conflicts between targets and instruments. The double disequilibrium induced and interacted with macroeconomic instability, such as capital flight, high inflation, and financial system weakness. Given these premises and contrasting the monetarist approach of the IMF, an alternative approach focused on other interpretations of inflation process rather than being simply due to a monetary supply expansion. Indexed contracts determine inflation inertia and behaviour of social actors does not correspond to rational expectations theory. But, especially when the tax system is not indexed to inflation, its efficiency drops as inflation goes up. It is due to lags between tax assessment and tax collection, as real receipts progressively decline as inflation runs faster (Olivera-Tanzi effect) 41 . Thus, inflation may increase government spending on entitlement programs (such as subsidies and transfers), or lead to an increase in tax revenue if tax rates are progressive, but it can also reduce real revenue in the presence of collection lags. The resource and foreign exchange gaps are not the only constraints. Based on the Gerschenkron hypothesis 42 that emphasised the crowding-in effect between private and public investment, Bacha in 1990, a few years after the Chenery and Bruno work appeared, stressed the role of fiscal gap. Fiscal limitations can open a gap between desired and feasible growth rate through the public sector accounts. From one side private investment can be crowded in by public capital formation through complementarities, rather than assuming the presence of a crowding out effect (as in neo-classical and monetarist approaches). And, from another side, public investment is drastically cut after a debt crisis, because foreign obligations owed by the government induce to channel its own resources to honour them. The fiscal gap is a subset of the savings gap, the former may be the binding constraint if there is some limit on public spending (say, through a borrowing target) and private investment is linked to public investment through a crowding in (or out) relationship. Capacity utilisation, i.e. the extent to which new and existing productive capacities (derived from past investments) are utilised, is of major importance for growth in developing countries. Government efforts to increase capacity utilisation are thus important, and involves spending on infrastructure, education, health services etc. Curbing these efforts to increase capacity utilisation can occur when government resources for investment and imports are insufficient, inter alia, as a result of large public debt service; indeed, evidence is available suggesting that government expenditure in the sub-Saharan African region has been curtailed by foreign debt service 43 . External resources directed to the government budget could thus facilitate filling this fiscal gap 44 . 41 The lack of considerations on exact timing of all the effects of macroeconomic policies proposed by monetarists is one of the main weaknesses of their approach. That is why the importance, stressed by M. Khan et al., of introducing expectations on the analysis of adjustment policies: «The introduction of short-run dynamic behaviour, say through lags in adjustment of prices to monetary disequilibrium or through slow revision of expectations to future inflation, while perhaps not changing the overall conclusions, would nonetheless yield useful insights on the time path of prices and thereby the adjustment process». M. Khan, P. Montiel e N. U. Haque (1990), op. cit. 42 A. Gerschenkron (1965), Economic Development in Historical Perspective, Harvard University Press, Cambridge. 43 See D. Fielding, (1997) «Modelling the Determinants of Government Spending in Sub-Saharan Africa», Journal of African Economies, No. 6. 44 P. Hjertholm, J. Laursen and H. White (2000), op. cit. 56 In such a way, the available resources to finance investment are: [(S-I) = private component] + [(T-G) = public component] + [(Ex-IM) = external component] = resources to finance investment (2.4) Given the presence of inter-linkages between all the components of available resources, the reduction of external resources can induce a reduction of domestic public components (public investment), thus discouraging also the private ones. In the case of both fiscal gap and resource gap, government can try to increase the primary surplus (i.e. the difference between taxation earnings and public expenditures) in order to balance the reduction of net foreign transfer. If the government can not move in this direction, then it can try to use deficit spending, but this option can provoke hyper-inflation pressure, particularly when devoted to consumption expenditures, which can be more appropriate than investment if they are less dependent on imports than investment. This will affect less negatively the balance of payments equilibrium. Also when fiscal gap and foreign exchange gap coexist, there is a reduction of investment due to the reduction of foreign net transfer, which can be faced in the short run through the usage of foreign exchange reserves or export-promotion policies. Chisari and Fanelli45 have stressed the inter-dependence between fiscal gap and investment level (which induces capital stock formation), by conceptualising the idea of an hysteresis effect on capital formation in the long-run46 . A reduction in capital stock, due to falling foreign resources, in fact may imply the fallacy of any policy devoted to restore the previous situation. Not only do the stabilisation policies for reducing expenditures appear problematic, but it also leads to an increasing need of foreign resources to restore growth, particularly in a 45 O. O. Chisari and J. M. Fanelli (1990), «Three gap models, optimal growth and the economic dynamics of highly indebted countries», Quaderni del Dipartimento di Economia Politica, N. 101, Università degli Studi di Siena, Siena. 46 Hysteresis represent the history dependence of physical systems. If you push on something, it will yield: when you release, if it does not spring back completely, it is exhibiting hysteresis, in some broad sense. The term is most commonly applied to magnetic materials: as the external field with the signal from the microphone is turned off, the little magnetic domains in the tape don't return to their original configuration. Hysteresis happens in lots of other systems: if you place a large force on your fork while cutting a tough piece of meat, it doesn't always return to its original shape: the shape of the fork depends on its history. Hysteresis loops happen when you repeatedly wiggle the system back and forth (cycle the field up and down). The magnetisation of a tape will “lag behind” as the field sweeps up and as it sweeps down. The memory in the tape is the magnetisation remaining as the field is released to zero from a large value. In magnetic tapes, this lag is repeatable: the shape of the loop after the first cycle is roughly the same as it is after many cycles. (This is convenient for doing multiple recordings on the same tape). Hysteresis was an unpopular subject for decades. Experimentalists generally tried to get rid of it, so they could get publishable equilibrium data. Theorists cringed from thinking about non-equilibrium, dirty materials with long-range elastic or magnetic forces. But styles change: dirt and non-equilibrium are now a major focus of research in physics. See: J. P. Sethna, K. Dahmen, S. Kartha, J. A. Krumhansl, B. W. Roberts, and J. D. Shore (1993), «Hysteresis and Hierarchies: Dynamics of Disorder-Driven First-Order Phase Transformations», Phys. Rev. Lett., N. 70, 3347. The concept of hysteresis (i.e. history dependence, persistence, memory) is linked to that of irreversibility (i.e. temporal asymmetry), as the main state of nature, which represents the basis of the new theories on dynamic equilibria and is directly derived from termo-dynamic literature. Quite interestingly, economic literature has followed the main shifts in physics paradigms: neo-classical equilibrium corresponded to dynamic models of classical physics, both being based on determinism, reversibility, steady-state and predictability; whereas new challenges for social sciences are specifications of dynamic and complex systems, based on uncertainty, bifurcation, probabilities, chaos. These dimensions correspond to theories of physics that emphasise irreversibility, probability, coherence, and they imply the importance of specific history (which means time and institutions) in determining every social phenomenon. Consequently, economics has increasingly devoted its attention to the so-called path dependency, historical inertia, dynamic equilibrium. See N. Georgescu-Roegen (1971), The Entropy Law and the Economic Process, Harvard University Press, Cambridge; N. GeorgescuRoegen (1976), Energy and Economic Myths, Pergamon Press, New York; R. Brunetta and R. Turatto (1992), Disoccupazione, isteresi e irreversibilità, Etas Libri, Milano. 57 context of limited capital stock. A dramatic exogenous shock can not be easily counterbalanced by a political reaction aimed at restoring the previous state. What makes the three-gaps model particularly useful as referred to Sub-Saharan African countries is the fact that it recognise the crucial importance of fiscal problems, rather than emphasising the use of external borrowing for investment purposes (filling the savings gap) in light of the expected future growth path of the economy or addressing the issue of external solvency in order to service debt obligations with foreign exchange (filling the foreign exchange gap). To stress the point written at the beginning of this chapter, it is important, particularly in the Sub-Saharan African countries, to remember that governments borrow to finance fiscal deficits, whereas countries borrow to finance Balance of Payments deficits. The importance of fiscal deficit has a particular nature in Sub-Saharan African countries, where the decision-making process leading to government budget is not based on the planning mechanism experienced in developed countries. In the latter case, in fact, governments fix a desired level of public expenditure, forecast ordinary revenue and define how to finance the residual gap. In Africa, on the contrary, fiscal deficits may reflect a surge in public expenditure, imposed by various lobbies as a way to escape from political conflicts, and in these cases decisions on expenditure and «easy and readily available» rather than optimal financing are simultaneous, at the best 47 . The need of resources as a way of war finance, support to a big push, nationalisation of private sector problem debt, finance of current expenditure implies the risk of excessive public spending financed through external debt. This phenomenon reflects the weakness of fiscal system in Africa, an heritage of structural political, economic and social problems that limits the ability of African governments to generate domestic revenue. The lack of a broad tax base is the most evident limitation, at this regard. Since the 1970s, the expansion of urban population to be subsidised rather than taxed, the increase of recurrent expenditures 48 , the sharp decline of export earnings which followed an initial commodity boom linked to the oil price dynamics, the mechanisms of quasi-fiscal deficit (originated from debt of public sector enterprises), the gradual dismissal of parastatals 49 concurred to increase fiscal deficits. These aspects will be investigated in chapter 4. 8. Some complicating macroeconomic factors. The fungibility argument Reality of foreign borrowing is always complex, as the amount and nature of debt significantly affect domestic savings, government expenditure and revenue (i.e. financing patterns). Debt is not simply financial amounts originated abroad, which are added to funds originated from domestic resources. Fiscal behaviour of governments that receive debt is directed affected by debt itself. An approach of development literature devoted to these complicating factors is the so-called «fungibility» argument 50 , which has been widely used to discuss aid. If an African government has $100 million to be allocated between building hospital or buying arms and it decides to build hospital, when a donor offers the government $100 million of soft loans for health programs, then the African government can free its own resources to buy arms. 47 V. Tanzi (1985), «Fiscal Management and External Debt Problems», H. Mehran (ed.), External Debt Management, IMF, Washington D.C. 48 Unproductive expenditures, such as military ones, crowd out development spending. 49 In a lot of African countries, the use of parastatals was an effective way of broadening the revenue base, through the taxation of agricultural exports. 50 H. Singer (1965), «External Aid: For Plans or Project ?», Economic Journal, No. 75. 58 Without any diversion of fund, the actual impact of debt, measured by the marginal expenditure that is related to it, is to increase military expenditures. Thus, fungibility implies an impact on the types of government expenditure supported, but it also affects the increase of consumption or investment. In fact, Griffin 51 , considering inter-temporal consumption decisions and aid fungibility, demonstrated that future consumption (in time t+1) will be (1+r) times the value of savings in time t (r being the return of capital). In the case of soft loans owed to bilateral donors, external debt can be treated like any other income, and shared between consumption and savings according to their marginal propensities (i.e. recipient Ct+1 wishes), rather than assuming that it is used to increase M investment only. For a given level of income, there is a budget constraint (in the K figure is KL, whereas consumption is at point P and Q domestic savings is L-C 1 t). C2t+1 External debt moves C1t+1 P consumption to Q and domestic savings fall to L-C 2 t. Thus, there is no one-to-one relationship between debt and C t 1 2 savings-investment, as in the 0 L N Ct Ct gap-models. Another approach used to analyse the complicating factors is focused on the fiscal response to external debt. Empirical evidence has been used to test how external debt induces higher capital expenditures (=investment), lower recurrent expenditures (=consumption), lower tax effort (=revenues). As a consequence, these effects can increase or decrease the budget deficit. 9. The additional problem of Dutch Disease Concerns on Sub-Saharan development perspectives have come to be focused around concepts of a «Dutch disease» 52 , which describe potentially negative outcomes associated with an economically significant minerals export sector. As a result of these concerns and other factors, most oil exporting countries have adopted conscious policies of diversification aimed at increasing the economic contribution of other sectors. These concerns are corralled under the term «Dutch disease», which made its appearance when Holland’s manufacturing sector experienced a loss of competitiveness, particularly export competitiveness, following the Schlochteren natural gas discoveries in the 1960s 53 . 51 K. Griffin (1970), «Foreign Capital, Domestic Savings and Economic Development», Bulletin of the Oxford University Institute of Economics and Statistics, No. 32. 52 W.M. Corden (1984), «The booming sector and Dutch disease economics: survey and consolidation», Oxford Economic Papers, No. 36(6). 53 J.J. Struthers (1990) «Nigerian oil and exchange rates: indicators of Dutch disease», Development and Change, No. 21. 59 The problem has been mainly laid at the door of an overvalued currency, the appreciation of which, in turn, was blamed on the strength of Dutch hydrocarbon exports 54 . The Dutch disease or, in its more generic format, the «booming sector model» is therefore the description of an economy in which there is co-existence of a booming minerals sector and a lagging or shrinking sector. The latter was initially identified as manufacturing, but the ideas have also been extended in other economies to cover agriculture. In fact, the standard Dutch disease model incorporates significant movement of resources out of manufacturing into the mineral boom sector. Developing countries have very little manufacturing and therefore few capital or labour resources that would move out even if oil were not an enclave. Main exports, being based on an export enclave model, have little or no connection with domestic economy. On the basis of the particular nature of such manufacturing as does exist in developing countries, Benjamin et al. therefore argue that the developing country variant of the Dutch disease model sees the resource-movement (and spending) effect adversely affecting agriculture rather than manufacturing. In any case, «Dutch disease» hypothesises a wider set of costs to be associated with a booming mineral sector. From the literature 55 , one can distil three main effects/problems that have been associated with a booming mineral exports sector: a) The spending effect. This occurs when part of the additional income generated thanks to the mineral boom is spent in-country on non-traded goods and services (education, health, welfare, construction, other services), leading to excess demand for these since imports cannot flood in to meet demand and since domestic supply constraints exist. As a result, there is price (and, hence, profit) appreciation. In comparative terms, local production of traded goods becomes relatively less profitable and this encourages its relative contraction. b) The resource-movement effect. This occurs when the boom in the minerals sector causes the marginal product of factors employed in that sector to be raised. In other words, as the output price of the extracted mineral rises, so does the apparent productivity of production factors such as labour and capital. Where these resources are mobile, they will be drawn out of other sectors into the booming minerals sector. Put more simply, investors will invest capital in minerals rather than other sectors because their investments will bring higher returns, and workers will also prefer this sector because they will be paid more. c) The currency appreciation effect. According to the standard description, the mineral boom will cause the local currency to strengthen and appreciate. This will hinder exports by increasing their price and will encourage imports by reducing their price, thus discouraging import substitution. In particular, this is felt to create problems for the manufacturing sector as it did in Holland 56 . Referred to foreign debt, the Dutch disease phenomenon basically describes a situation where an inflow of foreign exchange in any form (i.e. from export earnings, loans or grants) puts 54 N.C. Benjamin et al. (1989) «The Dutch disease in a developing country», Journal of Development Economics, No. 30. 55 A. Jazayeri (1986), «Prices and output in two oil-based economies: the Dutch disease in Iran and Nigeria», IDS Bulletin, No. 17(4) and M. Fardmanesh (1991), «Dutch disease economics and the oil syndrome: an empirical study», World Development, No. 19(6). 56 R. Heeks (1998), « Small Enterprise Development and the Dutch Disease in a Small Economy: The Case of Brunei», IDPM Discussion Paper Series, Discussion Paper No. 56, University of Manchester. 60 upward pressure on the real exchange rate of the recipient country by stimulating more rapid domestic inflation. Thus, a large inflow of foreign debt may therefore result in a loss of competitiveness of exports, counteracting other efforts to increase exports. The inflationary effects of foreign debt, however, may be mitigated by the inflow of foreign commodities purchased that increases the supply of commodities in general or reduce supply bottlenecks in the economy, have a deflationary impact, which may exceed the upward pressure on the real exchange rate as a result of the debt. Thus, there are some counteracting effects, and it is not possible a priori to determine what effects an increase in foreign funds will have on the recipient country’s exchange rate, and hence, on the competitiveness of its exports. In any case, in countries where foreign debt plays a crucial role in covering external deficits, it is dangerous an induced real appreciation of the rate of exchange and it is clear that the release of foreign exchange into the domestic economy needs to be in accordance with the absorptive capacity of the economy. An important general rule is that debt must help close resource gaps in the long run, rather than merely filling these gaps in the immediate future. While foreign loans may well be needed to help finance investment, imports and public expenditures, the longer-term perspective should always involve using loans for enhancing recipient countries’ ability to mobilise their own resources. This would imply provision of financial and technical support for mobilisation of the domestic savings needed for investments (closing the savings gap) 57 . The same rule must be applied to closing the forex-gap (the long-run development objective must be that a country’s own export earnings should be sufficient to meet import requirements58 ) and the fiscal-gap (the long-run objective of aid is that a country’s revenue efforts should be sufficient to cover its public expenditures 59 ). Other general rules seem to be much less evident: a case-by-case analysis is the most appropriate way to investigate the relationship between debt and development, as the context and its specific dynamics do matter a lot. 10. Last but not least, the important role of all the players The importance of all the players involved in the «debt game» is a further element to be seriously considered in order to define a more complex map of the inter-relationship between foreign debt and development. And these players change over the countries and years: official lenders prevail in SSA, whereas commercial banks are the bulk of lenders in Latin America. The importance of players’ behaviour in determining the real relationship between debt and development is demonstrated by a further element: the nature, effectiveness, stability and reputation of institutions involved in the debt game as creditors, which facilitate the mechanism of financial intermediation. As analytically described, economic theory assigns a basic role in development process to national and international finance. Loans are viewed as vital inputs in the resolution of growth problems. Since many people and firms are assumed to have unmet credit needs, the delivery of targeted loans becomes a major feature of development activities. 57 J. Kovsted (2000), «Financial Sector Aid», in F. Tarp (ed.) (2001), Foreign Aid and Development: Lessons Learnt and Directions on the Future, Routledge, London. 58 It means that debt should be used to support a conducive policy environment; to finance infrastructural development, and to direct support export promotion. 59 It means that debt should be used to increase government revenue in a manner, which minimises the distortionary nature of taxation, and to promote the effectiveness and efficiency of expenditures. 61 In the theoretical Arrow-Debreu world, based on no information or transaction costs, there is no need of financial intermediation. However, this world is clearly built upon unrealistic assumptions. Intermediaries are essential once imperfections and frictions are introduced in the model: market conditions are not perfect, the economic exchange is costly, information is costly and asymmetrically distributed across agents, technological frictions prevent individuals from having access to economies of scale. Financial intermediaries relax the frictions mobilising savings, better allocating resources, facilitating the trading, hedging, diversifying and pooling of risks. The work of Joseph Schumpeter 60 introduced the idea that the services provided by financial intermediaries are essential for technological innovation and economic development. Empirical work by McKinnon61 and Shaw 62 illustrated the close ties between financial and economic development, whereas influential economists such as Lucas 63 said that the relationship between financial and economic development is «over-stressed». Some recent studies 64 tested the positive relationship between financial depth and growth. They were based on regression analysis for large cross-sections of countries, having the following form: yt = β0 + β1 FDi + β2 Xi + ei (2.5) where yt is the rate of growth of country i, FDi is an indicator of financial depth, Xi is a set of control variables, and ei is the error term. Market imperfections are justification for national and international financial institutions. These perceived imperfections include both fairness and efficiency aspects. Centrally planned economies have been extreme examples of this where free markets are distrusted to do what it is efficient and fair, and where lending is often an integral part of fiscal policy. In mixed economies, imperfections in both financial and non-financial markets are cited to justify administered credit. These include – at national and international level – usurious moneylenders and institutions, dispersion in interest rates on loans, poor people and countries who lack access to market loans, asymmetric information, and bankers who fail to recognise the social and global positive externalities of lending to people and countries with credit needs. This perspective leads to the conclusion that there are many potential borrowers who are credit constrained, because of badly performing financial markets 65 . Given the failures of fifty-years directed loan programs - subsidised loans that are allocated by administrative decisions -, which have been used by governments and donors as broadspectrum policy tools and fuelled foreign debt trap, since the 1980s a contending view emerged. This switch to a new focus to the role, behaviour, interests and responsibilities of financial intermediaries, due to the failure of international borrowing mechanisms, was 60 J. Schumpeter (1912), The Theory of Economic Development, Harvard University Press, Cambridge (English version, 1934). 61 R. I. McKinnon (1973), Money and Capital in Economic Development, Brookings Institutions, Washington D.C. 62 E. Shaw (1973), Financial Deepening in Economic Growth, Oxford University Press, New York. 63 E. R. Lucas Jr. (1988), «On the Mechanics of Economic Development», Journal of Monetary Economics, XXII. 64 See R. G. King and R. Levine (1993), «Finance and Growth: Schumpeter Might Be Right», Quarterly Yournal of Economics, vol. 108, No. 3 and M. Khan and A. S. Senhadji (2000), Financial Development and Economic Growth: An Overview, IMF Working Paper, No. 209, Washington D.C. 65 See J. Stiglitz and A. Weiss (1981), «Credit Rationing in Markets with Imperfect Information», American Economic Review, vol. 71. 62 reinforced by general economic and financial reforms in numerous developing countries that provided an enhanced environment for financial markets. Nowadays, both the traditional Directed Credit Program approach, based on lending mechanisms, and the new Financial Market Paradigm, stressing developing durable and sustained relationships among financial intermediaries, creditworthy clients, and depositors and being much more concerned with the well-being of financial infrastructure, co-exist. That is why, in the meantime, on the international agenda, there are both the issue of debt relief and new lending and the issue of reforming the international financial architecture. The crisis of debt can not be simply reduced to borrower responsibilities. The nature, composition and strategies of lenders to SSA must be seriously considered in analysing debt problems, as they are main players of the “game”, and they are not the same players who play the “game” in Latin America. Referring to the SSA case, as external debt is basically owed to official lenders, political motivations of donors in providing aid and soft loans are particularly important. Quite apart from the measurement difficulties, quoting statistics on the volume, direction, and trends in development assistance is of little relevance without an understanding of the ultimate objectives of foreign aid. In fact, “there is no historical evidence to suggest that over longer periods of time donor nations assist others without expecting some corresponding benefits (political, economic, military, etc.) in return” 66 . ODA from the DAC countries has been generally allocated according to such criteria as strategic foreign policy interests, historic ties, rather than per capita income. “Granting aid to another country is basically a political decision and as such, therefore, a nation’s aid programme is first and foremost a tool of its foreign policy. The historical record of foreign aid provided by traditional donors has amply proved this point. Among the objectives they have tried to achieve through aid, thus serving as its primary motivations, most important have been those in the realms of security, politics, ideology, and economics” 67 . Besides, the biggest simple misconception about foreign aid and debt is that lenders send money abroad. They don’t. Ninety-three per cent of AID funds are spent directly in the United States to pay for equipment, raw materials, expert services, and food... 66 M. P. Todaro (1989), Economic Development in the Third World, Longman, New York, p. 485. 67 S. Hunter (1984), OPEC and the Third World: the Politics of Aid, Croom Helm, London, p. 53. 63 Bibliography E. L. Bacha (1990), « A Three-Gap Model of Foreign Transfer and the GDP Growth in Developing Countries», Journal of Development Economics, No. 32. R. Barro (1988), «The Ricardian Approach to Budget Deficits», NBER, Working Paper, no. 2685. R. J. Barro (1979), On the determination of the public debt, J. Polit. Econ. 87. N.C. Benjamin et al. (1989) «The Dutch disease in a developing country», Journal of Development Economics, No. 30. A. Buira (1983), «IMF Financial Programs and Conditionality», Journal of Development Economics, No. 1 H. B. Chenery and M. Bruno (1962), «Development Alternatives in an open economy: the case of Israel», Economic Journal, No. 72. W.M. Corden (1984), «The booming sector and Dutch disease economics: survey and consolidation», Oxford Economic Papers, No. 36(6). W. Easterly (1999), «The Ghost of Financing Gap. Testing the Growth Model Used in the International Financial Institutions», Journal of Development Economics, No. 60, December. E. Eshag (1971), “Comment”, Bulletin of the Oxford University Institute of Economics and Statistics, Vol. 33, N. 2, May. M. Fardmanesh (1991), «Dutch disease economics and the oil syndrome: an empirical study», World Development, No. 19(6). D. Fielding, (1997) «Modelling the Determinants of Government Spending in Sub-Saharan Africa», Journal of African Economies, No. 6. A. Gerschenkron (1965), Economic Development in Historical Perspective, Harvard University Press, Cambridge. K. Griffin (1970), «Foreign Capital, Domestic Savings and Economic Development», Bulletin of the Oxford University Institute of Economics and Statistics, Vol. 32, N. 2, May. K. Griffin and J. Enos (1970), “Foreign Assistance: Objectives and Consequences”, Economic Development and Cultural Change, Vol. 18. F. Hahn (1977) "The Monetary approach to the Balance of Payments", Journal of International Economics, vol.7. R. Heeks (1998), « Small Enterprise Development and the Dutch Disease in a Small Economy: The Case of Brunei», IDPM Discussion Paper Series, Discussion Paper No. 56, University of Manchester. P. Hjertholm, J. Laursen, H. White (2000), Macroeconomic Issues in Foreign Aid, DERGInstitute of Economics, University of Copenhagen. S. Hunter (1984), OPEC and the Third World: the Politics of Aid, Croom Helm, London. IMF (1977a), The Monetary Approach to the Balance of Payments. Washington D.C. IMF (1977b) "Theorical Aspects of the Design of Fund-Supported Adjustment Programs", IMF Occasional Papers, September A. Jazayeri (1986), «Prices and output in two oil-based economies: the Dutch disease in Iran and Nigeria», IDS Bulletin, No. 17(4) 64 H. G. Johnson (1977) "The Monetary Approach to the Balance of Payments - A nontechnical guide", Journal of International Economics, vol.7 M. Khan, M. D. Knight (1981), « Stabilisation Programs in Developing Countries: A Formal Framework», IMF Staff Papers, Washington D.C. M. Khan, P. Montiel and N. U. Haque (1990) "Adjustment with Growth: Relating the Analytical Approaches of the World Bank and the IMF", Journal of Development Economics, vol.31. M. Khan and A. S. Senhadji (2000), Financial Development and Economic Growth: An Overview, IMF Working Paper, No. 209, Washington D.C. R. G. King and R. Levine (1993), «Finance and Growth: Schumpeter Might Be Right», Quarterly Yournal of Economics, vol. 108, No. 3 J. Kovsted (2000), «Financial Sector Aid», in F. Tarp (ed.) (2001), Foreign Aid and Development: Lessons Learnt and Directions on the Future, Routledge, London. E. R. Lucas Jr. (1988), «On the Mechanics of Economic Development», Journal of Monetary Economics, XXII. G. Macesich (1983), Monetarism, Theory and Policy, Praeger Special Studies, New York R. I. McKinnon (1973), Money and Capital in Economic Development, Brookings Institutions, Washington D.C. R. I. McKinnon (1964), «Foreign exchange constraints in economic development and efficient aid allocation», Economic Journal. J. J. Polak (1957), Monetary analysis of income formation and payments problems, IMF Staff Papers, Vol. 6, Washington D.C. W. E. Robichek (1967) "Financial Programming Exercises of the International Monetary Fund in Latin America", Address to a seminar of Brazilian professors of economics, Rio de Janeiro J. Schumpeter (1912), The Theory of Economic Development, Harvard University Press, Cambridge (English version, 1934). A. Sen (ed.) (1970), Harmondsworth. Growth Economics. Selected Readings, Penguin Books, E. Shaw (1973), Financial Deepening in Economic Growth, Oxford University Press, New York. H. Singer (1965), «External Aid: For Plans or Project ?», Economic Journal, No. 75. J. Stiglitz and A. Weiss (1981), «Credit Rationing in Markets with Imperfect Information», American Economic Review, vol. 71. J.J. Struthers (1990) «Nigerian oil and exchange rates: indicators of Dutch disease», Development and Change, No. 21. L. Taylor (ed.) (1993), The Rocky Road to Reform: Adjustment, Income Distribution, and Growth in the Developing World, MIT Press, Cambridge . M. P. Todaro (1989), Economic Development in the Third World, Longman, New York. H. White (1992), “The Macroeconomic Impact of Development Aid”, The Journal of Development Studies, vol. 28, N. 2, January. 65 3. Stock, Burden and Sustainability of Debt 1. Classification of economies For operational and analytical purposes, the World Bank’s main criterion for classifying economies is now gross national income (GNI) per capita. For the first time, the 2001 edition of the World Development Indicators uses terminology in line with the 1993 System of National Accounts (SNA). Tab. 1 - From the past to the new terminology for classifying economies Previous terminology Gross national product, GNP GNP per capita Private consumption General government consumption Gross domestic investment New terminology Gross national income, GNI GNI per capita Household final consumption expenditure General government final consumption expenditure Gross capital formation Many countries continue to compile their national accounts according to the 1968 SNA, but more and more are adopting the 1993 SNA. A few low-income countries still use concepts from older SNA guidelines, including valuations such as factor cost, in describing major economic aggregates. GNI (formerly referred to as gross national product, or GNP) measures the total domestic and foreign value added claimed by residents. GNI comprises GDP plus net receipts of primary income (compensation of employees and property income) from nonresident sources. The World Bank uses GNP per capita in U.S. dollars to classify countries for analytical purposes and to determine borrowing eligibility. When calculating GNP in U.S. dollars from GNP reported in national currencies, the World Bank follows its Atlas conversion method. This uses a three-year average of exchange rates to smooth the effects of transitory exchange rate. Thus, based on its GNP per capita, every economy is classified as low income, middle income (subdivided into lower middle and upper middle), or high income. Other analytical groups, based on geographic regions and levels of external debt, are also used. Low-income and middle-income economies are sometimes referred to as developing economies. The use of the term is convenient; it is not intended to imply that all economies in the group are experiencing similar development or that other economies have reached a preferred or final stage of development. Classification by income does not necessarily reflect development status. The Bank's analytical income categories (low, middle, high income) are based on the Bank's operational lending categories (civil works preferences, IDA eligibility, etc.). These operational guidelines were established three decades ago, based on the view that since poorer countries deserve better conditions from the Bank, comparative estimates of economic capacity needed to be established. GNP, a broad measure, was considered to be the best single indicator of economic capacity and progress; at the same time it was recognized that GNP does not, by itself, constitute or measure welfare or success in development. GNP per capita is therefore the Bank's main criterion of classifying countries. The process of setting per capita income thresholds started with finding a stable relationship between a summary measure of wellbeing such as poverty incidence and infant mortality on the one hand and economic variables including per capita GNP estimated based on the Bank's Atlas method on the other. Based on such a relationship and the annual availability of Bank's 66 resources, the original per capita income thresholds were established. Thereafter, the original thresholds have been updated every year to incorporate the effect of international inflation, which is now measured by the average inflation of the G-5 countries ("SDR deflator"). Thus, the thresholds remain constant in real terms over time. The economies whose per capita GNP falls below the Bank's operational cutoff for "Civil Works Preference" are classified as low income economies, and those economies whose per capita GNP is higher than the Bank's operational threshold for 15-year IBRD Loans and lower than the threshold for High-income economies are classified as Upper-middle income economies. But as late as 1989, there were some anomalies in the countries included in the middleincome group (a holdover of earlier listings of what constituted "developing" vs. "industrial" countries). An explicit benchmark between the middle-income and high-income countries was established in 1989 at $6,000 per capita in 1987 prices. In general discussions in Bank reports, the term “developing economies” has been used to denote the set of low and middle income economies. Bank publications with notes on the classification of economies state that the term “developing economies... does not imply either that all the economies belonging to the group are actually in the process of developing, nor that those not in the group have necessarily reached some preferred or final stage of development.” These preliminary remarks are aimed at stressing the importance of being very cautious, not only because of problems of availability and reliability (transparency) of data referred to poor countries, but also because the euristic value of statistical concepts, measures and values are subject to criticism, too. 2. Definitions of groups The World Bank tables classify all World Bank member countries (183), and all other economies with populations of more than 30,000 (207 total). In terms of income group, economies are divided according to 2000 GNI per capita, calculated using the World Bank Atlas method. The groups, taking in account the 48 SubSaharan African (SSA) countries, are: Tab. (a) Low-income economies, $755 or less (63 countries, 38 from SSA) Angola Benin Burkina Faso Burundi Cameroon Central African Republic Chad Comoros Congo, Dem. Rep Congo, Rep. Cote d'Ivoire Eritrea Ethiopia Gambia, The Ghana Guinea Guinea-Bissau Kenya Lesotho Liberia Madagascar Malawi Mali Mauritania Mozambique Niger Nigeria Rwanda Sao Tome and Principe Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Zambia Zimbabwe 67 (b) Lower-middle-income economies, $756- $2,995 (54 countries, 4 from SSA ) Cape Verde Equatorial Guinea Namibia Swaziland (c) Upper-middle-income economies, $2,996- $9,265 (38 countries, 6 from SSA) Botswana Gabon Mauritius Mayotte Seychelles South Africa (d) High-income economies, $9,266 or more (52 countries, 0 from SSA). Thus, considering developing countries only (a+b+c), there are: Ø 23 from East Asia and Pacific Ø 8 from South Asia Ø 28 from Europe and Central Asia Ø 32 from Latin America and the Caribbean Ø 17 from Middle East and North Africa Ø 48 from Sub-Saharan Africa. Linked to the poor social and economic performance of countries, there is another sub-group of countries. Forty-nine countries 1 are currently designated by the United Nations as the "least developed countries" (LLDCs). The Economic and Social Council (ECOSOC) review the list every three years. The criteria underlying the current list of LLDCs are: 1. a low income, as measured by the gross domestic product (GDP) per capita; 2. weak human resources, as measured by a composite index (Augmented Physical Quality of Life Index) based on indicators of life expectancy at birth, per capita calorie intake, combined primary and secondary school enrolment, and adult literacy; 3. a low level of economic diversification, as measured by a composite index (Economic Diversification Index) based on the share of manufacturing in GDP, the share of the labour force in industry, annual per capita commercial energy consumption, and UNCTAD's merchandise export concentration index. Different thresholds are used for inclusion in, and graduation from, the list. A country qualifies to be added to the list of LLDCs if it meets inclusion thresholds on all three criteria. A country qualifies for graduation from the list if it meets graduation thresholds on two of the three criteria. For the low-income criterion, the threshold on which inclusion in the current list is based has been a GDP per capita of $800, and the threshold for graduation has been a GDP per capita of $900. In its July 2000 review, in the light of recommendations by the Committee for Development Policy, ECOSOC declared the eligibility of Senegal for designation as an LLDC (subject to the Government so desiring) and decided to postpone its consideration of Maldives' graduation. In such a way, there is a partial overlapping with the World Bank economic criterion, in fact most of the Low-income economies and only a few of the Lower-middle-income economies 1 In early 2001, following the triennial review of the list of LLDCs, Senegal was placed in the category, bringing the total to 49. 68 are within the LLDC group. The criteria for determining the list of LLDCs are under review. The Committee for Development Policy has recommended that the Economic Diversification Index be replaced by an Economic Vulnerability Index reflecting the main external shocks to which many low-income countries are subject, and incorporating the main structural elements of the countries' exposure to the shocks, including their smallness and lack of diversification. Within the LLDCs group, there are some 34 countries from SSA: (e) LLDCs (49 countries, 35 from SSA: 32 out of 38 from [a] and 3 out of 4 from [b]) Angola Benin Burkina Faso Burundi Cape Verde Central African Republic Chad Comoros Congo, Dem. Rep Congo, Rep. Equatorial Guinea Eritrea Ethiopia Gambia, The Guinea Guinea-Bissau Lesotho Liberia Madagascar Malawi Mali Mauritania Mozambique Niger Rwanda Sao Tome and Principe Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Zambia Real GDP per capita in the LLDCs grew at only 0.9 per cent per annum during the 1990s, and excluding Bangladesh, by only 0.4. 3 - External debt stock Developing countries can raise external funds from the international financial community and finance their development through a number of instruments, including attracting equity and foreign direct investment, receiving grants from donors or borrowing from foreign lenders. The Debtor Reporting System (DRS) was set up in 1951 to monitor these statistics by the World Bank, who used the DRS data, in combination with information obtained from creditors through the debt data collection systems of other agencies, such as the Bank for International Settlements (BIS) and the Organisation for Economic Co-operation and Development (OECD). Recently, the international task-force on financial statistics has been reorganised: a group of international agencies working together under the auspices of the Inter-Agency Task Force on Finance Statistics (TFFS) - an interagency task force endorsed by the UN Statistical Commission which was re-convened in 1998 to co-ordinate work among the participating agencies to improve the quality, transparency, timeliness and availability of data on external debt and international reserve assets - has produced a new Guide 2 . The purpose of the Guide is to provide comprehensive guidance for the measurement and presentation of external debt statistics; it can be considered an update of the Grey Book, which was published in 1988 by the BIS, IMF, OECD, and World Bank 3 . The Guide also provides advice on the compilation and analytical use of external debt statistics, and it is intended to contribute to an improvement in, and a greater understanding of, external debt statistics. In drafting the Guide, existing international guidelines are drawn upon, notably the Grey Book, the System of National Accounts 1993, and the fifth edition of the IMF's Balance 2 TFFS (2001), Draft of the External Debt Statistics: Guide for Compilers and Users, Washington D.C. 3 BIS, IMF, OECD, and World Bank (1988), External Debt: Definition, Statistical Coverage and Methodology, Washington D.C. 69 of Payments Manual (1993) 4 . The World Bank is the main repository for statistics on the external debt of developing countries on a loan-by-loan basis. The outstanding external debt has several important characteristics. A basic distinction is between total debt stocks and total debt flows. Fig. 1 – Total external debt stock of Ldcs (1980-2000) 3000 2563.6 2527.5 2500 2000 1459.9 1500 1000 586.7 500 0 1980 1990 1999 2000 Source: Global Development Finance, 2001 Concerning total debt stocks, for operational and analytical purposes, we have to consider: Ø The maturity: Short-term ( one year), medium and long-term (> one year) and IMF credits (very short debt). The problems of liquidity and temporary Balance of Payments deficits should be faced by short-term, whereas development strategies should require long-term debt, which is defined as debt that has an original or extended maturity of more than one year and that is owed to non-residents and repayable in foreign currency, goods, and services. The use of IMF credit denotes repurchase obligations to the IMF with respect to all uses of IMF resources (excluding those resulting from drawings in the reserve tranche) shown for the end of the year specified. Ø The nature of debtors : a sovereign government, a public or a private debtor. Ø The presence of guarantees: Referring to long-term external debt, available data permit a distinction between private nonguaranteed debt, public and publicly guaranteed debt. Private nonguaranteed debt is an external obligation of a private debtor that is not guaranteed for repayment by a public entity, differently from the publicly guaranteed debt. Public debt is referred to a public debtor, including the national government, a political subdivision, and autonomous public bodies. Ø The nature of creditors : Official or private creditors. Among the official creditors, there are multilateral creditors (The International Financial Institutions such as the IMF, the World Bank or Regional Development Banks) and bilateral creditors (governments or their appropriate institutions through [a] soft loans, which is an ODA 4 IMF (1993), IMF's Balance of Payments Manual, Washington D.C. 70 component, [b] commercial credits insured and indemnified by Export Credit Agencies, and [c] commercial banks’ debts guaranteed by State). Multilateral debt has a preferential status, i.e. it is the first to be repaid. Bilateral credit of industrial countries is negotiated by the informal group called the Paris Club 5 , as well as the London Club is the forum through which some commercial banks negotiate their credits. The Paris and London Clubs There are two main forums for debt renegotiations. The first one deals with the debt granted from official source, that is from governments or governmental agencies, regardless of whether the debtor is a public or a private entity. This forum is called the Paris club. On the other forum, which is known as the London club, the debtor governments negotiate with their commercial bank creditors. This categorisation reflects that the negotiations are organised from the perspective of the creditor, not the debtor. Neither club has a formal or permanent institutional structure, but a set of procedures is associated with each club. These procedures are used for negotiating agreements with debtors who are unable to meet their external debt obligations. The Paris Club is an informal group of official creditors whose role is to find co-ordinated and sustainable solutions to the payment difficulties experienced by debtor nations. There are 19 Paris Club permanent members, who are governments with large claims on various other governments throughout the world (all the G8 members are within the 19 permanent governments), plus other official creditors invited on a case-by-case basis (South Africa is among them). Paris Club creditors agree to rescheduling debts due to them. Rescheduling is a means of providing a country with debt relief though a postponement and, in the case of concessional rescheduling, reductions in debt service obligations. The first Paris club was convened in 1956 in response to Argentinean request for official debt relief. The main incentive was to ensure uniformity of treatment among numerous European creditors, to avoid time-consuming parallel bilateral negotiations and to make the bargaining power of lenders stronger for negotiations. Since then, the Paris Club or ad hoc groups of Paris Club creditors have reached 336 agreements concerning 75 debtor countries. Since 1983, the total amount of debt covered in these agreements has been $374 billion. In spite of such an activity, the Paris Club has remained strictly informal. It is the voluntary gathering of creditor countries willing to treat in a co-ordinated way the debt due to them by the developing countries. It can be described as a “non institution” with a strong power over institutions. Although the Paris Club has no legal basis or status, agreements are reached following a number of rules and principles agreed by creditor countries. The creditor countries meet 10 to 11 times a year, for negotiation sessions or to discuss among themselves the situation of the external debt of debtor countries or methodological issues on the debt of developing countries. These meetings are held in Paris. The Chairman is a senior official of the French Treasury. This informal organisation has made it possible for the Paris Club to work with limited staff and a flexible structure. Traditionally in the Paris club negotiations the debtor is a developing country and on the other side of the table are the governments of the country's official creditors which commonly are OECD members. Although in theory all official creditors can attend the restructuring negotiations of a given debtor, in practise only the largest ones do. The Paris club negotiations can be characterised with three interrelated principles: imminent default, conditionality, and burden sharing. The first principle determines that the club will not enter into negotiations before the requesting country shows evidence that it will default on its debt unless relief is given. In the beginning the club provided only short-term relief, but with the deepening of the debt crisis also restructuring agreements 5 71 including partial debt forgiveness, low interest rates, and extended maturity periods have been concluded. The second principle, conditionality, is implemented with the assistance of the International Monetary Fund. In practise, the club does not enter into negotiations with a debtor country, which has not concluded a standby agreement with the IMF covering the period for which the relief is requested. The existence of the IMF agreement is an ensurement for the creditor that the country will enact policies that improve the ability to service its debt. In addition, the IMF provides a monitoring mechanism, which benefits the creditor to some extent. The last principle requires the debtor countries to seek comparable relief from all of their various creditors. Similar principles are applied in the London club. The key difference between the London club and the Paris club is the status of the creditors. The former ones being commercial banks, they are more concerned with profit margins than the official creditors and thus the terms of the agreements are often less concessional. The London club, for instance, almost never reschedules interest obligations. Unlike the Paris club negotiations, the London club restructuring must include all commercial banks with exposure to the debtor country, even the minor ones. The agreements are concluded as concerted actions in order to eliminate the free rider problem among the creditors. Due to a great number of participants, the negotiations are time consuming and costly - they tend to be drawn over several months while the Paris club negotiates a comparable deal within a day. Regardless of some major differences in the procedures of the two clubs, they share the common purpose of controlling and restructuring developing country debt in the most timely and efficient manner possible. Because a good deal of emphasis is put on the comparability of treatment among the creditors, the restructuring exercise for a given country necessitates some degree of co-ordination between the two creditor clubs. The multilateral financial institutions often offer the forum for the cooperation and thus have a central role in the debt restructuring process. Participation of the IMF and the World Bank is seen to have a catalytic effect on the availability of funds as well as relief for the developing countries from both official and unofficial creditors. Their involvement in the debtor countries can reduce the cost debt restructuring to the creditors, because of the concessional nature of the IMF and World Bank loan packages. Also, as mentioned before, the existence of the standby agreement with the Fund is a precondition for creditor clubs to enter into restructuring negotiations. Fig. 2 - External debt stock and its components Debt components Short-term Long -term use of IMF credits By debtors private nonguaranteed debt public and publicly Guaranteed debt By creditor Official creditors multilaterals bilaterals Source: World Bank, 1996 72 Private creditors commercial banks bonds other Over time, the share of private debt (debt owed by private debtors) and the share of private claims (debt owed to private creditors) has increased, reflecting the increased role of the private sector in world economy. As of December 31, 1999, the total debt of developing countries (including countries in transition) was estimated by the World Bank 6 to be $2,550 billion, out of which 2,070 billion of medium-and-long-term debt, broken down as follows: Tab. 3 - total debt of developing countries Long term debt outstanding - Public-and-publicly-guaranteed - Official-creditors - Multilateral - Bilateral - Private creditors - Private non-guaranteed Source: World Bank 2001/b 1999 2 070,7 100% 1 580,1 76% 875,5 345,7 529,7 704,6 490,6 24% Medium-and-long-term debt can be broken down in bilateral, multilateral and private. Medium-and-long-term debt Bilateral 25% Private Multilateral 58% 17% Source: World Bank 2001/b This picture shows the prevalence of private over public debt (58% vs. 42%), and within public debt, the higher percentage of bilateral over multilateral debt (25% vs. 17%). Within the private category, the main components are commercial banks’ lending and government bonds. Particularly interesting is the role of private banks based in offshore centres. Public-and-publicly-guaranteed medium-and-long-term debt can be broken down as follows: 6 World Bank (2001/b), Global Development Finance, Washington D.C. 73 According to the IMF 7 , the total debt of developing countries has increased as follows over the past eight years: Source: IMF (2000) Concerning total debt flows, there are some components to be considered: Ø Disbursements : drawings on loan commitments during the year specified; Ø Principal repayments : the amount of principal (amortisation) paid in foreign currency, goods, or services in the year specified; Ø Net flows on debts: disbursements minus principal repayments; Ø Interest payments : the amount of interest (that can be fixed or variable, that is floating with movements in a key market rate such as the London inter-bank offer rate – LIBOR – or the US prime rate) paid in the year specified; Ø Net transfers on debt : net flows minus interest payments (= disbursements minus total debt service payments) Ø Debt service: the sum of principal repayments and interest payments made in the year specified. 7 IMF (2000), World Economic Outlook, Washington D.C., October. 74 However, the absolute stock and flows of external debt are not enough to define the profile of debt burden, particularly referring to SSA economies. 4. Debt Burden indicators and country classification Among the external sources of finance (FDI and portfolio equity flows, ODA and debt), debt results directly in future obligations for the borrower (debt must be repaid). This makes it necessary for the borrower to make sure that it will be in the future in a position to repay its debt, notably through an efficient use of the loans, in order to generate funds that will be used to repay the debt. This is why debt is often considered as a development tool. Traditionally, external debt was considered as contractual obligations, which had to be met in full and on time. If a country could not meet its service payments it was assumed to have adopted bad budgetary and monetary controls and fiscal discipline. Debt relief was only invoked in exceptional circumstances, to be treated on its own merits and without prejudice to other cases: an ad hoc and last resort measure, to be limited to the minimum amount necessary to restore the debtor’s credit-rating and to resume service payments, defined to minimise creditors’ losses. For the poorest and most indebted countries, the accumulated debt burden had become a drawback for development; this is why the international financial community designed mechanisms to rescheduling or cancelling debt. Without taking in account the possibility of debt repudiation, which arises when a debtor country refuses to recognise the legitimacy of the debt contract, indebtedness always implies the risk of default. This risk can induce to: 1. moratorium: service payments, on all or parts of its obligations, are (temporarily) suspended; 2. Deferral: a debt treatment may defer some debt due immediately or in the near future to a later date; 3. refinancing: creditors make a new loan that is used to repay the existing debt, particularly referred to long-term structural debts; 4. Reprofiling: part of the debt may be reprofiled over a few years, instead of a longterm period of time. The duration of a reprofiling is an intermediate between a deferral and a long term rescheduling; 5. rescheduling: creditors change radically the terms and conditions of the existing debt (through partial cancellation, extended maturities, concessional interest rates), without reducing the total debt outstanding; 6. swaps: debt conversion schemes such as buy-backs, or the discounted value of longterm bonds that were issued in exchange for outstanding debt realised to reduce debt stock. 7. forgiveness/cancellation: part of the claims or total debt (interest or debt stock) is written off, as the final stage in the process of debt consolidation. Any mechanism can imply concessional terms, if there is a reduction of the net present value of the claims. In the context of a concessional treatment, creditors may usually choose among a number of options to provide the required debt reduction in net present value of debt stock. When the creditor chooses the debt reduction option, the net present value reduction is achieved through a cancellation of part of the claims. 75 Otherwise, creditors can choose a debt service reduction, with which the net present value reduction is achieved through a rescheduling of the claims at an interest rate lower than the appropriate market rate. Interesting mechanisms offered to creditors by financial market are debt swaps. These operations may be debt for nature, debt for aid, debt for equity swaps or other local currency debt swaps. These swaps often involve the sale of the debt by the creditor government to an investor who in turn sells the debt to the debtor government in return for shares in a local company or for local currency to be used in projects in the country8 . Paris Club creditors and debtors regularly conduct a reporting to the Paris Club Secretariat of the debt swaps conducted. Given the default risk and its implication in terms of debt relief and creditor’s costs, international financial community introduced some indicators in order to monitoring the capacity of debtor countries to service their debt. These indicators are usually based on two ratios: Ø The ratio of the present value 9 of total debt service to GNP (EDT/GNP). This is considered a proxy of debt burden in terms of the broadest measure of income generation in an economy. Ø The ratio of the present value of total debt service to exports (TDS/XGS)10 . This ratio, called debt service ratio, is considered a proxy of debt burden in terms of the activities, which provide foreign exchange to service debt. It captures the impact of debt-service obligations on foreign exchange cash flow. This is a measure of a country’s foreign exchange constraint, taking in account debt relief provided by debtservice rescheduling and arrears on payments. A limitation of this indicator is that, given the uncertain nature of debt relief, a low value may reflect an unwillingness to pay. As a very rough rule of thumb, a ratio of 15-20 per cent or over has been considered the danger point, that is the limit beyond which debtor countries experience severe difficulties in servicing their foreign debts while maintaining existing levels of services domestically. But debt service ratios are essentially an index of short-term liquidity, and do not take in account of other major parameters such as changes in imports and new capital inflows, export potentials, foreign exchange reserves 11 . These two ratios measure two important aspects of an economy’s potential capacity to service the debt. Standard World Bank definitions of severe and moderate indebtedness are used to classify economies. Severely indebted means either of the two key ratios is above critical levels: present value of debt service to GNP (80 percent) and present value of debt service to exports (220 percent). 8 As in the strategy adopted in 2000 by Vatican – in particular through the Comitato Ecclesiale italiano per la riduzione del debito estero dei Paesi più poveri, Conferenza Episcopale Italiana - for Zambia and Guinea. 9 The net present value (NPV) of debt is a measure that takes into account the degree of concessionality. It is defined as the sum of all future debt-service obligations (interest and principal) on existing debt, discounted at the appropriate market rate. Whenever the interest rate on a loan is lower than the market rate, the resulting NPV of debt is smaller than its face value. The NPV of debt makes it possible to sum up loans with different maturities. 10 Earnings from goods and services, including worker remittances and without including official grants. 11 Moreover, in SSA countries, the improvement in the ratio occurred in the last 15 years is not due to any improvement in the ability to service the external debts, but to consolidation of debts or transformation into arrears. 76 Moderately indebted means either of the two key ratios exceeds 60 percent of, but does not reach, the critical levels: that is 48 % for the present value of debt service to GNP and 132% for the present value of debt service to exports. If both ratios are less than 60 percent of the critical value, the country is classified as less indebted. Combining these criteria with the definition of countries in terms of income group, there are six categories of economies 12 : (f) Severely indebted low income (33 economies, 27 from SSA) Angola Ethiopia Nigeria Benin Guinea Rwanda Burundi Guinea-Bissau Sao Tome and Principe Cameroon Liberia Sierra Leone Central African Republic Madagascar Somalia Comoros Malawi Sudan Congo, Dem. Rep. Mali Tanzania Congo, Rep. Mauritania Uganda Cote d'Ivoire Niger Zambia (g) Severely indebted middle income (10 economies, 1 from SSA) Gabon (h) Moderately indebted low income (19 economies, 9 from SSA) Burkina Faso Ghana Senegal Chad Kenya Togo Gambia, The Mozambique Zimbabwe (j) Moderately indebted middle income (24 economies, 1 from SSA) Mauritius (k) Less indebted low income (9 economies, 2 from SSA) Eritrea Lesotho (l) Less indebted middle income (41 economies, 7 from SSA) Botswana Cape Verde Equatorial Guinea Namibia Seychelles South Africa Swaziland The use of critical values to define the boundaries between indebtedness categories implies that changes in country classifications should be interpreted with caution. If a country has an indicator that is close to the critical value, a small change in the indicator may trigger a change in indebtedness classification even if economic fundamentals have not changed significantly. Moreover, these two basic indicators do not represent an exhaustive set of indicators of external debt. They may not capture the debt servicing capacity of countries in which government budget constraints are key to debt service difficulties. Countries that allow the use of free conversion of foreign currency - such as the Franc Zone countries - can face government budget difficulties, related to servicing external public debt, which are not reflected in balance of payments data. 12 Among the 48 low, lower-middle and upper-middle income SSA countries, only Mayotte is not classified by indebtedness. 77 In other countries, the servicing of domestic public debt may be a source of fiscal strain that is not reflected in balance of payments data. Raising external debt may not necessarily imply payment difficulties, especially if there is an increase in the country’s debt servicing capacity. Referring to debt service ratio, even an high value can be “sustainable” in the long-run if the export revenues are high compared to import demand. Moreover, where debt service is a fiscal concern – as it is in SSA – and export receipts are not totally translated into public income, then this ratio may be misleading. And, if the structure of debts’ maturity, in a given country, implies that most of them are grouped in the same years, then what is really a liquidity problem may be confused with an excessive indebtedness problem. Thus, there are some other indicators, which are used to assess the external situation of indebted developing countries: Ø Total external debt to export (EDT/XGS); Ø Total interest payments to export of goods and services, called the interest service ratio (INT/XGS); Ø Total interest payments to GNP ratio (INT/GNP); Ø The international reserves to total external debt ratio (RES/EDT); Ø The international reserves to imports ratio (RES/MGS); Ø The Short-term debt to total external debt ratio (STD/EDT); Ø The Concessional debt to total external debt ratio (CTD/EDT); Ø The Multilateral debt to total external debt ratio (MTD/EDT); Ø Total external debt to total debt service ratio (EDT/TDS) Ø The fixed imports to total imports ratio (INT/MGS); Ø The total imports to GNI ratio (MGS/GNI) Two other conventional measures aim at capturing the fiscal impact of the external debt burden, which is relevant in SSA context: Ø The ratio of scheduled interest payments to government revenue, which measures the country’s capacity to repay as scheduled (INT/T). This ratio is considered low of it is less than 0.2, high if it is above 0.5; Ø The ratio of scheduled interest payments to government expenditure, which measures the constraint imposed by debt servicing to the country’s ability to expand other (current and capital) expenditures. An alternative measure takes in account the discounted present value of future debt-service obligations. This measure is obtained by taking all future debt service obligations (including interest payments at the original rate of the loan and amortisation payments) until full repayment of the debt, and dividing them by a factor based on a given discount rate. For example, the World Bank uses the ratio of this value to current exports. The problem of these specific measures is that this measure is highly sensitive to the discount rate assumed to calculate present values, it does not provide information about the debt-service profile and it does require complex data (given the huge uncertainty on the future). For economies that do not report detailed debt statistics to the World Bank Debtor Reporting System (DRS) 13 , present-value calculation is not possible 14 . 13 Namibia is not reported in the World Bank’s Debtor Reporting System (DRS) 78 Tab. 4 – Debt Indicators, 2000 East Asia and Pacific EDT/XGS 74,8 EDT/GNI 32,6 10,8 TDS/XGS 4,0 INT/XGS 1,7 INT/GNI 17,0 STD/EDT 16,5 CTD/EDT 11,8 MTD/EDT Source: World Bank, 2001 Europe and Central Asia 114,4 46,4 14,6 5,3 2,2 15,4 5,8 7,4 Latin America and the Caribbean 172,6 38,5 35,7 11,8 2,6 15,6 3,7 11,6 Middle East and North Africa 93,8 31,2 10,9 4,7 1,5 24,3 29,3 11,1 South Asia SubSaharan Africa 156,0 26,5 13,1 5,4 0,9 3,6 54,7 37,3 180,2 66,1 12,8 4,3 1,6 16,5 41,2 26,1 Tab. 5 – Income and Indebtedness classification criteria Income classification Low income: GDP per capita less than $785 Middle income: GDP per capita between $786 and $9,655 Severely indebted low-income countries PV\XGS less than220% but higher than 132% or PV\GNP less than 80% but higher than 48% Moderately indebted lowincome countries Severely indebted middle-income countries Moderately indebted middleincome countries PV\XGS higher than220% or PV\GNP higher than 80% PV\XGS less than132% and PV\GNP less than 48% Less indebted lowincome countries Less indebted middle- income countries Source: World Bank, 2001 5. Creditworthiness. Another concept which varies from country to country The creditworthiness of a country is the estimate by potential lenders of the capacity of the country to repay its external debt. Being creditworthy is a key to success for developing countries, as this makes it possible for them to borrow larger amounts to finance growth development. In addition, a creditworthy debtor is in a position to borrow funds used to refinance its existing debt obligations. Governments of debtor countries have a significant impact on the creditworthiness of all borrowers of developing countries, since default of the government can have consequences for the capacity of other borrowers to repay their debt. A number of factors influence creditworthiness. Some are linked to economic factors, such as the capacity of the country to generate balance-of-payment receipts, the volatility of these receipts. Others are financial factors, such as the debt repayment profile. Political factors also play a role in the creditworthiness of a debtor country, when its government considers the cost of paying debt to be too high. 14 In this case, the following methodology is used to classify the non-DRS economies. Severely indebted means three of four key ratios (averaged over 1997-99) are above critical levels: debt to GNI (50 percent); debt to exports (275 percent); debt service to exports (30 percent); and interest to exports (20 percent). Moderately indebted means three of the four key ratios exceed 60 percent of, but do not reach the critical levels. All other classified low- and middle-income economies are listed as less indebted. 79 Creditworthiness usually takes a long time to build, as lenders tend to assess over time the capacity of the debtor to repay its debt before entering into large lending. In contrast, nonpayment of debt obligations can rapidly damage creditworthiness. Under the circumstances where debt restructuring cannot be avoided, countries that do not accumulate arrears 15 and take preventive steps to reach a co-ordinated solution with their creditors, notably in the Paris Club, can restore their creditworthiness more rapidly afterwards. In contrast, debtors that declare a unilateral moratorium tend to lose access to new financing for some time. Some organisations publish estimates of the capacity of debtors to repay their debt. For debtors with debt market exposure, rating agencies estimate creditworthiness. External debt as % of GNP, 1998 Source: OECD 6. The ambiguous concept of External Debt sustainability Implicitly, two different aspects can be considered to evaluate the sustainability of external debt: (1) debt capacity problems, in which the debtor is unable or unwilling to pay debt obligations; (2) economic development problems, in which a country’s external debt adversely affects development process, independently on the fact that it is well serviced or not. Obviously, it is not sufficient to consider sustainable a debt with a negative growth rate of indebtedness16 . In fact, in economic literature three different approaches, partially corresponding to the three gaps approach, have been experienced to analyse debt capacity. 15 Debt due and not paid as of a given date. Arrears may be late payments as well as debt due a long time before 16 M. Simonsen (1985), “The Developing Country Debt Problem”, in G. Smith and J. Cuddington (eds.), International Debt and the Developing Countries, The World Bank, Washington D.C.; L. Spaventa (1987), “The Growth of Public Debt”, IMF Staff Papers, Vol.34, N. 2, June, pp.374-399. 80 (1) Growth-cum-debt literature, to analyse whether foreign debt is supported by output growth. As development proceeds, changes in domestic income, rates of savings, accumulation of capital stock, and rates of return on investment can be expected and countries move from a debtor position to that of the creditor. But there is no automaticity in this process: the main condition to be met is high level of economic growth: it must exceed (or at least equal) the rate of interest. (2) External performance literature, to analyse external solvency through the focus on the external performance of an economy dependent on external borrowing. The basic idea is that preservation of debt capacity requires more than growth (even if it is high): foreign borrowing is conducted in foreign exchange, thus the savings surplus has to be converted into foreign exchange in order to effect the transfer of debt payments. Export performance relative to the cost of borrowing is of great importance. To maintain debt capacity, the rate of growth of exports must exceed (or at least equal) the rate of interest. When exports grow faster than debt, the borrowing country does not have to contribute any of its domestic resources when servicing debt. The debt dynamics approach stresses the need for adjustments in the trade balance in order to maintain debt capacity. (3) Fiscal dimension literature, to analyse the links between public debt and public revenue. In fact, exclusive focus on the balance of payments ignores internal constraints on debt capacity, even though in poor countries these constraints can be of a structural nature. This approach is focused on the “internal transfer problem”. If the government uses external debt for investments in infrastructure, education, health, the sustainable level of debt will depend, not only on the relationship between the marginal social return on these investments and the marginal cost of borrowing, but also on the governments ability to appropriate domestic resources for debt service (i.e. an expanding tax base). The first study, sponsored by the World Bank, to analyse debt capacity of Ldcs was written in 1958 by D. Avramovic 17 . In the context of post-war development policies, which involved both Europe and Ldcs, the main concern was the repayment flows due to increased international lending through the World Bank, among others. This issue was an extension of the macroeconomic rationale for external resource flows, described in the previous chapter. As Salop and Spitäller said 18 , debt capacity issue was mainly referred to the identification of the optimal level of debt, given the terms and conditions of available money, as it seemed to adequately adhere to the economic literature on optimisation. The basic theoretical idea was that the optimal level of debt is that at which the marginal benefits and costs of foreign borrowing are equalised. However, the translation of this idea into operative terms referred to specific countries was not easy. Quite differently, derived from the literature on the links between debt and development originated by the Harrod-Domar growth model, another approach considered debt capacity in terms of effects induced to growth and development. From this point of view, debt role changes as development process changes. Again, the basic condition to maintain debt service capacity is that the growth rate of output should equal or exceed the cost of borrowing (equal to the rate of interest). Debt capacity analysis is focused on debt servicing problems, in order to: 17 D. Avramovic (1958), Debt servicing capacity and post-war growth in International indebtedness, Johns Hopkins Press, Baltimore. 18 J. Salop and E. Spitäller (1980), “Why Does the Current Account Matter?”, IMF Staff Papers, No. 27, Washington D.C. 81 (a) find the critical level, that is the values of some particular ratios at which the economy shifts from performing to non-performing; (b) assess the amount of debt relief needed to solve the problems; (c) define appropriate policy remedies. Many countries – both developing and developed – continue to face the challenge of a sustainable pattern of debt. A portfolio review of debt must be assessed according to: (a) different debt characteristics, which have been previously described: external and domestic debt; short and long term; public, publicly-guaranteed and private lender-based; by the sector and type of project being financed; average life and duration; present value/grant element; (b) existence and percentage of non-debt financing: aid grants 19 , foreign direct investment, portfolio investment, bank flows, returning flight capital; the terms, sources and currencies of these flows; the destination sector and projects; the type of investors. Obviously, it is not recommended to rely on simplistic assumptions, nevertheless simple hierarchies of volatility among different flows (foreign direct investment and long-term loans are more stable than portfolio and short-term loans) can be a useful first step. The currency composition and maturity profile and the nature of foreign liabilities are other important concern: coeteris paribus, a higher incidence of foreign-currency denominated debt, a higher ratio of short-term debt to total debt, the bunching of debt redemption, and variable interest rates increase the risk of an external crisis. Based on this type of analysis, there are some preliminary broad rules to be taken in account: • firstly, the so called “Golden Rule”: government should not conduct net borrowing to fund recurrent spending over the life of an economic cycle; • secondly, the IMF rules: ceilings on overall borrowing, non- concessional loans and shortterm debt must be set, linked to the per capita GDP of the country; • thirdly, grants or soft loans should be used for projects without any direct economic rate of return. Then, some static ratios of external indebtedness can be considered 20 . All the debt ratios previously described can be adopted. The key issue is to define the levels considered sustainable for the various ratios. Tab. 6 – Unsustainable ratios Present value/export Present value/budget revenue Present value/GNI Nominal debt stock/exports Debt stock/GNI Debt service/exports Debt service/budget revenue Interest/exports Source: M. Martin, 1999. World Bank 220 n.a. 80 275 50 30 n.a. 20 HIPC-1 200-250 280 n.a. n.a. n.a. 20-25 n.a. n.a. HIPC-2 150 250 n.a. n.a. n.a. 15-20 n.a. n.a. DFID/EFA 140 155 n.a. n.a. n.a. 13 12 n.a. 19 A useful concept is the aid dependence, measured in terms of budget by aid as a proportion of capital or recurrent expenditure, or in terms of balance of payments as a proportion of imports. 20 M. Martin (1999), “Analysing the sustainability of development financing”, LSE Crefsa Quarterly Review, n. 1. 82 Table 6 shows the levels of some debt ratios that are considered “unsustainable” according to different institutions. The first criteria (World Bank) have been set by the World Bank in classifying severely indebted countries and based on the average ratios during 1980-87 for countries that fell into arrears or rescheduled during the 1980s. The World Bank and the IMF have set the second criteria (HIPC-1) in 1996 in defining Heavily indebted poor countries, an initiative that will be deeply described in chapters 7 and 8. The third criteria (HIPC-2) derive from the HIPC-1 and have been adjusted on the basis of negotiations between the G-7 and the IMF and the World Bank, assuming the potential cost of different levels of ratios and the availability of resources. The fourth criteria have been used in a study by External Finance for Africa undertaken in 1999 for the UK Department for International Development, covering the 1980s and 1990s and examining the ratios associated with occasions of default by low-income developing countries. In all these criteria, the averages on which the levels are based hide wide variations for individual countries, given different servicing problems, different changes in macroeconomic prospects and in domestic and international markets. Wide differences are related to the nature of the indebted country (low-income or middle-income country) and of debt (shortterm or long-term: when short-term is high, other indicators are more important, such as short-term external debt as a ratio of foreign reserves, or as a ratio of total external debt). Moreover, other institutions have set other levels of “sustainability”: one of the Maastricht convergence criteria set by the EU is the ratio of government debt no high than 60 percent of GDP. The UK government looks at the net position, assuming that net public debt must not exceed 40 percent of GDP. Recently, in the United States, both Congress and Administration have proposed budgets designed to reduce the deficit and reach a zero debt target by the years 2002 and 2015, respectively. Sustainability of debt and fiscal policies is one of the major and controversial issues in macroeconomics. Given these limitations to general rules, we have to consider another problem: the numerators are not very reliable. Present value is used to measure the concessionality degree of debt, considering the cost of future debt service payments against the alternative cost of borrowing. But the discount rates used to calculate the value are based on those used for OECD export credits, which do not represent a real alternative source of borrowing for poor countries. And the denominators are not very reliable as well. GDP is a vague and broad measure of national payment capacity, whereas export revenues are not very significant as the governments have scarce access to them in poor countries. Budget revenue and the present value of export seem to be more useful in the African country case. Usually and implicitly, all these ratios are referred to public and publicly guaranteed debt, without considering other types of debt (domestic debt and private sector debt). But the total level of government debt, total level of national external debt, total short-term debt are very important in order to better assess the sustainability of debt. The problem of domestic debt (Cameroon) or private sector external debt (Nigeria), respectively due to domestic overborrowing by government and to external over-borrowing by the private sector can introduce dramatic fiscal, balance of payments and currency problems as well as public external debt. In many African countries, domestic debt is not represented by treasury bills and bonds, but is due to the existence of payment arrears or liabilities contingent on the occurrence of some events (state guarantees of loans, insurance schemes on bank deposits, exchange rate guarantees, the assumption of debts of entities being privatised). Thus, external government debt service must be summed up to domestic debt service (including principal rollovers on it) and measured as a ratio of budget revenue to have a signal of risk of incurrence of arrears, default or rescheduling, because of the difficulties of fiscal adjustments in Africa. This 83 approach contradicts the so-called Lawson doctrine 21 , according to which current account deficits that result from a shift in private-sector behaviour (in the sense that they do not reflect government budget deficits) should not be a public policy concern. The Mexican crisis in December 1994 and the East Asian financial crisis in 1997 have shown that large and persistent current account deficit, basically financed through short-term capital inflows, can create a fertile environment for external crises. Obviously, it is not possible to define an optimal current account deficit in a normative sense: Norway, during the 1970s, run high current account deficits, but international capital markets did not press Norway, and no interest rates increase or krona depreciation occurred; Australia and Canada have experienced since the 1950s high current account imbalances, without any serious negative consequence. However, there are some indicators of potential difficulties 22 : (a) Inter-temporal solvency, that is the ability of the debtor country to repay in the very longrun its external debt, assuming that deficit of the current account represents the change in a country’s net foreign assets and that the present discounted value of future trade surpluses must equal the present value of foreign debt 23 . Expectations of the future stream of current account imbalances, future productivity growth, interest rates and access to foreign capital determines financial markets’ reaction, as international capital mobility opens the opportunity to trade off present levels of absorption against future absorption. The “equilibrium” response of the current accounts depends crucially on the expectation of whether the productivity surge is temporary or permanent (only a permanent productivity surge induces investment and a higher future capital stock, also causing consumption to rise more than output, lower savings, resulting in a strong current account deficit). (b) The path of sustainability, given by trade imbalances and debt dynamics. In this case, willingness to pay may be an important issue for external creditors. At this regard, it is not only important to know the sources of the current account deficit, but also the size and the time profile of the balancing adjustment. The currency composition and maturity profile of external debt can contribute as much to vulnerability to external shocks as does the total volume of debt24 . Considering an economy in steady-state, d are the liabilities as a fraction of the country’s GDP that foreigners are willing to hold in equilibrium (i.e. total external debt/GDP as an equilibrium portfolio share) and they are assumed to be constant in equilibrium and they are equal to the current account deficit (=CAD) plus the net accumulation of international reserves as a proportion of GDP (=FX), in proportion to long-run GDP growth (=γ). 21 In 1988, commenting the UK balance of payments situation in a speech to the IMF, the British Councillor Nigel Lawson said that, if in the past UK current account deficits were always associated with large budget deficits, low reserves and exiguous net overseas assets, the present situation is completely different. But, in a forward-looking rational-expectations framework, current account balances are always the result of privatesector decisions, with or without public-sector deficits. And current private-sector liabilities are often contingent public-sector liabilities: foreign creditors may force governments to turn private-sector debt into public-sector obligations, as happened in Latin American countries after 1982, and private-sector losses tend to be absorbed by the public sector, in terms of tax revenue. 22 A. North (1999), “Current Account Sustainability: Evidence from South Africa”, in LSE Crefsa Quarterly Review, n. 1. 23 G. Milesi-Ferretti and A. Razin (1998), “Current Account Reversals and Currency Crises: Empirical Regularities”, IMF Working Paper, No. 89, Washington D.C. 24 Financial markets’ concerns about Mexican risk were attributed primarily to the currency composition and maturity structure of the public debt rather than its size. See S. Griffith-Jones (1997), “Causes and lessons of the Mexican Peso crisis”, WIDER Working Paper, No. 132, Helsinki. 84 γd = CAD + ∆FX = net liabilities (3.1) Long-run GDP growth (=γ) indirectly impacts on debt dynamics through two other effects: (1) as the economy grows, the desired level of international reserves, which also depends on the level of imports, grows; thus, with η denoting real annual import growth: γd = CAD + {[(1 + η)/(1 + γ)] FX - FX} (3.2) (2) relative growth induces real exchange rate appreciation (=ε), which is driven by the evolution of productivity differentials between traded and non-traded goods in the domestic economy and in the rest of the world, that reduces both debt and foreign exchange reserves as a fraction of GDP. This is called the Balassa-Samuelson effect and it implies that also the ε factor should be taken in account, measuring (γ + ε)d. In any case, even if we use a system of structural equations, econometric models remain highly limited in their ability to predict a pending external crisis, as unforeseen shocks may show an external balance position to be unsustainable only ex-post. In sum, there is no simple rule to determine when a current account deficit is sustainable 25 . The East Asian crisis showed that an high exposure of domestic and international lenders and suppliers of non-debt capital flows can induce large government contingent liabilities for bailouts. Thus, the ratios of government liabilities to assets (considering the current stocks and the present values of associated flows compared to present values of future expenditures and revenues) must be seriously taken in account. There is always a risk of interaction between external and domestic financial crises, particularly when the domestic banking system holds high proportions of its liabilities or assets in foreign currency and foreign reserve cover is relatively low. Thus, foreign currency deposits to M2 ratios and M2 to reserves ratios can be useful indicators, too. Linked to the macroeconomic balances, it is important to monitor the non-interest current account deficit, the primary fiscal deficit, and the savings-investment gap, even though they are not directly related to the assessment of genuine demand for debt. But the current account balance reflects the difference between a country’s savings and investment, thus these levels have implications for the sustainability of the external position. Nonetheless, high savings and investment levels do not necessarily mean high output levels, as they may be allocated inefficiently. Under these circumstances, high levels of public investment do not automatically enhance external sustainability; its composition is very important. The same refusal of simplistic assumptions must be adopted referring to the importance of openness degree and trade issues. Propensity to import and export influences the sustainability of debt: countries with larger export sectors are able to better service external debt and implies large domestic constituency interested in good external relations (that is to avoid defaults on external debt commitments), but the more open an economy is, the more vulnerable it will be to external shocks. Exchange rate policy, fiscal and monetary policy have great importance in determining a country’s external competitiveness, but there is no definitive answer to the question of whether regime is the most adequate: it clearly depends on the relationship of the given economy within the global economic system, the structural characteristics of the economy; the susceptibility to external shocks. Caution is valid in analysing external factors, capital flows and the capital account regime as well. There is a critical relationship between the current account and the capital account, and 25 H. Reisen (1998), “Sustainable and Excessive Current Account Deficits”, OECD Development Centre Technical Paper, No. 132, Paris. 85 issues of flexibility and openness are very important to the sustainability of a current account deficit. But there is no regularity. Increased capital flexibility increases the country’s vulnerability to sudden capital reversal, but international investors perceive it as a more disciplined and responsible economy. On the other hand, the experience in Chile showed the importance of selective capital controls in order to mitigate the excessive volatility in capital flows. A deep analysis of debt sustainability should use also dynamic ratios, as all the ratios we have considered are static ratios, whereas debt problems are basically linked to the dynamic trend of payments. The growth of nominal debt, the effective interest rate (including all interest payments and other charges), present value compared to the growth of export revenues, budget revenue, domestic savings, GDP can represent the basic dynamic ratios. If debt growth is lower than revenue growth, then debt is becoming more sustainable. However, the same criticisms of static ratios must be referred to dynamic ratios using similar variables as numerators and denominators. Dynamic analysis represents an evolution of static approach, but it does not solve the practical problems in assessing the sustainability of debt. The models based on the savingsinvestment gap simply say that debt strategy can work if there is sufficient economic growth. Given that external financing requires foreign currency, as debt must be repaid in foreign currency, a more accurate analysis of external solvency considers also the relationship between export performance and the cost of borrowing. In fact, as the value of exports gives a more accurate measure of repayment capacity than GDP, the conclusion is that the rate of growth of exports must equal or exceed the rate of interest, in order to maintain debt service capacity. But, in reality it is wrong to assume a time-invariant growth path for income exports and the rate of interest, as all follow time paths. Given that also imports play an important macroeconomic role in the growth process, it should be more appropriate to include also the time path of imports. It leads to an extended debt dynamic model to define the conditions to maintain debt service capacity, where imports are treated as an endogenous variable. Solvency occurs if the growth rate of exports is higher than the growth rates of both the rate of interest and imports. When time is considered the analysis becomes more precise but also more complex and vague: the time paths of the considered variables are very difficult to predict. 7. High debt burden and debt overhang New lending and rescheduling of payments might only provide an indebted country with short-term relief. Resulting from the short-term relief the country might fall deeper into debt if the country's ability to service its debt does not improve in the future. In such situation it can easily be assumed that the debt servicing capacity of a country is dependent on the size of its debt. Thus the more new lending is granted to the country, the heavier its debt burden grows and the more difficulties it will have in keeping current with its external obligations on outstanding debt. When the difficulties grow too big to be handled, the country has only the options of either repudiating or requesting for a restructuring agreement. High debt burden can also be seen as an obstacle for growth. Empirical data shows that as a result of developing country debt crisis the net investments in those countries have suffered an extensive reduction. It seems that there is a two-way causality between the growth and the debt burden. Heavy debt burden can be seen to hinder capital formation and slow down economic growth through two channels: the illiquidity effect and disincentive effect. Illiquidity effect can be characterised by situation in which a heavily indebted country struggles with having to allocate the scarce resources between consumption, investment and external transfers to service outstanding debt. In such situations funds used for investments 86 are likely to reduce substantially because maintaining a certain level of consumption and keeping current in debt obligations easily exhaust most of the resources. Extensive cut-downs in funds used for consumption are politically very hard to make and thus short-sighted politics which cut down investments is often the easy way out for the country's leaders. The low level of investments is of course in the long run a severe obstacle for growth. An extreme situation of illiquidity is called a liquidity trap. Then ever larger and larger share of the country's output is spent in consumption, investments are driven down, and thereby also the future output. A country in a liquidity trap, due to external credit payments, which have not been serviced, is hardly able to receive financing for new, even productive, investments. This is because the new potential creditors realise that the senior debts will be serviced prior the later issued, and thus the senior creditors would pocket the profits from the productive investments. As the productive investments cannot be financed, the growth rate of the economy will further diminish. Debt overhang can develop as a result of a combination of the disincentive effect and the illiquidity effect or as a result of either one of the two effects. Debt overhang occurs when the effects are strong enough to discourage the country from fully exploiting its growth prospects unless some relief is granted by its creditors. In a country that suffers from debt overhang potentially productive investments are left unexploited. In such a situation also the country's creditors suffer as future income is sacrificed. Beyond a certain point, a high level of external debt acts as a marginal tax on investment because a fraction of the gains in output, resulting from increased capital formation, accrues to creditors in the form of debt repayment. High indebtedness can therefore lead to low investment, low growth, and ultimately to low repayment. Debt overhang has been divided to a weak version and a strong version. A country is said to suffer from a weak debt overhang when the debt burden is so large that simply issuing further financing for the country cannot solve the situation. In the case of a strong debt overhang, the debtor postpones the implementation of profitable investment projects until at least part of the debt is forgiven. The leaders of the debtor country have no incentives to participate in extensive structural adjustment program because the benefits of increased growth would end up to creditor’s pocket while the short-term cost would have fall solely on the debtor’s shoulders. The existence of a debt overhang can thus distort the incentives of a debtor. Krugman (1989) 26 defines debt overhang in terms of resource transfers. A country suffers from debt overhang if the expected present value of potential future resource transfers is less than its debt. The key is that the potential repayment of the country is not independent of its debt burden. When a country's obligation exceeds the amount it is likely to pay, these obligations act as a high marginal tax rate on the country because if the country succeeds in doing well, most of the benefits will go to the foreign creditors. This, of course, discourages the country from doing well. The government borrower is less likely to implement painful corrective policies if the benefits of improved economic performance go to foreign creditors. In addition, the burden of national debt will fall on domestic residents through taxation (in particular taxation on capital); therefore, debt overhang deters investment. Krugman postulates a relationship between the face value of the debt and the market value of the expected repayment. At low levels of external debt, creditors expect that the debt will be repaid in full. The secondary market price (which is equal to the market value of expected repayment divided by the face value of debt, ignoring risk and transaction costs) will be one and the value of the debt will lie on the 45-degree line. At higher levels of debt, the 26 P. Krugman (1989), “Market-based debt reduction schemes”, in J. Frenkel et al. (eds.), Analytical issues in debt, IMF, Washington D.C. 87 probability of non-repayment increases as the country has less incentive to invest; hence, the expected payments are below the 45-degree line. From a given point onwards, any level of debt is associated with lower secondary market price and at first, the market value of expected repayment would still rise. At very high levels of debt the disincentive effects of doing well are so large, that the curve turns down and the market value of the debt starts to fall when the face value of debt increases. 8. The solvency-liquidity problem of External Debt and need of debt relief Different features of sovereign lending compared to domestic lending, especially because it involves a sovereign state as a debtor cause the complexity of breakdowns in international debt contracts. In domestic debt the creditors are able to demand collateral to secure their outstanding claims, and the collateral is quite easily enforceable in the event of repudiation. While in sovereign lending, where the amount of principal is often much larger than in domestic lending, the collateral is often insignificant, and, in addition, the creditors have very limited instruments for attaining the possession of debtor’s assets due to lack of enforcement mechanism. In any case, if a sovereign debtor repudiated, it would more often be a case of unwillingness to pay than inability to pay. Another main difference between the domestic and sovereign lending is that the domestic bankruptcy negotiations are mostly held only once, when sovereign debtors renegotiate with their creditors time after time. The continuous nature is caused by the creditors' inability to make the sovereign debtor commit into a legally binding agreement. From the sovereign debtors’ point of view, bankruptcy implies further problems, as no new loans would be extended to sovereign states, thus the sovereign debtors must have some incentives to repay. Moreover, as international debt is a system of facilitating foreign trade, the more open the country’s economy is the more dependent it is on external credits. Therefore it is in the interest of an open sovereign state to build up a good reputation as a trustworthy debtor and thereby sustain the ability to enjoy the services. A rationally reasoning sovereign debtor will service its debts or enter into restructuring negotiations only if it is less costly than repudiating. Renegotiations of debt contracts as a preferred way of dealing with debt-service difficulties is a post World War II phenomenon. Prior to this, defaults on sovereign debt were common occurrence. In the past the difficulty in attaining the control of a debtor country’s resources to the creditor was often dealt with so-called “gunboat diplomacy”, that is with a use of armed forces. The main reason for the change in a way of dealing with the issue is that in the pre1930 period lenders in the international financial markets were mainly individual bondholders who did not co-operate with each other, while after the World War II, the lenders have rather been governments and banks who have been able to reach co-operative agreements through informal clubs, such as the Paris and London ones. Also the establishment of the International Monetary Fund (IMF) and the World Bank has influenced the international financial markets towards more co-operative direction. Another influential matter was a change in the relative bargaining strengths between the creditors and the debtors. Earlier the creditors were represented by private bond-holder committees which had limited option for retaliation and thus little power to force the debtor to service its debt. Now the creditor governments are able to intervene by imposing conditionalities, in order to influence the economic development and the structural adjustment programs of debtor countries. The fact that most credits granted from unofficial sources are syndicated loans, in which cross-default clauses are included, has increased the bargaining power of the creditors. In a situation when a country is unable to meet its debt obligations out of present income, the creditors have two options. They can either reschedule the payment stream or finance the 88 country in the hope that it will repay its debt or they can forgive part of the debt and reduce the debt to a level that the country can repay. The choice between the alternatives is a trade off. Refinancing and rescheduling give the creditor an option value: if the country turns out to do well in the future the creditors will not have written down their claims unnecessarily. On the other hand, heavy debt burden can distort the country’s incentives for development and growth and thus be detrimental to the creditors. Quite often, during the explosion of foreign debt crises, such as in the 1982-1987, a basic distinction in made: debt crisis can be due to liquidity rather than solvency problems. Countries can be seen as basically solvent, if they are able to repay (in terms of present value of future output flows), but illiquid, that is lacking at present the cash to service its debt. Paul Krugman expressed the fallacy of such an argument to explain debt crises: “If it is known to be solvent, a country can find voluntary lending, and there is no liquidity problem. The liquidity problem arises precisely because there is a possibility that the country will not be able fully to repay its debt” 27 . Agenor and Montiel28 pointed out two factors, which can be useful to understand the solvency-liquidity problem: the distinction between the ability and willingness to pay and the difference between the concept of solvency of a country at aggregated level and of the government of the country. As a consequence of this solvency problem, external debt can be sold at a discounted value on the secondary market that values debt below its face value. Hence, each kind of debt issued by that country will be discounted on the secondary market, on a par with existing debt, and new lenders would not enter in the market and no credit will be available for the country. Public sector is perceived to be solvent if the present value of its future resources available for debt service, currently discounted, is at least equal to the face value of its initial stock of debt. TDE/GDP = pd/GDP – sr/GDP + (ri – n)TDE/GDP (3.3) Where: TDE/GDP = total (domestic plus foreign) public debt over GDP Pd/GDP = primary deficit over GDP Sr/GDP = seignorage revenue over GDP ri = real rate of interest n = rate of growth of real GDP If (ri - n) < 0, then the government could do an “honest Ponzi game” 29 , i.e. it could borrow new credit for debt service since the outstanding debt contributes to the grow of total debt at a decreasing rate. If (ri - n) > 0, then it infers a debt growth without bound and it requires that the public sector is able to service the debt using their own resources, i.e. running sufficiently large primary surpluses and with seignorage revenues in order to limit debt spiral. At this regard, evidence suggests that in Sub-Saharan Africa voluntary lending from the private sector is very limited, the level of seignorage revenue is quite low and that money 27 Page 290, P. Krugman (1989), ibid. 28 P. R. Agenor and P. Montiel (1995), Development Macroeconomics, Princeton University Press, Princeton. 29 The debt burden can be assessed by evaluating external solvency, in a manner analogous to the assessment of fiscal solvency: the no-Ponzi game condition prevents a country from rolling over its external obligations indefinitely. Under this condition, external solvency requires the present discounted value of trade surpluses to be sufficiently large to repay the existing net stock of foreign liabilities. 89 printing is likely to increase inflationary pressure quickly and to evolve in a inflation-deficit vicious circle. TDE/GDP TDE/GDP ri - n < 0 ri - n > 0 t t As suggested by Agenor and Montiel, external debt crisis may arise as an effect of the perception of lenders that public sector of debtor countries are insolvent ex ante, i.e. the fiscal adjustment necessary to restore the solvency constraint are not forthcoming. In this case, the amount of debt is sold at a discounted value equal to the prospective debt service and secondary market values the debt below its face value. In presence of discounted debt on the secondary markets, the ability of collecting new credit on the market is reduced. Given these premises, the idea that a reduction in debt burden could be a Pareto improvement and therefore beneficial for both debtors and lenders has its theoretical justification in the fact that debt overhang would act as a marginal tax on the future output of a country. On one side, the government will have less incentive to implement policies to improve economic performance, especially if these are painful for the country, as it realises that the future benefits of such policies, in terms of increasing output, will mostly go to creditors. On the other side, private investors forecast increases in taxation of domestic income to repay debt and this acts as a powerful deterrent to investments. There are situations in which reduction of debt would be also in the interest of the creditor. For example, if the market value of the outstanding debt is increased by partial debt forgiveness then the creditor can benefit from the restructuring provided that the value of the claims is increased by more than what the restructuring has cost the creditors. The optimality of debt relief can be examined in a debt-relief Laffer curve Value -framework 30 . C L R D Debt 30 From this figure, we consider the present value of a country’s debt obligations on the horizontal axis, and the expected present value of its future debt service on the vertical axis. At a low level of Debt, expected repayments increase one for one with the contractual value of debt (i.e. P. Krugman (1988), “Financing versus Forgiving a Debt Overhang: Some Analytical Notes”, Journal of Development Economics, N. 29, December; E. Helpman (1989), “Voluntary Debt Reduction: Incentives and Welfare”, IMF Staff Papers, N. 36, September; J. Sachs (1989), “The Debt Overhang of Developing Countries”, R. Findlay, G. Calvo, P. J. Kouri and J. B. de Macedo (eds.), Debt, Stabilisation and Development, Basil Blackwell, Oxford. 90 there is non-risk of default). Beyond point R (the inefficient portion of the debt Laffer curve, with potentially sizeable losses) it is more efficient for the creditor to reduce its claims on the country and pass on at least part of the secondary market discount as the overall value of its debt will increase. This is because the loss of the option on the future output of the country is more than offset by the effect of the reduction of debt overhang on the incentives for the country and therefore on perspective payments to the creditors. A high debt burden can be a potent disincentive to investment, especially in the presence of uncertainty about future output (which captures the likelihood of being able to service debts in the next period). Debt relief, a form of aid, can then encourage higher levels of investment: within this framework, in certain situations partial debt forgiveness can be a Pareto improving solution. In fact, debt-relief Laffer curve suggests that the debt relief is beneficial to the creditors if the debtor lies on the falling part of the curve, which is often called the wrong side of the debtrelief Laffer curve. But in practice there are difficulties in ascertaining on which side of the curve a debtor country is: a useful technique is the comparison between the secondary market price of debt and the face value of debt claims, but unfortunately for poor countries there is no secondary debt market31 . Despite of these difficulties, the framework has been seen as an useful way of organising thinking and arguments when dealing with issue of developing country debt crisis. Infinitely large debt is comparable to a high marginal tax on efforts to expand debtor's foreign exchange earnings. Thus like a tax (that’s why the curve is comparable with the Laffer curve used in taxation models), the high debt burden has disincentive impact on domestic investment. As government may sometimes increase tax revenue by decreasing tax rates, creditors may increase expected payments by forgiving part of a country's debt. Therefore, debt reduction might be in the interest of both borrower and creditor if the borrower is on the wrong side of the Laffer curve. This can be illustrated by examining a situation when the currency income of the country increases with growth of the economy. The country being heavily indebted, a great deal of the increased currency income goes to servicing the earlier accumulated debt and does not rise the welfare of the country's citizens. Therefore political decision-makers of the indebted country will have little incentives to allocate funds to investments as the fruits generated by the investments would be pocketed by international creditors. They will much rather spend more on current consumption and thereby try to ensure their tenure in power. This effect of debt burden is what we called the disincentive effect of debt overhang. It arises from expectations that future debt burden reduce the incentives for current investment and adjustment. It can be argued that large external debt is a bottleneck to growth and development. If this is the case, then by reducing the debt of heavily indebted countries, the expected amount of repayments would eventually be increased. If the creditors committed themselves to absorbing a smaller part of the country’s output, the debtor country would be more likely to pursue the types of investment-orientated policies that lead to higher future levels of output, thereby increasing the total resources available to both parties. Sachs (1989) argues that most of these economic inefficiencies that hamper economic growth could be overcome by partial debt forgiveness so as to improve the position of both debtors and creditors. That is, debt forgiveness could in fact increase growth in the borrower country and increase the eventual repayments received by the creditors. This argument shows that debt relief initiative can have economic motivations, without implying any moral or solidarity priority over selfish. 31 Because of the relatively small proportion of debt owed to commercial creditors by these countries. 91 As an alternative to debt forgiveness to deal with the debt overhang problem, Paul Krugman32 observed that under adverse circumstances a “defensive lending” argument can justify quite substantial increases in creditor exposure. New lending that reduces the interest burden, even if it is at a loss (i.e. the expected loss on the new lending), may be worthwhile because it has a benefit, improving the expected value of the initial debt (i.e. the increase in the value of existing claims). As for debt relief, the free-rider problem can create a difficulty with defensive lending: it is to the advantage of any single creditor not to lend more and simply gain from the collective claims on previous debt; negotiating procedures and institutional arrangements are consequently required. Non-participation creates a “prisoner’s dilemma” in which it may well be to the collective advantage of all the creditors to forgive debt or to lend defensive funds but no individual creditor has the incentive to do so. Defensive forgiveness and defensive lending confront operational difficulties: how much debt should be forgiven or how much to lend? Forgiven by whom, within the map of all the lenders? At what level does outstanding debt become sustainable? A general rule may be that it is quite difficult to imagine the existence of a general rule, which is valid in presence of all the differences among debtors and lenders. 9. Free riding and moral hazard problems with restructuring debt Thus, a central problem with restructuring of debt is an accomplishment of co-ordination among the numerous creditors of a single debtor: this is a typical phenomenon that is linked to the problem of debt overhang and sustainability analysis. Even though debt reduction can be beneficial to creditors collectively, every single creditor has an incentive to withdraw from granting debt relief to a named debtor. This is because with other creditors cutting down the face value of the debt owed to them, the debtor’s ability to service the remaining debt increases. Fruits of this can be enjoyed also by the creditor not participating in the restructuring. Then, if one can ride free, why would another one pay for the same benefit? As a result it might be that nobody will participate in debt restructuring and agreements are simply not concluded. To provide the creditors with incentives to participate, equal treatment among them has to be ensured. In reality the free rider problem is partially eliminated with existence of the creditor clubs and the rules they implement to the restructuring agreements negotiated under the auspices of the clubs. When analysing the debtor’s incentives for debt restructuring, the related moral hazard problems cannot be ignored. The first moral hazard problem arises form a situation when a debtor country is granted debt relief. Then other countries struggling with debt servicing difficulties will have little incentives to keep current with their repayments but rather default and request for restructuring. All debts will not, however, be forgiven nor restructured. Often the decisions on which countries are eligible for debt restructuring are made on political bases. Of course the economic situation in the country and its past performance are also matters of concern. But, we can’t say that economic motivations legitimate a given choice by themselves. The complexity of economic systems of variables and determinants obliges us to identify a priority, which will orient our political strategy. The real priority can be the objective (we want to cancel all debt of the poorest countries) or the financial constraint (we have a limited amount of money at disposal) or a compromise between them. In any case, basically there is a political decision, which can be only supported by some “scientific” 32 P. Krugman (1988), “Private Capital Flows to Problem Debtors”, J. Sachs (ed.), Developing Country Debt and Economic Performance, vol. 1. 92 motivations. To think that this causal relation can be completely reverted is an illusion. Another moral hazard problem arises from the existence of debt relief system. Heavily indebted country that has been granted some relief has little incentives to mend its behaviour and utilise the whole potential of the economy if the decision-makers in the country can count on the availability of debt relief also in the future. To reduce the effect of moral hazard problems that arise from debt restructuring programs the relief is often made as una tantum and conditional on both past and future performance. The standard approach adopted in the debt renegotiations is a multilateral one. This is sensible because there are typically several creditors who have extended loans to the same debtor. The multilateral approach is able to ensure the equal treatment of all creditors involved. It also allows the restructuring process to be conducted in timely manner. Variety in the outcomes of restructuring negotiations suggests that there are major distinctions among the issues, interests and instruments which different debtor countries can bring to the bargaining table. It has been suggested that the difference in the bargaining outcomes of different countries can be explained with three factors: the size of the country's debt, its strategic and political significance, and its access to non-conditional resources. Large debtor countries such as Brazil and Mexico have been able to negotiate favourable deals with their international creditors. They have received lower interest rates, longer maturities, larger principal reductions and longer grace periods in their debt restructuring than countries with relatively smaller debt burden. This has happened partly because the large debtors have been in a position to threaten the international financial system as a whole. Thus they have been given more attention and have been taken more seriously as a result. Large debtor countries are often large economies as well, and hence capable of sustaining autarkic policy measures for longer time periods than small countries. Therefore the credibility of their threats to resist required policy reform is greater than the credibility of smaller countries. SSA countries are small debtors, in absolute terms, and they represent the easiest target of debt relief. To cancel their debt is not to threaten the international financial system. It is not expensive, it is manageable and it does not involve a lot of private lenders. And it can demonstrate the effectiveness of global governance guided by ethical principles and common responsibilities. Even more important than the size of the debtor country, might be the strategic significance of it. Due to strategic and political concerns creditor governments have used their influence in the boards of the IMF and the World Bank to press for greater leniency and to lobby bank advisory groups for expeditious settlement of restructuring negotiations. As an evidence of strategic importance it can be pointed out that the United States have been relatively more concerned with relief for Latin-American countries, the Germans for Eastern Europe, the French for their former colonies in Africa, and the British for indebted members of the Commonwealth. Latest implication of the strategic significance is the generous debt relief provided to Pakistan following its involvement in the allied coalition against the Taliban regime in Afghanistan. The access to non-conditional financial resources such revenues from scarce natural resources or proceeds from commodity price booms influence the debtor country's bargaining position and the financial terms it is likely to receive. The presence of additional resources can render a country less vulnerable to external pressure because it will probably have fewer political incentives to make difficult economic adjustments which a restructuring agreement or timely repayments would require. The credibility of a debtor country’s bargaining position is likely to increase if its creditors realise that it has access to non-conditional financial resources. 93 Also a number of different domestic factors can influence the outcome of the international financial negotiations. The political climate in the country can influence the leaders’ willingness the settle a deal with the creditors - if the length of their tenure in power is endangered, they are not likely to make hard decisions on restructuring of debt but rather ignore the external obligations of the country. On the other hand the economic situation determines the country's ability to withstand sanctions. The particular nature and condition(alitie)s of each debt re-negotiation directly affects the relationship, which is established, in the long-term, between debt and development. 10. An additional problem. The operationalisation of debt sustainability There is a clear gap between the conceptualisation of debt sustainability in theoretical literature and its operational guidance for designing adequate monitoring policies. That’s why, in the empirical literature, debt capacity is seen from an ex-post perspective, that is verifying tangible problems (failure to service external debt, accumulation of payments arrears, debt rescheduling), rather than from an ex-ante definition. Inter-temporal dynamic analysis must be extended to other crucial variables, which have been mentioned in the context of the static description. In fact, the use of external debt goes beyond the possibility to finance investment and imports, as it can also be used to maintain high level of consumption. Moreover, a complete analysis of debt sustainability should take in account lender countries as well. A debt situation may be consistent with the inter-temporal budget constraint of the borrower, but it may become unsustainable if the supply conditions change. Borrowers’ behaviour affects the behaviour of lenders and vice-versa. The decision-making and behaviour of lenders and of the institutional framework of international finance are very relevant to determine the sustainability of debt. Particularly when commercial lenders are involved, the perspective of creditors, using the concepts of creditworthiness and country risk (as well as a measure of the borrowers access to international capital market), should be an additional component to the solvency issue. In the credit-rationing context, directly linked to commercial lenders, a measure of the secondary debt market values can be used as a proxy of past difficulties or anticipation of future problems (a para-dynamic approach). The sustainability itself should be referred to both the borrowing and lender countries. What may be sustainable from the borrower’s point of view, it may result unsustainable from the lender point of view. To define the financial cost of any given debt relief initiative is a legitimate request from donor’s side. Given the financial constraint of the initiative, that is the amount of pledged resources, we can consider an initiative as “sustainable” only when its implementation does not incur in extra-costs. This is a correct way of defining the sustainable nature of a financial initiative, from the lender point of view. Obviously, taking account of such considerations makes the empirical application of theory much more difficult. Not only debtors’ situation interferes with sustainability, but lenders’ too. And we have to admit it. Unfortunately, rhetoric pushes international agencies to present to public opinion the sustainability as uniquely referred to debtors’ situation33 . Wherever possible, on the basis of the results from debt sustainability analysis, the idea should be to reduce reliance on debt to sustainable levels of the static ratios, with dynamic ratios moving in the correct direction. 33 Public opinion thinks that sustainable level means a level of debt, which can be managed by debtor country without negatively affecting development process. Donor countries prefer to hide the important lender component of sustainability, in order to demonstrate their high degree of responsibility towards the poor. 94 At this regard, a simple model of debt sustainability, can be derived from Jeffrey Sachs 34 . Given that: DSt = IPt + PRt (3.4) NBt = NDLt + INPt (3.5) Bt = NBt – PRt = Dt+1 - Dt = ∆D (3.6) NTt = Bt – IPt = NBt - DSt (3.7) Where: Dt = disbursed debt IPt = interest payments PRt = principal repayment DS = debt service NBt = flow of new debt NDLt = new disbursed loans INPt = past interests to be paid NTt = net transfer rDt = interests on current stock ∆D = variation of nominal debt stock And assuming the simplest economy, which produces only one tradeable good (=Q), there is a typical inter-temporal optimisation problem to be solved. What is the needed amount of debt in order to maximise the sum of utility/preferences of population? First, there is a national budget constraint: disposable resources must be enough to finance consumption and to pay interests on past debt: Qt + Bt = rDt-1 + Ct (3.8) Ct = Qt + Bt – rDt-1 = Qt + NTt (3.9) Second, if we assume the rationality of lenders constraint, which implies that borrowers can not continue receiving indefinitely new lending to repay past debt, then we have to take in account the solvency constraint. The value of what is produced (i.e. the stock of financial and real assets = Wt ), measured by income growth rate, can not be –at least in the long run - lower than the amount of debt: ∞ NTt = ∑ t t =0 (1 + r ) ∞ Ct − Qt ∑ (1 + r ) t =1 t ≤0 (3.10) Dt ≤ Wt (3.11) This implies that discounted present value of debt can not be permanently positive: Limt→∞ Dt /(1+r)t = 0 (3.12) Resource transfer from borrower to lender must be considered normal, if lender wants to realise a non-negative profit. If resource transfer from borrower to lender is never registered, then debt stock increases at least as much as interest rates, thus violating (3.12). Some points to be mentioned can be derived from these conclusions. Debt makes it possible to separate consumption growth from income growth. Thus, a country, which is serving the interests of people who do not want to postpone their consumption, can use debt. If a country has permanently high rate of growth, it can choose to maintain high levels of consumption, accepting debt in the short-term and paying it back in the long run, 34 Based on J. D. Sachs (1981), “The current account and macroeconomic adjustment in the 1970s”, Brookings Papers on Economic Activity, Vol.12, No. 1. 95 with increased income. A country can have sustainable present debt, in order to finance tomorrow investment, which will produce permanent increase of income growth rate. In terms of sustainability, it is important, even if quite difficult, to consider the political component too. Alesina and Tabellini 35 analysed political determinants of indebtedness. Berg and Sachs 36 tested the link between debt and political components, using Probit analysis on 24 middle-income countries and they showed that the level of inequality of income distribution is much more important to predict the probability of refinancing debt, rather than any other economic variables. In fact, due to inequality, there is more political instability and social unrest; governments are more exposed to strong pressure and budget discipline is less important. In these conditions, time horizon is quite short for both government and population, and indebtedness becomes a useful tool to keep high popular consensus. SSA continent seems to be an interesting area where these assumptions can be tested. The nature of sovereign debt implies the presence of specific sovereign risk. In case of insolvency, in fact it is unlikely that lenders can confiscate any real guarantee provided by the government. Nevertheless, contract enforcement can be provided by the presence of other nations or international organisations, which can use commercial sanctions, conditionalities, and the importance of reputation within international financial markets. 11. Debt sustainability in developed countries Given longer-term sustainability considerations and short-term aggregate demand conditions, the International Financial Institutions, and particularly the IMF, recommend appropriate fiscal policies. Tab. 7 – General Government Fiscal Indicators, at the end of 1998 (in percent of GDP) Canada France Germany Italy Gross debt 95.8 58.2 61.1 118.7 Net debt including social security assets 62.3 48.4 52.4 112.4 Net debt excluding social security assets 62.3 48.4 52.4 112.4 Overall balance including social security 0.9 -2.7 -2.0 -2.7 Overall balance excluding social security 2.9 -2.6 -2.3 1.3 Structural balance including social security 1.6 -1.3 -0.7 -1.5 Source: IMF, 1999 Japan 117.9 30.5 79.1 -5.3 -7.5 -3.8 UK 62.2 42.4 42.4 0.3 .. -0.3 In the previous table, estimates of debt indicators that are used to assess debt sustainability in formulating fiscal policy advice are reported for the G-7 countries. The first row refers to Gross debt, that is the government’s stock of outstanding financial liabilities. Italy and Japan are the most indebted G-7 countries, with gross debt of 119% and 118% of GDP respectively. The second row subtracts all government financial assets from gross debt, including assets of the social security system37 , often in the form of government debt. Japan is the only country 35 A. Alesina and G. Tabellini (1989), “External debt, capital flight and political risk”, Journal of International Economics 27 and A. Alesina and G. Tabellini (1990), “A positive theory of budgets deficits and government debt”, Review of Economic Studies 57. 36 A. Berg and J. D. Sachs (1988), “The Debt Crisis: Structural Explanations of Country Performance”, Journal of Development Economics, Vol. 29 (November). 37 This indicator does not include the future claims of the social security system against government assets, thus it may provide a very partial measure of the government financial situation. Chand and Jaeger (1996) add net future public pension liabilities – equal to the present value of future pension liabilities less the assets of the 96 USA 62.1 48.4 56.1 1.3 -1.7 1.3 where the latter is sizeable: when social security assets are included in government debt, Japan is the least indebted country; whereas Italy is always the highest (because the social security system holds no assets). The third row consider the social security system as independent, thus the debt held by government is not considered part of the government asset. In this case Italy is the most indebted country, followed by Japan. The forth, fifth and sixth rows indicate that the G-7 countries start with different fiscal balances The Japanese debt Japan is an example of huge public-sector debt. By the end of 2000, the Japanese government will be the biggest debtor (as a percentage of GDP) that has ever been among developed countries in peacetime 38 . The Japanese government is the champ in the public-sector debt-to-GDP ratio. This main characteristic has recent origin. Given the 1980s difficulties due to excessive private-sector borrowing, in the 1990s the government tried to re-start growth by spending and borrowing massively. It did mean a stimulus through an aggressive monetary expansion, after a long deflation and recession period. Public debt has increased rapidly as a result of the large deficits incurred during the current downturns and financial market concerns about the sustainability of this trend is shown by upward pressure on long-term interest rates 39 . As a result, in 2000 Japan’s general government deficit has widened to more than 8% of GDP; gross public debt amounts to 128% of GDP (compared to 69% in 1990) 40 . During past forty years non-industrialised country reached a public-sector debt higher than 135% of GDP, even though some developed countries such as Italy, Belgium and Ireland have been very deeply indebted public sector. The IMF forecasts that, by 2004, even though he government will reduce its budget deficit to 1.4% of GDP, debt will have increased to 150% of GDP, making Japan by a considerable margin the most indebted G-7 country. Net debt also increases. If we consider other potential liabilities, such as the government’s guarantees on bank loans, and the present value of future state-pension commitments (condemned to increase rapidly because of the ageing of Japanese population), then Japan public debt reaches 400%. Among developed countries, comparable levels of public debt were registered in United Kingdom at the end of the second war, when the ratio of public debt to GDP reached 250%. However, the Japanese situation is not very worrisome, particularly in comparison to SSA countries. Why? 1. Above all, Japan is the biggest creditor nation in the world: two decades of currentaccounts surpluses have produced $1.2 trillion of overseas assets, equal to 31% of public pension system – to net debt excluding social security to provide an indicator of net debt including future public pension liabilities. Their estimates, based on the 1995 juridical framework and on projections up to 2020, demonstrate that Italy and Japan would remain the most indebted G-7 countries. See S. K. Chand and A. Jaeger (1996), Aging Populations and Public Pension Schemes, IMF Occasional Paper No. 146, December. 38 The Economist (2000), “A tale of two debtors”, in The Economist, January 22nd 2000, p. 15. 39 IMF (1999), World Economic Outlook. October 1999, Washington D.C., p. 11-13 40 IMF (2000). 97 GDP. This is a very important characteristic, also compared to other developed countries having current-account deficits. 2. The Japanese are the biggest savers in the world. Thus the government can finance its deficit, through national investors rather than foreign ones. Foreigners hold only 10% of Japanese government securities, whereas this ratio is equal to 40% in the US and 23% in Italy. In terms of financial behaviour, we can assume that residents’ flows are less volatile than foreigners’ flows and they are not likely to give bonds up suddenly. 3. Moreover, half of all the outstanding government securities are held in the socialsecurity funds, owned by the government itself. 4. Net public-sector debt is only 40% of GDP, the lowest among developed countries. Nevertheless, a permanent growth in Japan’s debt is not sustainable in the long run: the rapid aging of Japanese population is pushing up social expenditures (health and pension costs), thus exacerbating the problem of deficit spending. However, it could be possible to increase revenues, broadening its tax base (as suggested by OECD) or raising tax rates, as Japan has the lowest tax revenues among developed countries. It is equal to 29% of national GDP, compared to 39% of other rich countries (and the European Union tax burden at 42% of GDP). The OECD also suggested to cut spending on public works, but reducing public investment, in order to reduce national budget deficit, while the economy is still fragile risks to generate recession, as the 1997 policy demonstrated. The Japanese case is instance where long-term and short-term aspects point to conflicting advice: short-term fiscal stimulus can effectively attack the weakness in aggregate demand, whereas the long-term unsustainability suggests that Japan need to return to fiscal consolidation. US Debt Compared with Japan, The US government looks like a paragon of fiscal virtue. After heavy borrowing in the 1980s, its budget has moved into surplus, allowing the government to repay some debt. When Bill Clinton moved into the White House in 1992, the US economy was starting to gather strength but the federal budget deficit was only getting bigger, leaving the new president with very limited room when it came to fiscal policy. His successor, George W. Bush, has inherited a very different situation. He has become president at a time when the economy is weakening, but the budget surplus seems to increase with each new calculation. As a result, Mr. Bush has more flexibility when it comes to the budget than any president in a generation. Mr. Clinton had to spend the first year of his first term pushing tax increases and spending cuts through a reluctant Congress to cut the deficit. Mr. Bush has a long menu of different choices made possible by the surplus: he can cut taxes, he can pay off the national debt, he can shore up Social Security and Medicare, he can provide health insurance to people who cannot afford it. Through most of Mr. Clinton’s two terms, fiscal policy has become centred on the long run: paying down the national debt, freeing capital for investment in the private sector and strengthening the federal balance sheet to deal with the costs of an ageing population. Mr Bush can decide how to pay off the national debt, or at least the $3.4 trillion of it held by the public in the form of Treasury securities (on top of the public debt, the government owes Social Security more than $2 trillion). And yet, even as the US government has weaned itself from its old borrowing habits, firms and US households have been borrowing still more, lifting their combined debts to a record 132% of GDP (compared to 80% in 1960, 93% in 1970, 102% in 1980, and 126% in 1990). The total debts of non-financial firms increased due to different reasons. First, to finance a high-tech investment boom. Second, to finance share buy-backs: during the past two years, non-financial corporations increased their debts by $900 billion, while they retired a net $460 98 billion of equity, in order to pay employees in share options without depressing the share price. Rising share prices have made households feel wealthier, and so encouraged them to borrow still more: total household debt has passed from 85% of personal income (1992) to 103% (1999). This increase in private debt may imply some elements to be worry about. Measures of corporate leverage, such as the ratio of debt to companies’ net worth (the value of tangible assets minus debt) have increased. Moreover, debt is fixed in value, whereas the value of assets, such as shares and property is not. Debts can only be serviced from income; assets can pay the interest bill only if they are sold, and if lots of debtors are forced to sell at the same time, asset prices will fall. A better measure of the debt burden therefore is debtservice payments as a percentage of disposable income. Despite low interest rates, households’ debt-service ratio is currently close to a record level (more than 13%). US’ private sector debt problem may not yet be as serious as Japan’s in the late 1980s, when banks engaged in imprudent lending on a massive scale. But US are vulnerable in a way that Japan never was, and it is more similar to the SSA countries. US are the world’s biggest foreign debtors, with net foreign liabilities of $1.5 trillion, around 20% of GDP. As a credit-fuelled spending devoted to more imports, US current account deficit has widened to about 4% of GDP. If it remained at this level, US net foreign debt would rise to more than 50% of GDP within ten years. This leaves the US economy dependent on foreigners’ willingness to hold more and more dollar-denominated assets. If foreigners’ interest in dollars falls, the currency will fall. US would then need to offer progressively higher interest rates to convince foreigners to put a growing share of their wealth into dollar securities. Higher interest rates, and hence lower share prices. Low inflation has encouraged heavy borrowing: the apparent defeat of inflation, combined with the expansion of US economy, has created the expectation that the boom will last forever. This, in turn, has encouraged firms and households to borrow more and lenders to relax their standards. And low inflation also makes excessive debt more dangerous: borrowers can no longer rely on inflation to erode their real debt burden if things go wrong. Given the differences of debt profile between USA and Japan, there is an important aspect, a part from the international reputation and usage, which makes these cases different from the SSA ones. Both the US and Japanese debts are almost all in their own currencies, unlike those of developing countries. And it does matter, too. 99 Bibliography P. R. Agenor and P. Montiel (1995), Development Macroeconomics ,Princeton University Press, Princeton. A. Alesina and G. Tabellini (1990), “A positive theory of budgets deficits and government debt”, Review of Economic Studies 57. A. Alesina and G. Tabellini (1989), “External debt, capital flight and political risk”, Journal of International Economics 27 A. B. Atkinson (1989), Poverty and Social Security, Harvester Wheatsheaf, New York. A. Berg and J. D. Sachs (1988), “The Debt Crisis: Structural Explanations of Country Performance”, Journal of Development Economics, Vol. 29 (November). N. Bhinda, J. Leape, M. Martin and S. Griffith-Jones (1999), Private capital flows to Africa: Perception and Reality, Forum on Debt and Development, The Hague. BIS, IMF, OECD, and World Bank (1988), External Debt: Definition, Statistical Coverage and Methodology, Washington D.C. C. Chang, E. Fernandez-Arias, L. Serven (1998), Measuring Aid Flows: a New Approach, Washington D.C. DAC (1996), Shaping the 21st Century: The Contribution of Development Cooperation, OECD, Paris. D. Gordon and P. Spicker (1999), The International Glossary of Poverty, Zed Books, London. A. J. M. Hagenaars (1986), The Perceptions of Poverty, Elsevier, Amsterdam. P. Krugman (1989), “Market-based debt reduction schemes”, in J. Frenkel et al. (eds.), Analytical issues in debt, IMF, Washington D.C. IMF (2000), World Economic Outlook, Washington D.C., October. IMF (1993), IMF's Balance of Payments Manual, Washington D.C. J. D. Sachs (1981), “The current account and macroeconomic adjustment in the 1970s”, Brookings Papers on Economic Activity, Vol.12, No. 1. A. Sen (1979), “Issues in the Measurement of Poverty”, Scandinavian Journal of Economics, n. 81. A. Sen (1981), Poverty and Famines: An Essay on Entitlement and Deprivation, Clarendon Press, Oxford. M. Simonsen (1985), “The Developing Country Debt Problem”, in G. Smith and J. Cuddington (eds.), International Debt and the Developing Countries, The World Bank, Washington D.C. L. Spaventa (1987), “The Growth of Public Debt”, IMF Staff Papers, Vol.34, N. 2, June. TFFS (2001), Draft of the External Debt Statistics: Guide for Compilers and Users, Washington D.C. UNCTAD (2000), The Least Developed Countries. 2000 Report, Geneva. UNDP (2000), Human Development Report 2000, Oxford University Press, New York. UNECA (2000), Economic Report on Africa 1999: The Challenges of Poverty Reduction and Sustainablity, Addis Ababa. 100 H. W. Watts (1968), “An Economic Definition of Poverty”, in D. P. Moynihan (ed.), On Understanding Poverty, Basic Books, New York. World Bank (2001), African Development Indicators 2001, Oxford University Press, Washington DC. World Bank (2001/b), Global Development Finance, Washington D.C. World Bank (1996), World Debt Tables. Volume 1: Analysis and Summary Tables, Washington D.C. 101 4. Debt crisis in Sub-Saharan Africa 1. - Introduction Sub-Saharan Africa (SSA) covers the whole of Africa south of the Sahara. Apart from South Africa 1 , SSA comprises 47 countries with a total area exceeding 22 million square kilometres. Obviously, these countries differ in size, population, geographical and climatic conditions, natural resources and other economic, political, social and cultural characteristics. Total population is over 600 million, ranging from Nigeria (with 124 million) to Seychelles (with a population of 80,000) 2 . Population of SSA countries, with more than 1 million (,000 people) with more than 10 million with more than 1 million Ghana Zambia Mozambique Somalia Cote d'Ivoire Senegal Madagascar Rwanda Cameroon Chad Guinea Angola Burundi Zimbabwe Benin Burkina Faso Sierra Leone Malawi Togo Mali Eritrea Niger Central African Republic 0 Liberia 5.000 10.000 15.000 20.000 Congo, Rep. Mauritania with more than 20 million Lesotho Namibia Nigeria Botswana Ethiopia Gambia, The Congo, Dem. Rep. Gabon Tanzania Guinea-Bissau Kenya Mauritius Sudan Swaziland Uganda 0 1 2.000 4.000 6.000 8.000 10.000 12.000 0 50.000 100.000 150.000 Which is considered as a “developed” country by the WTO criteria. 2 Cape Verde, Comoros, Equatorial Guinea, Mayotte, Sao Tome and Principe and Seychelles have each a population less than 1 million. 102 Similar disparities are reflected in population densities, with the smaller states in general having a higher density than the larger ones. Africa experiences almost every type of geographical and climatic condition, from tropical rain forests to semi-arid and desert conditions. There are different natural resources: petroleum (Angola, Cameroon, Republic of Congo, Gabon, Nigeria), minerals and precious metals (diamonds in Botswana, bauxite in Guinea, gold in Ghana, uranium in Niger, iron in Liberia and Mauritania, copper in Zambia, phosphate in Togo). There are differences in respect of economic, financial and political systems and forms of government as well as political “heritage”: francophone and anglophone areas. Within the same region, there are relevant differences, too. In fact, the European Commission was establishing, at the end of the 1990s, a new framework for EU-ACP (African, Caribbean and Pacific) trade relations, in order to define a post- Lome agreement 3 . Europe de facto imposed the need for the ACP to define geographical configurations for future trade arrangements with regional institutions representing these configurations. The EU has proposed free trade based Economic Partnership Agreements (EPAs) as the model for new trade arrangements and they will be negotiated largely on a regional basis. Unfortunately, there is a clear absence, in most ACP regions, of the institutional capacity for the negotiation of reciprocal preferential trade arrangements at the regional level. And there is a clear absence of the necessary economic convergence to allow the negotiation of region to region agreements, which accommodate the trade realities and needs of each participating country. The final appendix of this chapter tries to demonstrate how relevant is the heterogeneity of existing economic conditions and characteristics within the same SSA regional aggregations, which are not the final result of a long endogenous process of integration, as the EU is. Using continent-wide averages for SSA can be misleading because substantial population and GDP differences exist among the subregions of West Africa, Central Africa, East Africa and Southern Africa, and also within subregions. Considering the whole African continent, the five biggest economies of the continent (South Africa, Nigeria, Algeria, Egypt and Morocco) account for 37 per cent of the population, and 59 per cent of GDP. The 33 least developed countries have 45 per cent of the population, and only 17 per cent of GDP. From another perspective, the 11 oil-exporting countries of Algeria, Egypt, Libya, Tunisia, Côte d’Ivoire, Nigeria, Cameroon, Gabon, Republic of Congo, Angola and Equatorial Guinea account for 49 per cent of GDP, and 36 per cent of the population. Notwithstanding this heterogeneity, SSA countries exhibit common economic problems. Low per capita incomes, high population growth rates, agriculture remaining the dominant sector and poverty. 3 According to the new Cotonou Agreement, signed in 2000, which replaces Lome Convention between EU and the ACP, the parties agree to a preparatory period of 8 years before moving to new WTO compatible trade arrangements. Formal negotiations for these agreements will start in September 2002. The agreements will enter into force by January 2008 unless both parties agree earlier dates. The 8-years period is supposed to be used to prepare the ACP States for the EPA trade arrangements. In 2004 the EU will assess the situation of nonLDC ACP countries in relation to these agreements. If after consultations these countries decide they are not in a position to enter economic partnership agreements, the EU will examine alternatives in order to provide these countries with new trade arrangements equivalent to their existing situation, but in conformity with WTO rules. The ACP and the EU will then carry out a formal review in 2006 of the future arrangements planned for all countries to ensure that no further time is needed for preparations or negotiations. The ACP is yet to put forward a model for new WTO compatible trade arrangements, thus the EC's EPAs are the only option for future trade arrangements on the table at present. 103 Tab. 1 - Sub-Saharan Africa (excl. South Africa), Macroeconomic indicators, 1975-1999 1975 1980 1985 1990 1995 1999 GDP per capita, (Cur. US$) 326 541 353 392 310 325 GNP per capita (Cur. US$) 320 509 338 381 288 306 Gross dom. savings (% of GDP) 22.61 22.98 16.53 17.37 13.66 13.61 Gross dom. invest. (% of GDP) 22.03 19.307 13.82 15.92 18.017 20.32 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. The 1980s and 1990s saw a decline for the SSA region: this downward trend is reflected in the very poor growth performance of GDP and GNP levels, which appear to be below those observed in1980, and being back at the 1975 level. SSA region experienced a “two-lostdecades”: the saving and investment performance was really discouraging, and the observed trends much worse than the general trend of developing countries. Some 80 per cent of Low Human Development Countries - countries with high population growth rates, low income, low literacy, and low life expectancy - are in SSA. Four of every 10 Africans live in conditions of absolute poverty, and recent evidence suggests that poverty on the continent is increasing. At about 13.6 percent of GDP, the current levels of domestic savings are insufficient to finance the level of investment required for growth and development at the level needed to reduce poverty. Moreover, despite substantial policy reforms and concerted efforts to attract foreign direct investment by African governments, the continent’s share of FDI has remained very low. FDI flows to Africa represent a minuscule 2-3 per cent of global FDI flows 4 . Moreover, the flows are concentrated towards a few countries. Only about 20 countries are beneficiaries of FDI, with Nigeria, Egypt, Morocco, Tunisia and Angola together accounting for two-thirds of flows. The flows are concentrated on a few sectors and activities; more than 50 per cent go to support the oil and petroleum industry, and the rest goes to extractive, mainly mining, activities5 . ODA is a basic source of financial inflows 6 , but it is declining, and the heavy debt burden of African countries has dampened both private and public investment. 2. Current economic growth, social and poverty trends in Africa The Lldcs have clearly shown the worst social and economic performance during last years. Tab. 2 - Human development differences, 1999 areas Life expectancy (years) Adult literacy School enrolment Per capita rate (% popul. ratio (% of relevant GDP (at ppp) >15 years) age group) High income countries 76.4 97.4 Ldcs 64.7 72.3 Lldcs 51.9 50.7 Sources: UNDP and World Bank, 2000 4 86.0 60.0 37.0 24,430 3,410 1,790 Human Development Index (HDI) 0.89 0.64 0.43 UNCTAD (1999), Foreign Direct Investment in Africa: Performance and Potential, UN, New York. 5 UNECA (2000), Globalisation, Regionalism and Africa’s Development Agenda, Paper Prepared for UNCTAD X February 12-19, 2000, Bangkok, Thailand. 6 Measured relative to recipient GNP, the median value of aid to African countries now stands at nearly 10 times the amount received by Western Europe under the Marshall Plan. 104 The group of Sub-Saharan African countries still contains the hard core of the problem of marginalisation in the world economy. Tab. 3 - World GDP across areas AREA East Asia and Pacific South Asia Latin America and the Caribbean Europe and Central Asia Middle East and North Africa Sub-Saharan Africa Total, Ldcs Total, Lldcs Total, Industrial countries (Ics) % Ldcs/Ics % Lldcs/Ics Per capita GDP ($)1999 3.500 2,030 6,280 5,580 4,600 1,450 3,410 1,790 24,430 13.9 7.3 Per capita GDP % var. 1975-1990 1990-1998 1.7 - 0.3 2.0 3.3 0.9 1.5 1998-2002* 3.8 1.1 1.6 *forecast. Source: World Bank, 2000 According to the latest data, some 300 million Africans live on barely 65 cents a day. The average GNP per capita for the region is US$492, but in 24 countries GNP per capita is under US$350, with the lowest incomes found in Ethiopia (US$100), the Democratic Republic of Congo (US$110), Burundi (US$120), and Sierra Leone (US$130). This phenomenon is confirmed by the head count ratio measurement. The head count ratio, as argued by D. Gordon and P. Spicker (1999) is the most common measure of poverty. It refers to the proportion of individuals, households or families that falls under the poverty line. If q is the number of people identified as poor an n the total number of people in the community, then the head-count ratio measure H is q/n. The head-count ratio ranges from zero (nobody is poor) to one (everybody is poor). This simple indicator provides useful information on the incidence of poverty and the distribution of poverty among the population. However, the head-count ratio does not capture the intensity of poverty, i.e. how far the poor fall below a given poverty line 7 . The head-count ratio has been under severe attack for thirty years8 . In 1968, H. W. Watts 9 noted that it had “little but its simplicity to recommend it” and A. Sen10 has remarked that, considering its inadequacies, the degree of support commanded by this measure is “quite astonishing”. The head-count ratio can be dangerous for monitoring the effectiveness of propoor policies. Successful policies aimed at raising the wellbeing of the poorest of the poor will not affect the head-count ratio if their new living standard is still below the poverty line. On the other hand, successful pro-poor policies aimed at persons just below the poverty line will reduce the head-count ratio. 7 See A. Sen, 1981; A. J. M. Hagenaars, 1986. 8 A. B. Atkinson, 1989. 9 H. W. Watts, 1968, p. 326. 10 A. Sen, 1979, p. 295. 105 Head count ratio 1987/1998 50 1987 45 1998 40 35 30 % 25 20 15 10 5 0 East Asia East Asia But China Europe And Central Asia Latin America Middle East and North Africa South Asia S u b -Saharan Africa TOTAL TOTALE escluso Cina Source: UNDP, 2000 According to the latest edition of African Development Indicators 2001 - an annual World Bank compilation of key African social and economic data for the period 1970-99, covering indicators such as trade and external debt, communications, and aid flows - the slowdown in growth was the result of regional and civil wars, poor governance in some countries, and serious external shocks such as the rapid hike in oil prices at the same time that export earnings from primary commodities collapsed. Moreover, the report warns that growth is below the annual level needed to prevent a rise in the numbers of poor people on the continent. Growth in Africa slowed significantly after 1998, with average per capita GDP falling by almost 1 percent in 1998-99 11 . While growth trends for the region as a whole remain depressed, some African countries are doing well. Again, we have to remind that fourteen countries have grown on average by 4 percent a year during the 1990s, with rising annual incomes of 2-3 percent and even higher, with another 10 countries following close behind with growth rates above 3 percent a year. Some countries have grown at 7 percent a year or higher (Mozambique and Uganda, 7.1 percent). 11 World Bank (2001/b), African Development Indicators 2001, Oxford University Press, Washington DC. 106 In late 200, the World Bank gave US$155 million in credits to help seven African countries - Madagascar, Mali, Mauritania, Niger, Rwanda, Zambia, and Uganda - cope with an unexpected surge in oil prices and other losses in their terms of trade. These factors were causing serious hardship for the poor in terms of rising energy and transportation costs, which in turn were jeopardising the success of the countries’ reform programs. Civil conflict in the region has blunted and reversed growth prospects for war-torn countries. While the trend for many African countries during the 1990s was one of slow but steady economic improvement, those in conflict suffered negative growth and an alarming deterioration in basic conditions: Angola (0.2 percent), Burundi (-2.4 percent), Democratic Republic of Congo (-4.6 percent), Rwanda (-2.1 percent), Sierra Leone (-4.6 percent). Sierra Leone is a striking illustration of this trend with the region’s lowest life expectancy rate at just 37 years, and its highest infant mortality rate at 169 deaths per one thousand. In the early 1990s, there was the hope that the globalisation of finance and production, and the liberalisation of economic activity, would promote diminuishing income disparities between countries. During the 1990s there has been a process of economic liberalisation in many developing countries. Nevertheless, no real progress in increasing real incomes and social development, reducing poverty has arrived. GNP per capita (US$ thousands) 0.0 Seychelles 2.0 4.0 6.5 Gabon South Africa Namibia Egypt Cape Verde Equatorial Guinea Côte d'Ivoire Lesotho Zimbabwe Guinea Mauritania Kenya Zambia Gambia, The Togo Angola Tanzania Rwanda Mali Mozambique Eritrea Malawi Sierra Leone Congo, Dem. In view of the international target set by international community in the 1990s in order to halve poverty in the world by 2015 12 , UNCTAD addressed the question of how long it will actually take for each poor country to cross the $900 per capita threshold that currently forms the criterion for graduation from the Ldc category. If the trend growth rates of 1990-1998 persist, only Lesotho, in Sub-Saharan Africa, will cross this threshold by the end of 2015 13 . 12 See chapter 13 and DAC (1996), Shaping the 21st Century: The Contribution of Development Cooperation, OECD, Paris. 13 UNCTAD (2000), The Least Developed Countries. 2000 Report, Geneva. 107 Tab . 4 - Forecast to escape from poverty, given the current trends, in 30 SSA Lldcs Already there By 2015 By 2025 By 2050 By 2100 Later than 2100 Negative growth Cape Verde Equatorial Guinea Djibouti Lesotho Sudan Guinea Mozambique Uganda Benin Eritrea Ethiopia Mauritania Burkina Faso Central Rep. Africa Malawi Mali Angola Burundi Chad Dem. Rep. Congo Gambia Guinea-Bissau Madagascar Niger Rwanda S. Tomé & Principe Sierra Leone Togo Tanzania Zambia Source: UNCTAD, 2000 Projections of social indicators on the basis of the 1990s trends fall well short of the international targets. Social indicators show mixed progress with a welcome rise in literacy and school enrolments for girls on one hand, but declining immunisations for children, and a widening HIV/AIDS epidemic across the region, on the other. Overall Africa’s demographic transition remains slow. Fertility has started to decline, particularly in countries with higher incomes and better access to contraception. Still, some countries in the region have the highest fertility rates in the world: Niger (7.3); Somalia (7.2); Angola and Burkina Faso (6.7). Even though the age dependency ratio has changed little, the percentage of the population aged 0-14 has fallen slowly during the past two decades. Growing urbanisation, and the rapid exodus of rural Africans to the continent’s cities, has given Africa the largest rate of urban population growth in the developing world. Moreover, on current trends, the continent’s urban population is expected to outnumber the number of people living in rural areas by 2025. In countries like Nigeria, Kenya and Tanzania, there are now twice as many people living in urban centres today than 20 years ago; in Mozambique, the percentage has almost tripled during the same period. Child mortality is a particularly acute problem for many countries in Africa. Infant mortality is close to 10 percent, and on average 151 of every 1,000 children die before the age of 5, although in many countries the mortality rate exceeds 200 per 1,000. The region has had the smallest improvement in under-5 mortality since 1970, and some countries - including Kenya, Zambia, Mozambique and Côte d’Ivoire - saw infant mortality increase in the 1990s. This compares with 53 in East Asia and 9 in high-income countries. Even allowing for Africa’s low incomes, its under-5 mortality rates are exceptionally high. Although life expectancy has risen slightly in Africa, this is happening at a slower rate than elsewhere and, since 1990 the HIV/AIDS epidemic has caused it to decline, especially in countries with high adult infection rates. In Zimbabwe, for example, life expectancy has fallen by five years, while in Botswana, it has fallen by over ten. Today in 21 African countries more than 7 percent of adults live with HIV/AIDS, with the highest absolute number of cases found in South Africa, where one in every five adults has contracted the virus. Countries like Niger and Sudan, which have some of the lowest incidence of AIDS in the region, offer great potential for control. Analysis of income distribution in Africa shows a fairly high degree of inequality. Compared with other regions of the world, Africa has the second most unequal income distribution next to Latin America. The Gini coefficient for Africa as a whole is 44.4 per cent. The highest 108 values for inequality are for South Africa, Kenya and Zimbabwe. The lowest are for Northern Africa (Egypt and Algeria) and Ghana. Shares of total expenditure by quintiles confirm the picture of relatively high inequality. 3. Domestic financial conditions to meet international development goals The issue of development finance involves the analysis of three interrelated issues: resource requirements for economic growth; the efforts to mobilise domestic resources; the need for, and availability and effects of, external sources of finance. To reduce poverty by half in Africa by the year 2015, that is the target set by international community of donors, requires a 4 per cent reduction in the ratio of people living in poverty each year. Change in poverty can arise for two reasons: a change due to growth in mean consumption expenditure (appropriately adjusted for the change in the poverty line); and a change in the distribution of income (the inequality measure). For Africa as a whole, GDP growth of about 7 per cent per annum would be required to achieve this annual reduction in poverty. Increases of 5-6 per cent are needed for North Africa and Southern Africa, 6-7 per cent for Central Africa, and 7-8 per cent for the West and East African sub-regions 14 . In order for Africa’s GDP to grow on average at 7 per cent per annum, additional investment will be needed. An estimate of the amount can be made by employing the Harrod-Domar model, which uses the savings rate and incremental capital output ratio to derive the rate of growth of GDP. Thus, for a desired rate of growth of GDP, required investment can be calculated by assuming an incremental capital output ratio. Domestic sources of finance are defined as gross net domestic savings, which are measured as gross net investment minus the net inflow of external finance. Investment is measured as the additions to physical capital stock, which aims at measuring additions to the production capacity of the economy 15 . The determinants of savings are analysed after being disaggregated into private and government savings. The average private consumption ratio for SSA countries has fluctuated at around 85 % of GNP, about 20% higher than the average for other developing countries. This phenomenon, associated with falling per capita consumption levels in SSA Lldcs during the past two decades, implies a long period of slow and in some cases declining per capita income growth. In fact, recent econometric analyses have shown a robust long-term relationship between private consumption and income, with an income elasticity of consumption of about 0.8 16 . For Africa as a whole, investment of 33 per cent of GDP would be needed to reach 7 per cent per annum growth. The current domestic savings rate is about 15 per cent. 14 UNECA (2000/b), Economic Report on Africa 1999: The Challenges of Poverty Reduction and Sustainablity, Addis Ababa. 15 Even though investment and savings in the national accounts exclude investment in human capital formation, which plays a significant role in enhancing production capacities in the economy, and therefore miss an important component of development finance. 16 This phenomenon supports the idea that an important role of foreign aid in Lldcs, which are subject to external shocks, is indeed the smoothing of consumption. See UNCTAD (2000). 109 Investiment shares in GDP (percentage), 1998 Source: DAC, 2000 Africa must address the key issue of raising domestic savings rates, but in the short run, expectation of significant change is unrealistic, in view of the existing low levels of income. Stabilisation of the macro economy will stimulate savings by creating an economic environment where private agents can plan their future with a large measure of confidence. Moreover, prudent government behaviour and fiscal discipline will be expected to contribute to increased savings. Financial liberalisation will theoretically lead to higher savings through the effects of high real interest rates on savings. However, most empirical work suggests that the effect of interest rates on gross savings is weak or non-existent. The most important determinant of savings in Africa has been found to be the level of real income. Very poor people save little or nothing and income must rise above the subsistence level before increases in income result in higher savings. It would take 18 years of 5.3 per cent of GDP growth for sub-Saharan Africa to reach the income threshold where further increases result in increased savings rates. More research is needed to advance understanding of the factors determining savings rates in sub-Saharan countries. Current understanding of the linkage between interest rates and savings indicates that African governments have few policy instruments to increase savings in the medium run and for as long as incomes remain low. Unfortunately, there is significant hysteresis, because there are ‘sunk costs’ in promoting savings that can operate in such a weak environment, and while such savings ultimately force an improvement in economic environment, unless a certain threshold is reached, it will simply perpetuate and entrench the weak environment. With the prevailing levels of domestic resources available for finance, very low levels of per capita income and private consumption, and vulnerability to frequent and large external shocks, the poor economies can only rely on external financing. 4. External financial conditions to meet international development goals There are three main sources of external finance: aid, private capital flows and external debt. 110 According to African Development Indicators 2001, two important sources of finance, foreign direct investment (FDI) and official aid, are declining in size, and tend to favour those countries with lucrative mining and oil industries in the case of FDI, or countries with sound social and economic policies in the case of aid. Recent foreign resource flows to Africa have been far short of the volume needed to meet the poverty reduction objective. Commodity prices are beyond the control of African policy makers, flows and maintaining exemplary domestic economic management can influence dealing with the debt overhang and Official development assistance (ODA) 17 only indirectly. 17 Foreign aid is conventionally measured on the basis of the OECD’s ODA, a concept introduced in the early 1970s. ODA comprises official financial flows with a development purpose in the form of grants (inclusive of those tied to technical assistance) and highly concessional loans. Loans are defined as highly concessional when their grant element - i.e., the subsidy implicitly included in the loan, relative to the loans’ face value - is at least 25 percent, as measured by a formula to be analysed in depth in the next section. The leading measure of foreign aid flows is the so-called Net ODA, which is the net disbursement amount, i.e. disbursements minus amortisation, of those flows classified as ODA. Conceptually, international finance flowing to capital-scarce developing countries may involve efficiency gains even if the flows accrue on market terms as long as the funds are used appropriately. Such efficiency gains translate into net financial gains for the recipient countries. The main purpose of measuring foreign aid flows is to assess the portion of those gains that is due to a pure transfer of resources from donors to recipients through below-market, subsidised financial terms -- i.e. to assess the donors’ net financial cost, rather than the (presumably larger) recipients’ benefit. Net ODA, however, does not accurately measure the cost that donors incur in connection with their aid (especially debt) flows, and as a result the evolution of Net ODA over time, as well as across donors and recipients, likely provides a distorted picture of aid trends. The five main problems: (1) ODA includes the full face value of both grants and highly concessional loans without distinguishing between the two. However, concessional loans entail repayment obligations, and, therefore, the aid they involve, i.e. the net financial cost to donors, is only a fraction of their face value. The inclusion in ODA of the full face value of these loans overestimates their aid content. Only grants, that is to say pure unrequited transfers, should be accounted at full value; (2) Under the ODA definition, non-concessional loans include loans on market terms as well as concessional loans with low degree of concessionality. The aid content of the latter – i.e., the donors’ cost involved in these loans – is therefore not captured by ODA. (3) The inclusion in ODA of official technical assistance grants by their full value can be seen as another shortcoming. In this case, the donor benefits from payments received in return for the technical assistance supplied, and this may greatly reduce the donor’s net financial cost. (4) In order to reflect donors’ opportunity costs, the discount rates used for present value calculations should correspond to applicable market rates. The fixed 10 percent discount rate utilised in ODA fails that test on at least three important dimensions to which it should be sensitive, namely time, currency, and maturity: (a) Time. Discount rates should evolve over time with market conditions prevailing at the time the aid content of loans is estimated. For example, to measure the donors’ cost as seen at the time of loan disbursement, the market terms prevailing at that time should be used. (b) Currency. At any point in time, market rates, and therefore appropriate discount rates, are currency specific. The discount rate should follow the currency in which debt service is payable. (c) Maturity. At any point in time and for any given currency, market rates depend on the length of the repayment period according to the socalled yield curve. Therefore, the discount rates applied to the debt service stream should vary over the life of the loan according to the timing of service payments. (5) In the case of variable rate loans, the construction of the future debt service stream requires a forecast of interest rate charges. This is especially important for floating rates linked to future market conditions (e.g., indexed to six-month LIBOR). ODA makes no attempt to predict these conditions and implicitly assumes that, like in the case of fixed-rate loans, variable rates will remain constant at their level at the time of disbursement. For more details, see C. Chang, E. Fernandez-Arias, L. Serven (1998), Measuring Aid Flows: a New Approach, Washington D.C. 111 The global environment and exogenous shocks are not changing in Africa’s favour. ODA is stagnant or declining, little progress has been made in reducing the debt burden, protectionist tendencies continue in Africa’s major markets, and erratic weather conditions persist. As investment of 33 per cent of GDP would be needed to reach 7 per cent per annum growth in Africa, it should be financed partly by domestic savings and the rest by foreign inflows. Given that the current domestic savings rate is about 15 per cent, a further 18 per cent would be needed from external sources. ODA for the continent averages about 9 per cent, which leaves a residual financing gap of about 9 per cent. Overseas development assistance (ODA) as % of GNP, 1998 Source: DAC, 2000 Africa-wide averages hide large variations among the sub-regions. North Africa needs only about 5 per cent of GDP in external resources to complete the financing needed to generate a GDP growth rate high enough to halve the poverty level in the sub-region by 2015. ODA to the sub-region has averaged about 3 per cent of GDP, leaving a financing gap of about 2 per cent of GDP. Financing investment for needed GDP growth is most difficult in Net Official Aid Per Capita to Central Africa where the residual financing Sub-Saharan Africa, 1990-99 gap is about 27 per cent. Policy issues that can be addressed directly by African policy 40 makers relate to domestic savings and external resource inflows other than 30 conventional ODA, such as foreign direct investment, and the causes of capital flight. 20 Official aid has followed a similarly selective trend over the same period, and 10 falling in terms of total volumes. Aid levels in 1999, for example, were US$10.8 billion 0 compared to US$ 17.9 billion in 1992 when development assistance to Sub-Saharan Africa reached its highest-ever levels. 112 The World Bank report shows that during the 1990s the region attracted an annual average of US$15.8 billion dollars in aid (nominal). As in keeping with FDI flows, most of this development assistance went to a small number of countries, which bilateral and multilateral donors considered to have adopted modern economic and social policies and were performing well. These included Cameroon, Côte d’Ivoire, Ethiopia, Ghana, Kenya, Mozambique, Senegal, Tanzania, Uganda, and Zambia, all of them received well over US$500 a year. Tanzania was the leading recipient of official assistance during the period with more than US$1 billion a year. The report also reveals how official aid to Sub-Saharan Africa has been falling from US$32 per head in 1990 to US$19 by 1998. Africa has suffered massive capital flight, estimated to total $22 billion between 1982 and 1991. At the end of 1991, the average ratio of capital flight to debt was estimated at over 40 per cent for a sample of 18 countries for which data were available. For four countries, the rate exceeded 60 per cent: Nigeria (94.5 per cent); Rwanda (94.3 per cent); Kenya (74.4 per cent); and the Sudan (60.5 per cent). Another important source of external funds is foreign direct investment (FDI) that is needed as a non-debt-creating form of resource inflow. But experience shows the share of FDI flows to Africa is very small and that it is highly biased in favour of mineral-rich countries. FDI in Africa seems to be caught in a vicious circle since it requires a hospitable economic environment and sustained high growth. Yet, FDI is needed to help create that environment and achieve that rate of growth. Moreover, FDI is the bulk of neo-liberal approach. The underdeveloped human resources base - exacerbated by out-migration of skilled Africans - and the weak physical infrastructure of the continent deter foreign direct investment. Furthermore, political and civil instability, weak institutional capacity and inefficiencies have not created an investment-friendly climate. These conditions have had important negative implications for resource mobilisation and utilisation in Africa, including exacerbating capital flight. They will therefore need to be given urgent attention. Of the US$2.52 billion in FDI that flowed into Sub-Saharan Africa during the last decade, just three countries accounted for much of that totalAngola (US$626 million), Lesotho (US$170 million), and Nigeria (US$876 million). If South Africa is excluded (as both a recipient and source of FDI), five other countries accounted for another US$576 million Republic of Congo, Cote D’Ivoire, Equatorial Guinea, Namibia, and Sudan - leaving the remaining 40 countries of Sub-Saharan Africa to compete for just $US275 million in annual FDI flows. Given the slowdown of ODA and the particular difficulty in attracting FDI18 , SSA economies have heavily relied on external debt. This picture demonstrates that debt relief is important as well as to guarantee new lending for the future. Thus, in empirical terms, it can be very useful to investigate the complex relationship between development and debt in SSA, in terms of both three-gaps approach and links between different sources of external finance. The combination of a large number of different types of variables, the complex interrelationship between “levels” of actors, the emphasis on qualitative factors, all make the empirical verification or falsification of theoretical propositions particularly difficult. The situation is made even more difficult by the fact data availability and reliability is a critical point. 18 See N. Bhinda, J. Leape, M. Martin and S. Griffith-Jones (1999), Private capital flows to Africa: Perception and Reality, Forum on Debt and Development, The Hague. 113 5. - The African Problems of Balance of Payments International relations are of enormous importance to Sub-Saharan African countries. The limited relatively “modern” sector of the economy of the African states shaped in the era of colonialism has been highly dependent on international trade. The way the international trade has expanded in the past has affected the distribution of income and investment both within and between African countries. The concept of the balance of trade is based on the notion that in the long run a nation should export to all countries a total amount of goods and services equal in value to its total imports from all other countries. The balance of trade of African countries shows that most of the African states have been importing an amount of goods and services as great by value as they have been exporting in recent years. It is not an unexpected result, particularly referring to the whole SSA region. Quite simply, what this shows is that we have an ex-post balanced pattern of trade in goods and services, which could clearly reflect the fact that imports are constrained by the limited amount of available resources and by the need of financing debt servicing rather than imports. In fact, the servicing of the external debt erodes the meagre foreign exchange available for imports: this has led to the import compression problem that adversely affected both public and private investments. Moreover, the cited differences among SSA countries do matter. Total Merchandise SSA Exports and Imports (current US$) 60.000.000.000 Exports Imports 50.000.000.000 40.000.000.000 30.000.000.000 20.000.000.000 10.000.000.000 5 9 4 9 1 1 9 9 3 9 2 1 9 9 1 1 9 9 0 9 1 1 9 9 9 8 8 9 1 1 9 8 7 8 6 1 9 8 5 9 1 1 9 8 4 8 3 1 9 8 2 1 9 8 1 9 8 1 9 1 1 9 8 0 0 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. More interesting, we can compare trade balances of each of the SSA countries, taking in account some possible outcomes, simply measured through the arithmetic mean of the 19651999 period and concerning: - Net trade in goods that is the difference between exports and imports of goods. Trade in services is not included 19 . - Net trade in services. 19 This category includes goods previously included in services: goods received or sent for processing and their subsequent export or import in the form of processed goods, repairs on goods, and goods procured in ports by carriers. 114 - As a way of weighting the absolute value of the African economies, we can consider Net trade in goods to total Trade Balance ratio and external Size of the economy. Tab. 5 – Comparison between trade balances of all the SSA countries Liberia Somalia S. Tome &Principe Comoros Guinea-Bissau Burundi Eritrea Sierra Leone Rwanda Cape Verde Central Afr. Rep, Gambia, The Lesotho Chad Niger Burkina Faso Mauritania Seychelles Togo Benin Malawi Mozambique Mali Equatorial Guinea Uganda Guinea Sudan Zambia Madagascar Ethiopia Tanzania Swaziland Senegal Namibia Botswana Congo, Rep. Congo, Dem. Rep. Cameroon Zimbabwe Ghana Kenya Mauritius Gabon Cote d'Ivoire Angola Nigeria South Africa (1) J (2) (3) (4) L L L L L L L L L L L L L (5) L L L (6) K L L K L L K L K K L L L L K L K L L L L L L J L K L K L K K L L L L L L J L J L L L L L L K L L K L L L L L L L J J J J J L L L J J J J J J J L L L L L The outcomes are: 115 (7) M H M H H M H M M H L M H L M H L M M H H H L M H M H H M H H M H M H M M M M H H M H M H H H (8) .. .. 17 34 56 62 66 92 109 135 178 199 226 260 290 292 370 378 511 518 542 586 650 711 726 829 832 842 908 914 1.163 1.264 1.534 1.614 1.650 1.730 1.937 2.244 2.545 2.567 2.600 2.673 2.775 5.024 5.467 12.871 33.321 (1) Exports of goods > Imports of goods (2) Exports of goods < Imports of goods (3) Exports of goods = Imports of goods (4) Exports of services > Imports of services (5) Exports of services < Imports of services (6) Exports of services = Imports of services (7) Net trade in goods to total Trade Balance ratio: high (H), medium (M), low (L) (8) 1999 External Size of the economy: Absolute value of Total exports (million US$) There are some points of significance showed by table 5. First, South Africa is effectively an outlier compared to other SSA economies, as its external size demonstrates. It makes sense to differentiate between oil-importing and oil-exporting SSA countries. All the five oil-exporting economies (in order of metric tons of exports they are Nigeria, Angola, Gabon, Republic of Congo, Cameroon, all being in bold and italics fonts) have a net trade in goods surplus and they have a big external size, compared to other SSA countries. Variations in export prices and earnings for the five major oil exporting countries (which in the 1980s have accounted for about half of SSA export earnings) have dominated terms of trade trends in the region as a whole. The others net exporters of goods show an export concentration in natural resources (Botswana: meat and copper; Cote d’Ivoire: oil palm products, cocoa, cotton and coffee; Democratic Republic of Congo: petroleum and diamonds; Guinea: forest products, and coffee; Liberia: forest products, iron, coffee, oil palm products; Zambia: copper; Zimbabwe: tobacco, meat, sugar, cotton), but also manufactured goods (Cote di’Ivoire and Zimbabwe). In general terms, the performance of the manufacturing sector has been disappointing and not able to provide the growth and employment which SSA countries need. Concerning services (previously referred as nonfactor services), which are the economic output of intangible commodities that may be produced, transferred, and consumed at the same time, there are only three countries (Mauritius, Kenya, Seychelles) that registered exports higher than imports. All of them are countries being classified as Medium human development 20 and with an high level of GNP. There are significant limitations in using the arithmetic mean: for example, Mauritania’s balance of net trade in goods is in equilibrium because this country has passed from high negative to high positive balance as well as the Comoros’ balance of trade in services. Due to the civil war problems, Liberia and Somalia have no data on size of their current external economy; however in the past the value of total Liberia’s exports was similar to Madagascar, and the value of Somalia’s exports was similar to Burundi. 20 Every year since 1990, the United Nations Development Programme has commissioned the Human Development Report by an independent team of experts to explore major issues of global concern. The Report looks beyond per capita income as a measure of human GDP per capita (1985 PPP US$). The annually updated Human Development Index (HDI) ranks 162 countries by a composite measure that includes life expectancy, educational enrolment and adult literacy, and income per person. In the 2001 Report, Sierra Leone, where a child born today will probably die before reaching the age of 39, and only 32 percent of the adults can read, is ranked last. The bottom 28 countries on the Index are all in Africa. And 19 African countries are suffering setbacks in the human development index. See UNDP (2001), Human Development Report 2001, Oxford University Press, New York. 116 In any case, in absolute terms, we can affirm that Africa has not experienced the benefits deriving from globalisation, that is the intensification of the global interdependence of economies. Africa’s share of global exports of goods and services has declined trend-wise from 4.2 percent in 1985 to a mere 1.8 percent in 1999. The comparison of trends of total external debt21 and GNP in SSA shows the existence of a link between poor economic performance and external debt. Tab. 6 – Comparison between economic performance and external debt in SSA (Cur. US$ million) 1975 1980 1985 1990 1995 1999 Total External Debt 19,597 60,612 107,220 177,050 210,160 202,660 GNP 97,605 179,660 138,150 180,110 155,280 183,570 Between 1980 and 1995, the SSA (excluding South Africa) countries’ total stock of external debt grew rapidly from US$ 60.6 billion to US$ 210.2 billion (1988-89). On the other hand, gross national product decreased from US$ 179.7 billion (in 1978-79) to US$ 155.3 billion. In other words, the debt-to-GNP ratio increased from 20.1% in 1975 to 135.3% in 1995. Even though the rescheduling terms of the Paris Club became increasingly concessional for SSA countries, and also included more and more debt reduction on eligible bilateral debt, more and more heavily indebted African countries were unable to pay their debt service due during the 1990s. And SSA countries are still facing at least five major possible reasons for weakness and vulnerability in the balance of trade. First, the export enclave has come to be dependent on imports of consumer necessities including even packaged foodstuffs – tinned sardines, milk, tomatoes – because of the truncated expansion of the domestic economy. Not infrequently, the rich groups in the export enclave use their incomes to purchase and import luxury items (biased path of development). And dramatic is now the food situation in a continent that as recently as 1970 was selfsufficient in food. “Now almost one in four Africans is undernourished and the food deficit is estimated at more than 30 million tons per year - 10 million tons for SSA alone - and likely to rise to 50 million tons by the end of the present decade” 22 . Second, the prices of the country’s primary exports may relatively fall on the world market, making it difficult to pay for the imported manufactured consumer goods on which those living in t he export enclave have come to depend (declining terms of trade). In fact, since the independence in the 1960s, Sub-Saharan African terms of trade has never registered a positive trend, except the period between the first (1973) and the second (1979) oil shock 23 . 21 Total external debt is debt owed to non-residents repayable in foreign currency, goods, or services. Total external debt is the sum of public, publicly guaranteed, and private nonguaranteed long-term debt, use of IMF credit, and short-term debt. Short-term debt includes all debt having an original maturity of one year or less and interest in arrears on long-term debt. 22 UN (1990), Africa’s Commodity Problem: Towards a Solution, UN, New York, p.18 23 Paul Cashin and Catherine Pattillo have examined the persistence of shocks to the terms of trade, using annual data on 42 SSA countries between 1960-96 and they found that the persistence of terms of trade shocks varies widely. For about half the countries such shocks are short-lived, while for one-third of the countries (those that have large shares of petroleum imports in total imports, small shares of non-fuel commodity exports in total exports, and highly concentrated in exportable commodities) such shocks are long-lived. See P. Cashin and C. Pattillo (2000), Terms of Trade Shocks in Africa: Are They Short-lived or Long-lived?, IMF Working Paper, N. 72, Wahington D.C. 117 ”During the period 1980-1987, Africa’s terms of trade deteriorated from 100 to 66 - from 100 to 70 those of SSA -. The share of SSA exports in the total world has declined from 2,4% in 1970 to 1,3% in 1987” 24 . International prices for the main export items of African countries, such as coffee, cotton, or copper, have fluctuated widely creating uncertainty. The African inability to cope with external shocks has contributed to Africa’s debt problems and very low rate of economic growth. Many governments such as Zambia responded to commodity price booms in the late 1970s by sharply expanding public expenditure for import-intensive public investment programs that they financed with foreign borrowing when revenues fell with subsequent steep declines in commodity prices, assuming that these negative price shocks were short-lived. SSA Merchandise Terms of Trade (price index '87) 140 120 100 80 60 40 Merchandise Terms of 20 Trade (price index '87) 9 9 7 1 9 9 5 9 1 9 9 3 1 1 9 9 1 9 9 1 9 8 7 9 1 9 8 5 1 1 9 8 3 8 1 1 9 8 9 9 1 9 7 7 1 1 9 7 5 7 3 1 9 7 1 9 1 9 7 9 1 1 9 6 7 6 9 1 1 9 6 5 0 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. Third, there is a high level of concentration of SSA export in a very few items, which increases the vulnerability to exogenous shocks. As average, three commodities still represent around 70 percent of merchandise exports. 24 S. Sideri (1992), “External Financial Flows: The Case of Africa”, African Review of Money Finance and Banking, N.1, p.93. 118 SSA Exports of top 3 commodities as % of merch. exports 90 80 70 60 50 40 30 20 10 99 98 19 97 19 95 96 19 19 94 19 19 93 92 19 91 19 19 90 89 19 19 88 86 87 19 19 19 85 84 19 19 83 81 82 19 19 19 19 80 0 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. Fourth, any country seeking to implement a development strategy must import capital equipment and machinery. These imports even if reduced to the critical minimum required to ensure that development does take place, tend to contribute to a balance of trade deficit (technological transfer). Fifth, increased market access for African countries requires a reassessment of priorities on the part of their developed trade partners, who should focus their efforts on lowering protectionism in key markets for African exports. The share of African countries in international trade has been eroding since the 1970s, and even as a proportion of developing countries’ exports African countries’ share has fallen over the same period. Compared to a vulnerable situation of trade balance, the perennial imbalance in the services sector, driven by external debt payments and the cost of transport and financial services, which can limit the growth rate of imports, continues to put pressure on the current account balance, and to claim an inordinate share of foreign revenue from merchandise exports. 119 SSA Balance of Payments (net, current US$) 15.000.000.000 Current account balance Trade balance 10.000.000.000 5.000.000.000 8 9 6 1 9 9 4 9 1 1 9 9 2 9 0 9 1 1 9 9 8 8 6 1 9 8 4 1 9 8 2 1 9 8 0 9 1 1 9 8 8 7 6 9 1 1 9 7 4 7 2 1 9 7 0 9 9 7 1 -5.000.000.000 1 1 9 6 8 0 -10.000.000.000 -15.000.000.000 -20.000.000.000 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. From the graph, we see that the trade deficit is usually much smaller than the current account deficit. It shows that it is not just a trade problem: even with a trade account on balance or in the positive, this group of countries would have had negative current account, mainly because of the external debt service. The trade account and in particular the current account and their deficits give an indication of the size of the external finance needed by these countries. From the point of view of the ‘foreign exchange constraint’ the current account matters, and it is important to notice that the trade deficit is smaller than the current one. Thus the trade balance needs less foreign exchange to match its deficit, which is now due to a debt trap. This is a clear indication of additional fragility by these countries with respect to trade relations only25 . Large and persistent current account deficits can create a fertile environment for external crises, especially when those deficits are financed through short-term capital inflows. While the specific causes are likely to vary, common precursors to external crises include falling stocks of foreign reserves, lax monetary and fiscal policies, and rising levels of (especially short-term) foreign debt. Periods of external crisis are typically associated with increased exchange rate and interest rate volatility, falling asset prices, declining foreign reserves. It is in these immediate postcrisis periods that punitive reductions in both domestic consumption and national output occur, with associated increases in both unemployment and general socio-political instability. Private agents and policy makers often view persistently large current account deficits as an 25 The actual debt service has always been a relevant share of exports, in the range of 20%. These countries have a negative, large and increasing net factor income. ODA flows have substantially increased from the eighties to the nineties: the figures seem to indicate that aid has been used to service the debt, at least partially. Moreover, these data seem to confirm the existence of three gaps in SSA - given the low level of domestic savings (9.26 percent of GDP is the average for the nineties) and public revenues -, and it means that these countries have to resort to foreign savings to finance their growth process. 120 indicator of future difficulties associated with external debt and the balance of payments, particularly when it is referred to fragile economies. A structural component of unbalance of SSA economies is surely shown by the fact that there is an enormous, permanent and pervasive overall budget deficit in all the 47 countries, for all the 1980-1999 period, apart from the Botswana case (overall surplus) and Nigeria (balance). And these structural unbalances reach an averaging value of around 20 percent as share of GDP in Sao Tome, Angola, Eritrea, Cape Verde. We have mentioned the possibility of interrelations between external and fiscal debt: in Africa servicing of debt placed fiscal pressure, which has an adverse effect on public investment, inducing their reduction and concentrating expenditures on current part, which is unable to produce future revenues. Thus, in the SSA case it is important to analyse the complex link between the fiscal and the external deficit, in order to define a sustainable escape from debt crisis. 6. – Nature and causes of external debt crisis in Africa Concerning the causes of external debt crisis in Africa, there is a preliminary point to be stressed. The problem of foreign debt is neither a specific African problem nor a new one. Quite the contrary: in absolute terms, SSA debt is 7.8 percent of total world external debt, which equals to 2,460 US$ billion. In fact total SSA countries’ external debt is around 191.6 US$ billion26 ; less than the Brazilian external debt, equals to 244.7 US$ billion27 . However, its relative burden is very high: ten countries have the present value of debt (% of GNP) higher than 100 percent 28 . Gabon 108.21 Cote d'Ivoire 116.67 Sierra Leone 136.19 Mauritania Sudan Zambia Angola 169.48 171.52 172.15 286.37 Congo, Rep. 286.59 Guinea- S. Tome & Bissau Principe 347.48 449.71 Therteen countries have debt service ratio higher than 20 percent. Mozamb. Angola Uganda Cameroon Zimbabwe 20,00 21,06 23,68 24,30 25,30 Cote d'Ivoire Kenya Mauritania S. Tome Rwanda & Principe Sierra Leone Burundi Zambia 26,24 26,71 29,88 45,61 28,44 29,08 29,56 46,63 Compared to very lower level in 1971: respectively, 20% and 9 %. The importance of debt servicing component confirms the nature of vicious circle of external indebtedness. By comparison, the 1995 level of external debt per capita and two other important ratios confirm the growing problems of SSA. 1975 1980 1985 1990 1995 1999 Total external debt per capita (Cur. US$) 68,8 183,5 264,8 377,8 390,5 337,3 Total External Debt to export ratio, % 82,2 105,9 256,0 327,3 365,3 313,5 Total External Debt to GDP ratio, % 19,7 31,7 74,4 95,4 125,7 103,8 26 It is 215.8 US$ billion, including South African debt. 27 Among developing countries, the other main external debtors in the world are: Russian Federation, Mexico, China, Indonesia and Argentina. 28 And twelve other countries have the ratio higher than 60 percent. 121 Debt burden is so high that it is only thanks to aid that net flow of international resources is positive 29 . The problems of SSA region with servicing debts are shown by the principal and interest arrears components of total debt, which have grown from less than 750 US$ million (1975) to around 62 US$ billion (1995). Cur. US$ million Total stock of arrears (Int. and Prin.) 1975 1980 1985 1990 1995 1999 743.2 3,441 10,722 26,831 61,870 45,125 The arrears/total external debt ratio has shifted from 3,8 (1975) to 5,7 (1980), 10,0 (1985), 15,2 (1990), 29,4 (1995), 22,3 (1999). Thus, the “arrears-burden” has significantly affected the debt burden, making it unsustainable. Is it possible to affirm that current SSA debt crisis is an anomaly in history? The contrary is true. History experienced a lot of external debt crises, as when Haiti’s ratio of debt to exports was of 484 percent, in the 1890s, without considering the important question of German debt and debt relief plan during the inter-war period 30 . For SSA countries themselves, the problem of debt is not new. The colonial period witnessed a flow of loans from European centres to the colonies, in order to finance public infrastructure development - such as railways and roads to link ports to export the primary commodity production sites – and to fund military presence. The repayment of this debt by colonial administrators created serious difficulties, because of the instability in the world commodity market and the vulnerability of the African colonies to this. After the 1929 great depression, African exports declined by about 42 per cent, international credit decreased and the colonies were induced to highly concentrate their production in a few cash crops to repay their debts. Not only does the pre-II World War situation demonstrate that debt crisis is not new, but it also justifies some analogies with the current situation and vicious cycles 31 . SSA debt crisis has its origin during the colonial period, which expanded the enclave of primary commodity exporting economies and generated a situation of vulnerable specialisation, indebtedness and dependence on external finance and on cash-crops as sources of foreign exchange required for repaying debt (commodity export-led strategy). But the current African debt crisis has been exacerbated by some other international factors. After independence, the inadequacy of domestic capital markets, the inflationary pressure of monetary expansion, added to the structural inability to generate domestic revenue induced 29 But 35% of aid goes to expatriate technical experts. And the low level of FDI is partly due to the fact that debt stock creates a debt overhang problem that affect the confidence of investors who are usually sensitive to uncertainty. 30 Keynes gave one of the most interesting analyses of the dynamics of debt, which is currently proposed to solve external debt problems, when he studied the German difficulties to pay its war reparation debts. Keynes argued that Germany was obliged to extract the necessary funds from domestic resources, that is to cut the budgetary deficits and reduce the rate of domestic absorption (the budgetary problem) and to convert them into foreign currency. This induced a loss of purchasing power, by entailing a depreciation of local currency (the transfer problem), causing domestic incomes and prices to fall but stimulating exports and a surplus of external account in order to provide additional foreign exchange to meet debt service obligations. See J. M. Keynes (1929), “The German Transfer Problem”, The Economic Journal, Vol. XXXIX, March. 31 A. Geda (2001), “Debt Issues in Africa: Thinking Beyond the HIPC Initiative to Solving Structural Problems”, WIDER Conference on Debt Relief, Helsinki, August 2001. 122 the SSA countries to link deficit financing to external borrowing, inducing a fiscal origin of external debt. In fact, given the colonial structure, after independence it was needed spending on social and physical infrastructure, as well as to keep a social consensus and support to governments (urban-biased expenditure). Most SSA countries engaged in a massive programme of statesponsored industrialisation. At international level, SS countries imported capital and intermediate goods to develop infrastructure. Foreign borrowing supported this process. The oil price shocks of 1973-74 and 1978-79 was followed by the rise in commodity price and in public expenditure. It was tackled, partly, by resorting to external financing, particularly the private sources, interested in circulating OPEC-surplus through the international banking system into productive investments. The expansion of the Eurodollar market facilitated the access to loan. When the commodity price fall, as a consequence of the recession in the industrialised nations, African governments were unable to cut expenditure and projects. Thus, they were obliged to increase borrowing, also thanks to improved credit worthiness, as prices of export commodities decline was believed to have cyclical nature. Assuming an eventual rise in commodity prices and given the prevailing low real interest rate commercial banks and SSA governments were both attracted by external debt mechanism to finance massive budgetary deficits. In addition, SSA governments refused to devalue their currencies, they used tariffs, import controls and other quantitative restrictions to protect domestic markets, following importsubstituting industrialisation policies, rather than devaluing their currencies, to correct the distortions. Consequently, rates of exchanges became seriously over-valued, leading to a loss of markets and valuable foreign exchange earnings. This contributed to SSA countries’ balance of payments problems, by over-pricing exports and under-pricing imports. Public expenditure by African governments following increases in commodity prices during the early 1970s was maintained during all the decade, because the need of building independent states. The situation was reinforced by a second oil price shock, in 1979. The new loans were characterised by harder terms, with commodity prices continuously deteriorating, exacerbated by the recession in the industrial countries. It induced a rise in real world interest rate 32 , due to fiscal and tight monetary policy of the Reaganomics in the US. In fact, the adoption of tough anti-inflation monetary policy led to sharp rises in the dollar interest rate and in the dollar’s foreign exchange market value. Given that most of developing countries’ loans had been contracted at floating interest rates and that loans are denominated in dollars, the real value of the debt service increased substantially. In the meantime, non- concessional and private credits became the only option at SSA countries’ disposal. Short-term real interest rates in the United States rose from an annual average of -0.7 per cent in 1972-1975 to 5.0 per cent in 1980-1982. The index of the terms of trade of non-petroleum exporting developing countries fell from 110 in 1973-1975 (1980=100) to 94 in 1981-1983 and further to 84 in 1989-1990. The net flows of private capital declined from over US$ 70 billion in 1979-1981 to barely US$ 28 billion in 1985-1986, while capital flight from 13 32 By 1981 the real foreign interest rate was 17.4% compared to -17.9% in 1973. 123 highly indebted countries rose from US$ 47 billion at the end of 1978 to US$ 184 billion at the end of 1988. In sub-Saharan Africa excluding Nigeria, the net deterioration in the external financial situation from these three factors amounted to US$ 6.5 billion per annum over the period 1979-1981 to 1985-1987. These amounts, which take into account debt rescheduling but ignore capital flight, attained roughly one third of the total annual imports of goods and services of these countries in the early 1980s and about 45 per cent of average annual export earnings. In addition, policy mismanagement in African countries also explains the debt crisis. Budget deficits, overvaluation of the currency, capital flight and current account deficits have to be counted as the domestic sources of repayment difficulties. Inadequate time horizon (shortterm loans to support long-term investments), wrong investment, unproductive investments (arms) and capital flights contributed to exacerbate the situation. Both lenders (who were not cautious in their policy) and debtors’ governments were responsible for the crisis. In the 1980s, new funds were used in a more defensive manner than in the past: rather than being used for infrastructure, they were used to avoid social and political discontent due to cut in public expenditure and to repay arrears. Significantly, the proportion of non- concessional debt was rising faster than concessional, which increased the overall debt burden, and there was an increase in short-term debt, too. Originally, these debts were not considered part of the debt problem, assuming that, as traderelated credits, they would liquidate themselves in the normal course of events. Given the drastic reduction in imports, SSA countries had severe problems with financing their trade flows and an high proportion of short-term debts, also including accumulation of arrears of interest on public long-term debts, were converted to medium- and long-term debts. Since the half of the 1980s, SSA countries embarked in structural adjustment programs, but they faced declining commodity prices and the deterioration of terms of trade. The 1984-1985 period represented the turning point for SSA region crisis. In 1985 the debt service to export ratio reached the peak of 26.2 percent (compared to 6.8 percent in 1975 and 11.7 percent in 1980), and then, with the collapse in debt service capacity, this ratio was pushed to around 20 percent. Net transfers of debt (being net flows minus interest payments or disbursements minus total debt service payments), after having been positive, in 1985 became negative (-811.4 US$ million). A basic measure of the link between GNP and the debt-to-GNP ratio is the statistical correlation, which is a measure of the strength of association between two variables. Unlike regression, it is not necessary to define one variable as the independent variable and one as the dependent variable 33 . 33 The correlation coefficient is a number that varies between -1 and +1. A correlation of -1 indicates there is a perfect negative relationship between the two variables, with one always decreasing as the other increases. A correlation of +1 indicates there is a perfect positive relationship between the two variables, with both always increasing together. A correlation of 0 indicates no relationship between the two variables. The formula of the correlation coefficient is: where covariance is divided by the product of the stardard deviations. 124 The (Pearson product moment) correlation coefficient for the entire period from 1975 to 1999 is equal to –0.636, and indicates there is a strong negative relationship between economic development and debt burden in SSA region, with one always decreasing as the other increases. But, if we split the entire series into two sub-periods, and we consider the first period of “sustainable indebtedness” (1975-1983), in this case we have a significant positive association (+0.487). In the remaining period after the turning point (1986-1999), we have a significant negative association (-0.794), which is very high if we drop out the period of implementation of the recent multilateral initiative to cancel debt (Heavily Indebted Poor Countries Initiative), and consider the sub-period 1986-1996 (-0.843). This result confirms the existence of a complex, non-linear relationship between debt and development. Given particular conditions, the relation can be positive, whereas it becomes negative if we assume different context. Again, the crucial issue is how to define the correct connection between politics and economics, in order to escape from over-simplification on the links between debt and development and to reconstitute social science, primarily aimed at overcoming the artificial distinctions between economic-political, on the one hand, and domestic-international, on the other. GNP per capita Turning point 1983 1986 1996 1975 n l a ti o e r r co N egat iv e c o rre la t io n -0 i ti v e 487 s o P 0. .843 + Debt-to-GNP ratio This picture can be consistent with the existence of debt overhang problem, provided that a large external debt overhang – corresponding to the 1984-85 turning point, where the curve turns down and the additional stock of debt implies disincentives to development - affects negatively economic development. Moreover, the increase of positive real interest rate and the capitalisation of amortisation and interest payment through the Paris and London clubs rescheduling had also started pushing the debt stock upward. African economies became extremely indebted by the 1990s and dependent on external finance for securing imported intermediate inputs. Northern protectionism for agricultural products implied an high cost for badly diversified economies, such as the SSA ones. In the 1990s, multilateral development banks - particularly the World Bank, IMF and African Development Bank - replaced commercial banks, to facilitate their bail-out, and to implement adjustment program, thus becoming the main creditors of SSA governments. The international financial institutions enjoyed “preferred creditor” status and debts to these institutions could not be rescheduled: they had to be serviced, irrespective of what was 125 happening elsewhere in the domestic and international economy34 . In fact, several SSA countries are net creditors to the IMF. Nowadays, ten years later, over 80 percent of the SSA debt is a public sector liability, derived from public sector borrowing and public assumption of past private sector debt. Among the official creditors, multilateral institutions have gained importance (accounting now for 1/3 of long-term debt and ½ of long-term debt service, and being 29.9 percent of total debt, compared to 12.4 percent in 1980). Less than 20 percent are short-term debt accounts. From this rapid historical review, it derives the importance of several factors, both domestic and external, which determined a severe debt problem. It seems to confirm the idea discussed by Keynes on the German war reparation problem, even then the causes of debt crisis were clearly different. Debt crisis, as managed during the 1970-90s, has created a budgetary problem, as SSA governments need no only to cut budgetary deficit by reducing the level of domestic absorption, but also to extract the additional domestic savings out of current national income in order to secure the necessary domestic resources for servicing external debt. Therefore, a budgetary surplus needed first to be created, reducing the level of investment, cutting government social expenditure and restricting the expansion of the money supply (measures induced by structural adjustment programs, financed by the World Bank and IMF). In a short-term horizon these measure seemed more feasible than expanding the tax base, restructuring tax rates and improving system of administration, all of them being measures, which take time to implement. The implemented measures affected domestic incomes, prices, wages and savings, creating serious economic, social and political problems 35 . Moreover, as creditors do not accept domestic currencies, given the asymmetrical relationship between a debtor country and its creditors, the mobilised domestic resources have to be converted into foreign exchanges, and this creates the transfer problem. The SSA ability to increase the foreign exchange earnings depends basically on the rate of growth of exports, which in turns depends on the supply performance – very low - and on the creditors’ availability to accept SSA imports – very low -. Unfortunately, inadequate supply and demand prevented SSA countries from solving the transfer problem, particularly in the given context of the 1980s and 1990s, when all SSA countries were required to increase exports at the same time, increasing competition with each other and reducing the space for expansion of intra-regional trade. Anglo-American neo-liberalism has tended to prevail, based on a set of institutions and practices, which promote a Social Darwinist reconfiguration of priorities, policies and outcomes real priority. The Structural adjustment programmes were the mainstream, promoting the liberal vision of the Washington Consensus; and according to it a fully integrated and self-regulating developing economy must be based on international links (FDI inflows). No real priority was given to the building of an efficient, strong, redistributive revenue system: the third-gap of development financing problem received very marginal attention. Given this long history of domestic and external problems, there are three contending ideological explanations of the current African crisis, which can be derived from the background theoretical analysis presented in Part I. These explanation provide different political solutions. 34 In 1996, the HIPC Initiative, described in chapter seven, changed significantly the situation, introducing the possibility to cancel debt owed to multilateral banks. 35 The same effects described by Keynes as referred to the reparation of Germany’s war debts. See J. M. Keynes (1929), op. cit. 126 The Washington Consensus approach is based on the World Bank's Agenda for action (1981) 36 report saying that the main determinants of African crisis are domestic: underdeveloped human resources, political instability, inadequate institutions, geography and population growth. External shocks - a rise in oil prices in the periods 1973-74 and 1978-80 and deterioration in SSA terms of trade – are not the crucial direct cause, which is the decline in the volume of exports, volumes, rather than declining prices, due to structural changes in the composition of world trade, inadequate supply capacity and Northern countries’ trade restrictions and agricultural subsidy policies. Orthodox macroeconomic management, based on export promotion and adjustment, represents the road to economic recovery. An opposite structuralist view is represented by the Economic Commission For Africa (1989) 37 , saying that there are some interlinked constraints, being the inadequate economic and social infrastructure, research capability, technological know-how and human resource development. Inflation, balance of payments deficit, debt burden and instability of exports, in a context of an excessive outward oriented industrialisation policy, strongly interact. It created a structural external dependence and vulnerability of SSA, as a direct continuum with the colonial experience, still producing raw materials and agricultural goods for Europe. Exogenous factors, such as terms of trade and world interest rates, have caused the crisis. Policy implications, which can be derived from this explanation, are improving income distribution, focusing on the basic needs, reducing import dependence. Opposed to both the previous approaches, a post-Marxian approach, based on the Wallerstein idea of world-economy, says that SSA crisis is part of long-term secular effects of imperialism and of the world capitalist crisis. From this point of view, free market international system is inherently a source of exploitation of poor countries. Thus, opposite to the other approaches, free market mechanism, rather than being itself instrument of best resource allocation and income distribution within a short-term perspective (Washington Consensus) or being the fundamental instrument to be adequately and structurally transformed in a long-term perspective (ECA), it is source of systemic problems. History and data seem to demonstrate that probably it is the combination of internal and external factors (i.e. a combination of all these different explanations) which has given the debt crisis in SSA countries its peculiar characteristics. Some additional useful information38 : - total external debt grew nearly 20 fold, from US $12.6 billion in 1971; - the basic component is now long-term debt outstanding, owed to official creditors; - bilateral debt is the most important component, followed by multilateral debt. Private inflows are the residual part, and with a declining trend; - a larger share of the official debt is on concessional terms; - since the 1980s, the use of IMF credit became an important support to structural adjustment; - capitalisation of interest and principal arrears constitute a quarter of the external debt, confirming the fact that external debt is now a self-reproductive mechanism. 36 World Bank (1981), Accelerated Development in Sub-Saharan Africa: An Agenda for Action, Washington D.C. 37 UNECA (1989), African Alternative Framework to Structural Adjustment Programs for Socio-Economic Recovery and Transformation (AAF-SAP), Addis Ababa. 38 These figures are useful information, but their interpretation must be cautious, as: (i) problems of data coverage and reliability, (ii) differences among SSA countries, (iii) there have been changes in the value of the dollar, (iv) it is not always clear whether data on debt payments represent scheduled (ex ante) or actual (ex post) payments. 127 7. – Fiscal component of external debt crisis in Africa External debt problem can be due to fiscal reasons. The fiscal nature of the problem is evident in Sub-Saharan Africa. Hjertholm 39 looked at the external debt-financed government spending experience of SSA, confirming the existence of an abuse of foreign borrowing to finance large fiscal deficits. He stressed that SSA countries: • inherited very weak fiscal policy and administration from colonial period; • were unable to generate domestic revenue, due to structural economic (trade-dependence) and political considerations; • experienced very high level of current (urban subsidies) and capital (infrastructure) expenditure to create political consensus and economic development; • exploited the international commodity boom (in the 1970s) to expand external debt as the main source of deficit financing, producing favourable expectations among the creditors; • when commodity prices fell sharply (in the 1980s), the rigidity of government expenditures increased fiscal deficit and produced further external debt (particularly multilateral debt); • central governments guaranteed the repayment of debt acquired by para-statal enterprises, the economic failure of many “productive” public sector investments and the high share of military expenditure (15 percent of total government expenditure) compromised the repayment of debt. The aggregate picture of SSA region masks the differences in individual countries, nevertheless there is a common characteristics, which can be stressed. High levels of external debt were associated to high and permanent levels of overall deficit. And central government debt represented the bulk of this deficit, confirming the existence of a vicious circle between development, external debt and public revenue. The more developed is a country, the more capacity this country has to generate revenue, then the more domestic resources are available to finance development and the less external burden is needed. Tab. 7 - SSA Overall surplus/deficit, incl. all grants, No. Of countries Deficit, as share of GDP (%) surplus n.a. 4 3 7 12 6 3 3 13 10 7 1 >10% >5% >2.5% >0.1% 1986-91 7 15 10 1992-95 8 14 1996-99 6 9 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. In fact, at a more disaggregated level, it is quite clear that the revenue problem is more evident for the sub-group of SSA countries being classified as “Severely indebted lowincome countries” and, within this group, for the poorest non-oil-exporters countries. 39 P. Hjertholm (1997), An Inquiry Into the Fiscal Dimension of External Debt: The Case of Sub-Saharan Africa, PhD dissertation, Institute of Economics, University of Copenhagen. 128 Tab. 8 - Government revenues, excl. all grants as % of GDP 1985 1990 1995 1999 Burkina Faso 9,8 12,3 10,8 16,5 Burundi 13,4 15,1 20,1 17,9 Cameroon 21,0 14,3 12,9 15,3 Central African Republic .. 10,7 9,2 9,2 Comoros 10,5 17,0 17,4 10,9 Congo, Dem. Rep. .. .. 13,2 12,8 Ethiopia 16,7 17,7 17,5 19,5 Guinea .. 15,8 10,9 10,3 Guinea-Bissau 7,6 18,7 12,7 18,4 Madagascar 12,9 12,0 8,5 11,9 Malawi 22,7 21,0 17,8 16,6 Mali 12,0 17,6 12,9 17,8 Mozambique 9,9 12,8 11,3 10,5 Niger 10,8 10,2 7,2 6,8 Rwanda 9,9 10,1 6,9 9,2 Sierra Leone 5,2 9,2 9,5 10,6 Sudan 9,4 6,3 8,7 8,4 Tanzania 17,4 14,7 11,0 10,6 Uganda 9,1 6,8 9,8 10,9 Zambia 21,9 20,3 19,9 17,6 Source: World Bank (2001), World Bank Africa Database 2001, Washington D.C. The table shows twenty poor SSA countries with a permanently weak fiscal position, due to their inadequate revenue performance, which means weak debt servicing capacity. Moreover, considering the government revenue indicator as compared to a common measure of debt distress, that is debt service ratio (with an export-based denominator), some interesting points emerge. Revenue based indicator of debt distress appear higher than the export-based indicator, confirming that government revenue represented the main constraint in SSA region. And, as SSA region’s debt is mainly public debt, the revenue constraint is clearly more stringent, whereas export earnings facilitate the servicing of both public and private debt and represent an added public debt service capacity solely through their conversion into government taxation. Conversion that is not complete and transparent, given the administrative weakness of African countries and discrepancies of data. In sum, revenuebased debt indicators seem to capture a significant component of external debt burden, at least as important as the export-based indicators, which are commonly used. A relationship that should be more deeply empirically investigated, as well as the relationship with indicators referred to the other side of the public budget, that is government expenditure. 8. – Different components of foreign capital flows to Africa External debt flows to SSA countries are not the only channels of international financial resources to Africa. And capital flows are not all the same. We have to consider several dimensions: equity versus debt, short term versus long term, public versus private. Considering total external debt, grants, foreign direct investment, portfolio investment and remittances - both private and official finance – as total external financial inflows, we can compare this aggregate to gross domestic savings (calculated as the difference between GDP and total consumption). 129 SSA's Financial aggregates (% of GNP) 30 total external capital inflows 25 Gross dom. savings 20 15 10 5 19 98 19 96 19 94 19 92 19 90 19 88 19 86 19 84 19 82 19 80 19 78 19 76 19 74 19 72 19 70 0 During the 1970s, domestic savings represented a growing component to be mobilised to finance investment in SSA’s economies, and they reached 29.7 percent of GNP in 1979, whereas external finance was much lower and represented 20.0 percent of domestic savings in 1979. During the 1980s, this trend has abruptly changed: ruined by the dramatic slowdown of African economies, domestic savings dropped suddenly and arrived at 10 percent of GNP in 1992. In the meantime, external finance was still growing and arrived to represent a very high level of 9.8 percent of GNP (equal to 97.6 percent of domestic savings). Thereafter, domestic savings went up again but in the last two years both domestic and external flows have dramatically gone down. This picture is different from the South African experience: domestic savings were never lower than 17 percent of GNP and external finance was never higher than 30 percent of domestic savings. In general terms, in SSA as a whole external finance flows represent a relevant component in relation to GNP and domestic savings. And Feldstein notes that the close association between foreign inflows and domestic investment suggest a lack of deep integration in the financing of investment 40 . But a more precise analysis requires taking in accounts the differences across types of flows (long-term and short-term or private and official capital flows) and countries. Intuitively, the net inflows of investment to acquire a lasting management interest in an enterprise operating in an economy other than that of the investor, that is the sum of equity capital, reinvestment of earnings, other long-term capital, and short-term capital as shown in the balance of payments is one thing. And we call it foreign direct investment. External debt, official aid, portfolio investment and current transfers by migrants are quite different things. These differences are relevant in terms of impact on development. 40 M. Feldstein (1994), “The Effects of Outbound Foreign Direct Investment on the Capital Stock”, NBER Working Paper, N. 4668, Cambridge. 130 We can consider a rough picture of all these foreign capital inflows to SSA. Foreign direct invest. 11.000,0 Foreign capital inflows to SSA (Net, cur US$ mill.) Official grants, excl tec. coop. Workers' remittances (Credit) Portfolio invest. excl LCFAR 9.000,0 Total debt flows 7.000,0 5.000,0 3.000,0 1998 1996 1994 1992 1990 1988 1986 1984 1982 1980 1978 1976 1974 1972 -1.000,0 1970 1.000,0 In terms of flows, from the 1970s external debt has represented the main source of international capital to SSA up to 1996. In 1996, the international financial institutions launched an important initiative to reduce the indebtedness of poor countries (the HIPC initiative, analysed in chapter 7), which has a direct negative consequence, in terms of dramatic reduction of new lending. What seems to be a great worry is the sudden and parallel tremendous decline of aid during the last five years. Aid represented the main source in the 1990s and is widely considered a more adequate alternative to debt for poor countries. Quite interesting is the FDI dynamics, with a recent counter-trend compared to aid and debt. Obviously, referring to private flows, it is important to consider the between-countries differences. Remittances represent a growing, stable source of capital inflows, whereas portfolio investments have always been very marginal in SSA’s case and it reflects negative investor sentiment. How should we interpret the recent upsurge in FDI? It is true that, at a very simplistic level, it could be argued that FDI is a good capital flow from the point of view of the potential vulnerability of the capital importing countries. It represents equity rather than debt; it is long term rather than short term, it is associated with increased domestic capital formation, it tends to go to the private sector. Thus, it is common to attribute crises to short-term and debt capital inflows, while FDI is seen as a safer form of finance. In this case ,looking at the picture of foreign capital inflows to SSA, we should appreciate the recent reversal of course 41 . But, at a deeper level, FDI means different things, and different forms of FDI are likely to have rather different implications for the safety of capital inflows. Debt liabilities are not inherently less safe than FDI. Fernández-Arias and Hausmann note that the common view is inappropriate as FDI is not a physical asset of a firm, but only (and simply) one of its 41 As expressed by the World Bank: “FDI also is less subject to capital reversals and contagion that affect other flows, since the presence of large, fixed, illiquid assets makes rapid disinvestment more difficult than the withdrawal of short-term bank lending or the sale of stock holdings”. See World Bank (1999), “Foreign Investment Resilient in the Face of Financial Crisis”, Chapter 3, Global Development Finance 1999, Analysis and Summary Tables, Washington D.C. 131 liabilities42 . And they conclude that the share of FDI in total capital inflows is not a measure of anything good happening in the economy, in fact in rich countries lower growth prospects and higher risks lead companies to prefer more equity and less debt in the composition of their capital. Also, poorly functioning debt markets can make FDI a more efficient way to access capital. In all of these cases, the fact that the share of FDI in capital inflows is rising is not bad in itself, but is instead an optimal response to a deteriorating environment. In order to analyse FDI to SSA, we can start from the Hausmann’s approach43 and look at two concepts, and consider both total external financial inflows as a share of GNP (i.e. how large is the flow of liabilities to foreign players relative to GNP) and, second, the share of FDI in those liabilities. We can also add a time dimension and draw the line of historical trend (from 1970 to 2000). Thus, we can look at the ratio of FDI to GNP, that is the product of the first two ratios (= [total external financial inflows / GNP] [FDI / total external financial inflows] = [FDI / GNP]). We can decompose the share of FDI to GNP as a consequence of a volume effect (reflected in the total flow of external capital), and a composition effect (the proportion of FDI). Foreign capital flow vs. GNP: SSA total external capital inflows/GNP (%) 12,0 10,0 8,0 6,0 4,0 2,0 0,0 4,5 4,6 4,7 4,8 4,9 5,0 GNP (Log) 5,1 5,2 5,3 5,4 The figure shows that external capital liabilities as a share of GNP are low in SSA countries, with flows reaching no more than 10 per cent of GNP. In East Asia and East Europe they reach 15 per cent; in Latin American countries they reach 22 per cent; in developed countries they are much higher. The effect of debt (and aid) reduction is a drastic drop of total capital inflows. Hence, we find that the proportion of FDI clearly reflect this crisis in new lending, and the story is reversed. The flow of FDI represent a low percentage of the total external capital 42 E. Fernández-Arias and R. Hausmann (2001), “Capital Inflows and Crisis: Does the Mix Matter?”, Development Centre Seminars, Foreign Direct Investment Versus Other Flows to Latin America”, OECDIADB, Paris. 43 E. Fernández-Arias and R. Hausmann (2001), “Foreign Direct Investment: Good Cholesterol?”, Development Centre Seminars, Foreign Direct Investment Versus Other Flows to Latin America”, OECD-IADB, Paris. 132 ionflows in SSA countries, until the recent years, when it goes up to 80 per cent. And the most important problem is that FDI continue to target a narrow range of SSA countries and sectors, with extractive industries dominating, though there has been interest in telecommunications and tourism. Composition of External capital inflows vs. GNP: SSA 100,0 FDI/total external capital inflows (%) 90,0 80,0 70,0 60,0 50,0 40,0 30,0 20,0 10,0 0,0 4,5 4,6 4,7 4,8 4,9 5,0 GNP (Log) 5,1 5,2 5,3 5,4 Then, we can consider the flow of FDI as a share of GNP. In SSA there is a growing trend in the recent years. And the figure shows that the recent experience has been different. But we have to consider that Latin America and the industrial countries share a similar reatio of slightly over 7 per cent, followed by East Asia with 6.5 per cent. Africa and Asia are still below 4 per cent of GNP. FDI vs. GNP: SSA 4,0 3,5 FDI/GNP (%) 3,0 2,5 2,0 1,5 1,0 0,5 0,0 4,5 4,6 4,7 4,8 4,9 5,0 GNP (Log) 5,1 5,2 5,3 5,4 In sum, total capital flows tend to increase with the level of development. But the share of those flows that take the form of FDI tends to decline with the level of development. The ratio of FDI to GNP is high in industrial countries because it is a small share of a very large 133 total volume of external capital inflows. In SSA the ratio of FDI to GNP is still low, because low volumes of total external inflows. From this rough picture, it derives an important aspect to be investigated. The increase of the share of FDI is not necessarily a consequence of a general improvement in the perception of development prospects and institution building. Quite the opposite, it can be an indication that markets are working badly, the institutions are not accountable and the risks are high. In the context of debt relief, the lenders can be reticent to extend loans. Provided that portfolio investments are very marginal and that workers’ remittances have their own trend, what about the interaction between the major inflows, debt, aid and FDI? Their relationship does not seem to be linear and probably the effect on development depends on their inter-relationship as well. The effects of external capital inflows on development cannot in general be derived by their statistical definition (debt vs. equity, private vs. official source) but on the basis of other considerations as well (such as currency and maturity in the composition of liabilities, complementarity or substitutability with other sources of finance). The econometric analysis on the relationship between debt and development should consider this reciprocal effects and control for other forms of foreign capital inflows. We think that the impact of various classes of international capital inflows cannot be assessed by looking at each flow separately. 9. – Multidimensional nature of African problems African case demonstrates that external debt problems are strictly linked to other aspects of external relations (trade and aid) as well as to domestic problems (fiscal and savings), directly interacting with poverty reduction policies. The coherence of policies matters and it calls for political will by all partners. During the 1980s, 37 countries belonging to SSA underwent at least one adjustment programmes, due to their external debt crisis. The social, political and economic impacts of these programmes are direct effects of debt crisis. External debt is not only a financial and economic problem, but also a social and political problem. Given these premises, we have found some points to be underlined. First, as we repeatedly stressed and the following appendix tries to demonstrate, specific context does matter, thus every country has its own different profile in terms of external relations, domestic problems, and poverty situation. Second, poverty reduction, in the context of sustainable development, remains a major challenge in Africa. Poverty is not a static condition among individuals, households, regions or countries. Instead, it is recognised that while some individuals or households are permanently poor, other become impoverished. The dynamic concept of vulnerability must be used to understand these processes of change. Poverty encompasses different dimensions of deprivation that relate to human capabilities including consumption and food security, health, education, rights, voice, security, dignity and decent work. Despite wide consensus on the importance of poverty reduction strategies, an enormous gap remains between rhetoric and practice. Social and economic inequality within and between nations is a relevant obstacle to sustainable poverty reduction. And, in spite of international efforts and Africa’s urgent needs, performance still falls far short of goals. Third, in terms of future perspectives, Africa confirms to be the region being much far from achieving the seven International Development Goals (IDGs). Again domestic and external components seem to be crucial to address these targets. Fourth, external debt is a big problem in SSA, and it is related to some structural weaknesses of the balance of payments. 134 Fifth, external debt is not a new problem for Africa. We found some historical structural components of this crisis. Sixth, linked to this historical problem and to the external face of debt, being represented by the balance of payments problem, there is a direct link to the fiscal deficit problem. At the end, the picture is more complex than it may appear. External debt problem is not a financial problem in itself; it is a main component (both source and effect) of the multidimensional problem of development in Africa. It is not useful to separate domestic and external causes as opposite determinants. The reality is the interaction between all these factors. Correctly, Rodrik 44 argued that domestic social conflicts are a key to understanding why growth rates lack persistence and why so many countries have experienced a growth collapse after the mid-1970s. He emphasises, in particular, the manner in which social conflicts interact with external shocks on the one hand, and the domestic institutions of conflict management on the other. And empirical methods are used by Rodrik to investigate these relationships. Countries that experienced the sharpest drops in growth after 1975 were those with divided societies (as measured by indicators of inequality, ethnic fragmentation, and the like) and with weak institutions of conflict management (proxied by indicators of the quality of governmental institutions, rule of law, democratic rights, and social safety nets). Empirical methods have been widely used to investigate the complex relationship between external finance (both aid and debt) and development. Chapter 5 will present a survey of this literature, whereas chapter 6 will present an econometric application to analyse this issue. 44 D. Rodrik (1997), “Where did All the Growth Go?: External Shocks, Social Conflict and Growth Collapses”, Kennedy School, Harvard University, mimeo. 135 Bibliography N. Bhinda, J. Leape, M. Martin and S. Griffith-Jones (1999), Private capital flows to Africa: Perception and Reality, Forum on Debt and Development, The Hague. P. Cashin and C. Pattillo (2000), Terms of Trade Shocks in Africa: Are They Short-lived or Long-lived?, IMF Working Paper, N. 72, Wahington D.C. C. Chang, E. Fernandez-Arias, L. Serven (1998), Measuring Aid Flows: a New Approach, Washington D.C. DAC (1996), Shaping the 21st Century: The Contribution of Development Cooperation, OECD, Paris. M. Feldstein (1994), “The Effects of Outbound Foreign Direct Investment on the Capital Stock”, NBER Working Paper, N. 4668, Cambridge. E. Fernández-Arias and R. Hausmann (2001), “Capital Inflows and Crisis: Does the Mix Matter?”, Development Centre Seminars, Foreign Direct Investment Versus Other Flows to Latin America”, OECD-IADB, Paris. A. Geda (2001), “Debt Issues in Africa: Thinking Beyond the HIPC Initiative to Solving Structural Problems”, WIDER Conference on Debt Relief, Helsinki, August. D. Gordon and P. Spicker (eds) (1999), International glossary of poverty, Zed Books, London. A. J. M. Hagenaars (1986), The Perception of Poverty, North-Holland, Amsterdam. IMF (1999), Direction of Trade Statistics, Washington D.C. J. M. Keynes (1929), “The German Transfer Problem”, The Economic Journal, Vol. XXXIX, March. A. Sen (1981), Poverty and Famines: An Essay on Entitlement and Deprivation, Clarendon Press, Oxford. A. Sen (1979), Collective Choice and Social Welfare, North-Holland, Amsterdam. UNCTAD (2000), The Least Developed Countries. 2000 Report, Geneva. UNCTAD (1999), Foreign Direct Investment in Africa: Performance and Potential, UN, New York. UNDP (2001), Human Development Report 2001, Oxford University Press, New York. UNECA (2000), Globalization, Regionalism and Africa’s Development Agenda, Paper Prepared for UNCTAD X February 12-19, 2000, Bangkok, Thailand. UNECA (2000/b), Economic Report on Africa 1999: The Challenges of Poverty Reduction and Sustainablity, Addis Ababa. UNECA (1989), African Alternative Framework to Structural Adjustment Programs for Socio-Economic Recovery and Transformation (AAF-SAP), Addis Ababa. H. W. Watts (1968), “An Economic Definition of Poverty”, in DP Moynihan (ed.), On Understanding Poverty, Basic Book, New York. World Bank (2001/b), African Development Indicators 2001, Oxford University Press, Washington DC. World Bank (2001), World Bank Africa Database 2001, Washington D.C. 136 World Bank (1999), “Foreign Investment Resilient in the Face of Financial Crisis”, Chapter 3, Global Development Finance 1999, Analysis and Summary Tables, Washington D.C. World Bank (1981), Accelerated Development in Sub-Saharan Africa: An Agenda for Action, Washington D.C. World Bank, African Development Bank, ECA, OAU and African Economic Research Consortium (2000), Can Africa claim the 21st Century, Addis Ababa. 137 5. Review of econometric literature on the linkages between foreign capital inflows and development There are a large number of empirical studies that investigated the most relevant determinants of a Ldc’s development rate. The role of foreign capital has been explored in this context. However, this literature, mainly based on the usage of comparative crosscountry analysis, has not adequately explored: (i) the long-term nature of the relationship between foreign capital and development, (ii) the role of different conditions between countries within the same geographical region (which we stressed through the “flower” ideogram, in chapter 4), (iii) the importance of social, political and institutional conditions, (iv) the difference between the level and evolution of foreign capital intensity, which implies that case-specific features may overwhelm any mechanical link between the two. (v) the implication for sustainability of different nature and composition of foreign debt (short-term or long-term debts, its maturity and grace periods, the level of interest rates, the structure of interest and principal arrears, arrears to actual payments ratio, its frontloading,…), in fact the total debt outstanding involves debt of varying maturities, varying interest rates, and varying grace periods; the debt service owed in any given year thus is only loosely connected to the debt outstanding, (vi) the different impact of the various types of capital inflows (foreign debt, aid, foreign direct investment, portfolio and equity flows, workers’ remittances) on development, and particularly the relationship between aid and debt, (vii) the fiscal dimension of external debt. Moreover, the bulk of this literature studied the debt experience of Latin American countries. If we are interested in SSA, then we have to consider a lot of other studies, which investigated another relationship that is between aid and development. In fact, as SSA foreign debt is basically concessional and owed to official lenders - differently from Latin American debt -, most of the empirical studies on SSA were referred to the effectiveness of foreign aid, foreign debt being implicitly assumed as part of it. Much of the emphasis in the literature on fiscal implications of external inflows in SSA countries has been on the role of external aid, rather than debt1. Measured relative to recipient GNP, the median value of aid to African countries now stands at nearly 10 times the amount received by Western Europe under the Marshall Plan. This emphasis reflects the historical importance of aid relative to external debt in these poor countries, compared to the case of Latin American economies2. The literature has been dominated by cross-section studies using single-equation estimation 1 P. Cashel-Cordo and S. G. Craig (1990), “The Public Sector Impact on International Resource Transfers”, Journal of Development Economics, N. 32; S. Devarajan., A, S. Rajkumar and V. Swaroop (1998), “What Does Aid to Africa Finance?”, African Economic Research Consortium (AERC), Nairobi, mimeo; T. Feyzioglu, V. Swaroop and M. Zhu (1998), “A Panel Data Analysis of the Fungibility of Foreign Aid”, World Bank Economic Review, N. 12.; and I. N. Gang and H. A. Khan (1991), “Foreign Aid, Taxes and Public Investment”, Journal of Development Economics, N. 24. 2 As the external debt burden of African countries has become considerable, however, increasing attention is being paid to implications of debt for the these economies. In addition to the finding that the debt burden may have reduced economic growth. See A. K. Fosu (1999), “The External Debt Burden and Economic Growth in the 1980s: Evidence from Sub-Saharan Africa”, Canadian Journal of Development Studies, Vol. 20, N. 2; C. Lancaster (1991), “African Economic Reform: The External Dimension”, Institute of International Economics, Washington, DC. 138 techniques, producing mixed empirical results. Empirical studies have failed to provide statistically significant and clear conclusions. What most of all these studies seem to reject is the complexity of the relationship debt/aid/Fdi-development, the ambiguous nature of external debt, which can be both positively and negatively related to development. When we offered a re-examination of the theoretical literature on development and debt in chapters 24, we have stressed this nature, and Barro and Sala-i-Martin (1995) and Aghion and Howitt (1998) remind us that the development process depends on an intricate range of interacting characteristics and dynamics. Nevertheless most of existing studies prefer to assume a strict, often linear, and ideological relationship to be tested - confirmed or rejected - between debt and development. Debt (or, separately, aid) is good or bad for development. If we proceed from an historical point of view, it is clear that, as time goes, empirical studies reflect the growing complexity of new development theories. Recent advances in growth theory have allowed more sophisticated empirical growth equations to be specified. The regressions include more and more variables, as well as the econometric techniques improve their quality. In other words, the history of econometric analysis is a mirror to the history of theoretical literature and reflects, given stronger techniques, the same disenchantment. Because theory is inconclusive, the relationship between foreign capital inflows and development becomes an empirical matter. But the empirical evidence does not seem to agree on whether the relationship is generally positive or negative. Our basic idea, and we test it in chapter 6, is that the assumption of such a simple relationship is wrong. In this chapter, we offer a review of those quantitative analyses, which are based on crosscountry (and not on single-country case) studies at macroeconomic level (rather than at micro- level). This review of methodological approaches is a preliminary step to present, in chapter 6, our econometric analysis of debt/aid and development relationship in SSA countries. This chapter reviews econometric applications and takes account of the main relations, which have been discussed. Our aim is to stress the importance of recent theoretical and empirical studies and to point out that foreign capital inflows (external debt and aid) seem to have positive but decreasing returns and start to have a negative effect on development after a certain level. We think that this level is not the same for different countries. Debt overhang exists, and a country can receive too much debt or aid and becomes dependent. We think that theory supports the idea of an inverted U-shaped relationship (a foreign capital inflow-Laffer curve) between concessional finance and development. And empirical evidence confirms it. But we also think that “how high flows should become to have a negative effect on development?” is a misleading question as historical, social, political, and economic contexts do matter as well. And that different kinds of foreign capital inflows interact among themselves in affecting development. 1. The links between national saving and foreign debt The links between savings and international capital inflows have been the core of the first analyses, based on the early Harrod–Domar traditional growth theory and saving-gap approach. Foreign capital was perceived as an exogenous increment to the capital stock of the recipient country. The implicit causal chain runs from foreign capital to growth via savings. A huge amount of savings is translated into higher investment (i.e. accumulation of physical capital), the investment function being governed by the acceleration principle. 139 The extent to which shifts in government saving induce offsetting changes in private saving has been a central issue in much of the 1960 and 1970 literature in macroeconomics. Rosenstein-Rodan (1961) affirmed that each dollar of foreign resources would induce an increase of one dollar in national savings and investment. Thus, foreign flows were not considered as a component of national income adding to both consumption and investment, and fungibility was not taken in account. Quite the opposite, Griffin and Enos (1970)3 said that capital inflows from abroad reduce the rate of growth of the receiving country. Weisskopf confirmed, analysing 38 developing countries during the 1950s and 1960s period, a negative impact of foreign resources on domestic saving4. Foreign capital inflows become a substitute for, instead of an addition o domestic savings. Chenery and Eckstein5, Landau6, Grinols and Bhagwati7 confirmed the negative correlation between foreign resource and domestic savings, implying severe critics to the gap theories. In this case, the foreign resources dependency is the mechanism described to explain the negative effect of foreign resources on domestic saving, which was assumed as dependent variable. Foreign resources may support both investment and consumption8, that is they can replace saving, which becomes a residual component. Given the income level, increasing consumption implies reduced savings9. Nevertheless, all these studies realised in the 1970s have simply showed the presence of correlation, without deriving any causal relationship. In fact, Morisset10 demonstrated that this correlation does not imply any causality: both national savings and foreign resources are affected by exogenous (interest rate on debt) and endogenous (public deficit, money supply) variables, and this creates the illusion of causal relationship between domestic savings and foreign debt. Thus, in broader terms, an issue that remains unresolved in the empirical literature is that of causality: external debt may cause policy changes (owing to conditionalities), but policy changes may also results in higher levels of debt, by increasing international confidence in the country’s economic prospects. The general relation assumed by this kind of analysis, combining the acceleration principle and the multiplier theory, is: St = a0Yt + b1FFt and Yt = d0 + d1FFt St = a0 d0 + a0 d1FFt + b1FFt = a0 d0 + FFt (b1 + a0 d1) 3 (5.1) (5.2) K. B. Griffin and J. L. Enos (1970), ibidem. T. E. Weisskopf (1972), “The Impact of Foreign Capital Inflow on Domestic Savings in Uncerdeveloped Countries”, Journal of International Economics. 5 H. B. Chenery and P. Eckstein (1970), “Development Alternatives for Latin America”, Journal of Political Economy, N. 78, 4. 6 L. Landau (1971), “Savings function for Latin America”, H. B. Chenery (ed.), Studies in Developmental Planning, Harvard University Press, Cambridge. 7 E. Grinols and J. Bhagwati (1976), “Foreign Capital, Savings and Dependence”, Review of Economic Statistics, N. 58. 8 Also Dornbush considered relevant the role of foreign finance in terms of consumption-smoothing, allowing the indebted country to keep its consumption and expenditures levels inaltered in the short term. See R. Dornbush (1983), “Real Interest Rates, Home Goods and Optimal External Borrowing”, Journal of Political Economy, N. 91, 1. 9 K. B. Griffin and J. L. Enos (1970), “Foreign Assistance: Objectives and Consequences”, Economic Development Cultural Change, N. 18, 3. 10 J. Morisset (1989), “The Impact of Foreign Capital Inflows on Domestic Savings Re-examined: The Case of Argentina”, World Development, N. 17, 11. 4 140 Where: St = savings a0Yt = the average (and marginal) propensity to save FFt = foreign flows (aid and external debts) All these studies used only aggregated data on foreign inflows, due to lack of data on different flows and their inability to separate the various aggregate foreign inflows measures. Hansen and Tarp (2000)11 compared 131 regressions, identified from 29 articles: they investigated a sample of 41 aid-savings regressions. Only one study reported an estimate of α 1 which is significantly greater than zero. Thus, the idea of a direct, positive, on a one-toone basis impact of foreign inflows on savings is not confirmed. But a negative impact is not confirmed either, and in only one study does aid lead to lower total savings. The overwhelming evidence from these studies is that aid leads to an increase in total savings, although not by as much as the aid flow. The more recent empirical studies indicate12 that: • an increase in public saving tends to be associated with higher national saving, rejecting the Ricardian assumptions (a rise in the budget deficit has no effect on the national saving rate: the so-called Ricardian equivalence); • increases in per capita income raises the private saving rate; • increases in foreign saving (measured by the current account deficit), affect national saving negatively, but only partially: this suggests that foreign saving is a substitute, less than perfect one, for domestic saving, and an increased availability of external financing support both consumption and investment and however the negative coefficient is less than one (i.e. less than proportional effect), implying that the crowding out effect is less than one to one; • reduction in high external debt burden increases national saving. This result can be due to the fact that if the public sector’s foreign debt deteriorates, the private sector will anticipate a significant increase in taxation in the future, and this substitution effect due to such expectations will reduce consumption. Thus, from this point of view, a high level of debt will reduce the national saving rate. In every case, there are some statistical problems linked to his kind of analysis. Ø There is the problem of assuming as parameters what are in reality variables (for example, propensity to save). Ø It is also important to consider the fact that important exogenous shocks are removed by analysis. Ø It is not easy to correctly identify the exact amount of foreign resources (for exemple, there is the capital flight phenomenon) or to combine different forms of flows (grants, loans, investment)13. 11 H. Hansen and F. Tarp (2000), “Aid Effectiveness Disputed”, in F. Tarp (ed.), Foreign Aid and Development: Lessons Learnt and Directions for the Future, Routledge, London. 12 P. R. Agénor (2000), The Economics of Adjustment and Growth, Academic Press, San Diego. 13 Sengupta proposed an easy weighting system to determine the length or time horizon to reach independence from different forms of foreign flows: dependency vanishes faster in the case of grants, it takes longer in the case of debt and the longest dependency in case of investment. This is simply due to the fact that grants are free of interest payments, and investment requires an interest rate (equal to loans) plus a risk premium, which makes it the most expensive. Obviously, this simplification assumes 141 In fact, Michaely14 showed the difficulty in estimating a dependency indicator (equal to the foreign resources divided by total resources). Bhaduri15 stressed the nature of complex system: a dependency effect linking foreign resources and domestic saving can not be reduced and progressively removed through the correct usage of funds to support productive investment, if a foreign exchange gap exists as well. He reflected the fact that the original Harrod–Domar model, assuming only a savings constraint on growth, was expanded in the sixties in the influential Chenery and Strout (1966) two-gap model. In this case some bifurcations are possible, and both domestic structure and trade balance should be addressed. If the trade gap is the larger of the two, actual savings are supposed to fall short of potential savings, and if the prospective savings gap is the larger, actual imports will be greater than those needed for growth (see Chenery and Eckstein 1970). Robinson (1971) introduced the trade balance in a cross-country growth regression. He found that the trade balance variable was highly significant and that foreign exchange can operate as a limiting factor for growth, lending support to the two-gap model. It represented a step towards the definition of a more complex set of relations, which link foreign debt and development. 2. The links between investment and foreign debt Foreign inflows can raise investment rates or, alternatively, they may be used to raise current consumption, potentially reducing saving. In the past decades, there has been a longstanding interest in the extent to which the resource inflows are invested or consumed16. The decision to invest in Ldcs has been shown to depend on a variety of factors. An ambiguous effect seems to have foreign debt on national private investment, because from one side, by increasing the fiscal deficit and reducing available credit, it crows out private capital formation, but from another side public investment (for example, in infrastructure projects) may be complementary to private investment. Considering the relationship between investment and foreign debt, the empirical studies assumed that a large ratio of foreign debt to output has an adverse effect on investment, against the hypothesis of positive effects. Again, he idea was to test a linear, direct and clear relationship between debt and investment, rather than assuming a “complex” nature of interactions. The introduction of investment in econometric analysis did not change the focus on capital accumulation - investment is the major direct determinant of growth - and it resulted consistent with both the Harrod–Domar and Solow growth models. Hansen and Tarp (2000) identified a sample of 72 cross-country foreign inflows-investment regressions. These equations considered domestic savings and various aggregates of foreign inflows as separate determinants of investment, which was assumed as dependent variable. that there is no difference in the way how the three considered forms of foreign finance affect the savings/income ratio. See A. Sengupta (1968), “Foreign Capital Requirements for Economic Development”, Oxford Economic Papers. 14 M. Michaely (1981), “Foreign Aid, Economic Structure and Dependence”, Journal of Development Economics, N. 9. 15 A. Bhaduri (1987), “Dependent and Self-Reliant Growth with Foreign Borrowing”, Cambridge Journal of Economics, N. 11. 16 M. Obstfeld (1998), “Foreign Resource Inflows, Saving, and Growth”, in K. Schmidt-Hebbel and L. Servén (eds.), The Economics of Saving and Growth: Theory, Evidence and Implications for Policy, the World Bank and Cambridge University Press, Cambridge. 142 In such a way, the specification of the linear regression included a more detailed set of explanatory variables: It = b0+ b1St + b2ODAt + b3FDIt + b4ETDt+ b5PEIt (5.3) Where domestic savings (=St), aid (=ODAt), foreign direct investment (=FDIt)17, total external debt (=ETDt), and Portfolio and equity investment (=PEIt) are separate regressors18. These regressions suggest that there is a positive link between foreign inflows and investment whenever there is a positive link between savings and investment. In fact, in the majority of their regressions, when the foreign inflows coefficients are insignificant, then the savings coefficient is insignificant, too (b1=0). It should be correct to conclude that foreign inflows do not work only when savings do work (b1≠0). In these regressions the real paradox is that savings does not seem to promote growth, giving rise to the suspicion that the underlying structural model is not appropriate. Quite recently, Bosworth and Collins (1999)19, attempted to evaluate the implications of capital inflows for the recipient countries by developing a panel data set covering 58 countries over the period of 1978-95. FDI, portfolio and external debt result not significantly correlated with one another over time or across countries. The correlations are all low; indeed, the only statistically significant one (between FDI and external debt in the time dimension) is just 0.0920. It seems to confirm little evidence of complementarity between different types of international capital inflows. Soto21 analysed the effects of the different component of private capital inflows on the growth of 44 developing countries, during the 1986-97 period. The dependent variable is the change in the logarithm of GNP. This paper uses dynamic panel with yearly data in growth regression following the method proposed by Barro and Sala-i-Martin in their empirical research22. In fact, there are a certain number of “state” variables (in particular, the lagged level of GNP), which determine the initial conditions of the economy, and “control” 17 Recent literature on the potential role for FDI to raise growth through technological diffusion can be found in E. Borensztein, J. De Gregorio, and J. W. Lee (1998), “How Does Foreign Direct Investment Affect Growth?”, in Journal of International Economics, N. 45; G. M. Grossman and E. Helpman (1991), Innovation and Growth in the Global Economy, MIT Press, Cambridge. 18 This is the consequence of history. “Although the private international capital market revived in the late 1950’s, for LDCs capital flows remained almost entirely official. This remained the rule in the 1960’s, with the notable exceptions of a few dramatically successful countries” (A. O. Krueger [1987], “Debt, Capital Flows, and LDC Growth”, American Economic Review, Vol. 77, N. 2, May, p. 159). “For a number of years aid was calculated as the sum of official and private capital flows, although now these two items are listed separately” (M. P. Todaro [1989], Economic Development in the Third World, Longman, New York, p.481) and “...’aid’, a term which had been ambiguously used and sometimes actually applied to the entire flow, including private investment and export credits.” (W. T. Newlyn [1977], The Financing of Economic Development, Clarendon Press, Oxford, p.98). 19 B. Bosworth and S. M. Collins (1999), “Capital Flows o Developing Economies: Implications for Saving and Investment”, The Brookings Institution, Meeting Draft, March 15. 20 This correlation is due to the fact that, since the problem of foreign exchange shortages and external debt means that spare parts cannot be imported and therefore existing plants are forced to operate below capacity or are completely paralyzed, any attempt to increase the inflow of DFI depends on the solution of the debt problem. 21 M. Soto (2000), Capital Flows and Growth in Developing Countries: Recent Empirical Evidence, OECD Technical Paper, N. 160, Paris. 22 R. Barro and X. Sala-i-Martin (1995), Economic Growth, McGraw-Hill, New York. 143 variables, those which determine the steady state. The standard specification include oneyear lagged values of the growth rate, investment rate, government consumption, change in the logarithm of the terms of trade and the degree of openness to international trade. In other regressions, the investment rate is replaced by the national saving rate and the current account among the instruments, so that the effects of national saving and foreign saving can be distinguished. It is found that a negative sign is associated with national saving and the coefficient associated with the current account is not significant. The idea of decreasing returns on saving, which is the excessive saving hypothesis, is modelled by using a non-linear relationship between the saving rate and growth. Non-linearity indicates that beyond a certain threshold, additional savings can not be productively absorbed by the economy. A squared term for the saving rate is introduced The saving rate recovers its positive sign, and the squared term has a negative one, confirming the hypothesis. The categories of capital inflows introduced are FDI, portfolio equity flows, portfolio bond flows, bank credits and trade-related credits. All financial inflow variables are measured as a ratio to GNP. FDI and portfolio equity flows exhibit a robust positive correlation with growth. Portfolio bond flows are not significantly linked to economic growth. Bank-related inflows are negatively correlated with the growth rate. The degree of openness displays a significant and positive link with growth, confirming Frankel and Romer’s results, who found a positive correlation between trade openness and growth23. Government consumption presents a negative sign, which is in agreement with the hypothesis of Barro and Sala-i-Martin about the distortions introduced in the economy by the government intervention. The terms of trade present a very strong positive sign: the magnitude of its coefficient (+0.1) shows that a 10 percent increase of the terms of trade has a short run (= one year later) positive impact of 1% in the GNP per capita. Frankel and Rose24 find in a panel of 100 LDCs from 1971 to 1991 that a high ratio of FDI to debt is associated with a low likelihood of a currency crash. Hence, there is some presumption that since FDI is determined by long-term considerations, generates positive externalities and does not exert pressure on the real exchange rate. However, Reisen25 considered this optimism as misleading, as capital is fungible. The more recent empirical studies on debt and investment relationship, even though they may be affected by specification bias and spurious correlation results, indicate26 that debt service ratio has had a strong27, negative effect on private investment. Again, it may confirm the ambiguous effect of foreign debt on development, which shifts from being positive o negative when the debt burden and pressure becomes so disproportionately large, given the specific context of the debtor’s situation, to determine a debt overhang situation. 23 J. Frankel and D. Romer (1999), “Does Trade Cause Growth?”, American Economic Review, Vol. 89, N. 3. 24 J. Frankel and A. Rose (1996), “Currency Crashes in Emerging Markets: Empirical Indicators”, NBER Working Paper, N. 5437, Cambridge. 25 H. Reisen (1999), “Sustainable and Excessive Current Account Deficits”, in J. Gacs, R. Holzman and M. Wyzan (eds.), The Mixed Blessing of Financial Flows, Edward Elgar Publ., Cheltenham. 26 P. R. Agénor (2000), op. cit. 27 Using the standardised regression coefficients, which are unit free and measure the relative importance of the independent variables, the debt service ratio has the largest effect on private investment in poor countries, and also the ratio of external debt to exports has an adverse effect. 144 3. Empirical evidence of debt overhang The empirical evidence on existence of a debt overhang has been rather mixed. Claessens (1990) only finds a debt overhang for a very limited number of LDCs. Oks and Van Wijnbergen (1995) test the debt overhang hypothesis for Mexico and conclude that it does not exist. Borenzstein (1990), based on a simulated growth model for a typical debtor country, concludes that debt relief does not have important effects on growth. Cohen (1993) finds no evidence for the general existence of a debt overhang using data for sample of 81 LDCs. Yet, for the Latin American countries he shows that high debt had a negative impact on their growth performance. This results is reconfirmed by Cohen (1997), who states that for African countries high debt is not a major cause for low levels of economic growth in the 1980s and 1990s, and by Weeks (2000). Desphande (1997) shows that a debt overhang might exist for 13 severely indebted countries. Kaminsky and Pereira (1996) find evidence for a debt overhang for Latin American countries, once social inequality and its impact on government policy and consumption is taken into account. In current discussions on the need of debt relief for the HIPCs, critics of the debt overhang hypothesis have pointed out that the adverse incentive effects of high debt cannot be an important issue for most HIPCs. HIPCs have experienced positive net resource transfers during the 1990s28. Some other authors, as an alternative to the debt overhang hypothesis, hypothesise that it is not the amount of debt that may hamper economic growth, but the uncertainty about the annual debt service payments. They argue that there is a close link between uncertainty and instability: the annual instability of payments contributes to uncertainty of debt payments. Lensink and Morrissey (2000)29 show that instability of annual aid receipts negatively influences the effectiveness of development aid. Gemmell and McGillivray (1998)30 argue that aid inflows are more volatile than other government revenues and this volatility influences government spending and taxation. Bleaney, Gemmell and Greenaway (1995)31 show that government revenue instability is highest for Sub-Saharan African countries and that revenue instability is associated with expenditure instability and instability in the sources of deficit finance. Oks and Van Wijnbergen (1995)32 stress the importance of the uncertainty of debt service payments on growth. Sachs et al. (1999)33 stress the relationship between the instability of debt service payments and economic growth. 28 W. Easterly (2001), “How Did Highly Indebted Poor Countries Become Highly Indebted? Reviewing Two Decades of Debt Relief”, Unpublished Paper, The World Bank, Washington D.C. 29 R. Lensink and O. Morrissey (2000), “ Aid Instability as a Measure of Uncertainty and the Positive Impact of Aid on Growth”, The Journal of Development Studies, Vol. 36, N. 3. 30 N. Gemmell and M. McGillivray (1998), Aid and Tax Instability and the Government Budget Constraint in Developing Countries, CREDIT Research Paper 98/1, University of Nottingham, Nottingham. 31 M. Bleaney, N. Gemmell, and D. Greenaway (1995), “Tax Revenue Instability, with particular Reference to Sub-Saharan Africa”, The Journal of Development Studies, Vol. 31, N. 6. 32 D. Oks and S. van Wijnbergen (1995), “Mexico After the Debt Crisis: Is Growth Sustainable?”, Journal of Development Economics, Vol. 47. 33 J. Sachs, K. Botchwey, M. Cuchra, and S. Sievers (1999), “Implementing Debt Relief for the HIPCs”, Center for International Development, Harvard University, Cambridge, Mimeo. 145 Borensztein (1990)34 tested the impact of debt stock on private investment by introducing a debt overhang effect in an econometric regression, where investment is the dependent variable. The results were not significant. Moreover, the basic data are not available or they are very fragile. 4. The links between growth and foreign capital inflows New growth theory has inspired a more complex analysis in distinct ways. First, empirical research focuses on growth as dependent variable. A preliminary consideration must be referred to the importance of adequate indicators of growth. In measuring national income, economic literature has sometimes shown confusion between GNP and GDP35. As well shown by King36, different specifications of income can produce different results: in general, GNP seems to be more appropriate to be analysed in terms of disposable income (for consumption or investment functions) to be compared to the cost of indebtedness. In fact, the GDP, measured on the basis of expenditure, income, or value added, includes interest to be paid on debt arrears. It is an accurate guide as to the productive performance of the economy, but it does not adequately reflect the welfare of the population in monetary terms, net of indebtedness costs37. At the end of 1970s, Wasow38 estimated the relationship between GNP, foreign finance flows, savings and GDP/capital ratio. In attempting to define an indicator of dependence on foreign finance, Wasow assumed net foreign flows to GNP ratio as its proxy and its gradual convergence to zero as the condition to escape from dependency in the long term. Simulating different situations (constant growth rate of net foreign flow, or equal to GNP growth ratio, or at constant rate needed to reach a planned GNP growth), this study shows the existence of a unique condition to escape from permanent dependence. The marginal propensity to save multiplied by the incremental GDP/capital ratio must be higher than the growth rate of net foreign flow. Garavello39 suggested a more flexible approach, that should be based on the nature of the prevailing gap (saving gap or foreign exchange gap) to select the indicators to be related to GNP and debt. In case of saving gap, it is useful to consider the relationship between investment and net foreign flow as well as between domestic and foreign capital stock, but the results are equivalent to those of Wasow. In the case of foreign exchange gap, export must be taken in account, and the result is that the condition to escape from permanent dependence on foreign flows is that export earnings become higher than net foreign flows. 34 E. Borensztein (1990), Debt Overhang, Debt Reduction and Investment: The Case of the Philippines, IMF Working Paper WP/90/77. 35 For example, Solomon specified a consumption function based on GDP, as if it were correct to consider as part of disposable income for consumption the amount of GDP owed to abroad as interests on foreign debt. R. Solomon (1977), “A Perspective on the Debt of Developing Countries”, Brookings Paper Economic Activities. 36 B. King (1968), “Notes on the Mechanics of Growth on Debt”, World Bank Staff Occasional Papers, N. 6, J. Hopkins Press, Baltimora. 37 GNP (=Y) is equal to GDP minus the interests paid on external debt (iTED): Y = GDP - iTED. 38 B. Wasow (1979), “Saving and Dependence with Externally Financed Growth”, Review of Economic Statistics, N. 61. 39 O. Garavello (1981), “Processi di sviluppo e dipendenza dai flussi esterni di capitale”, Rivista Internazionale di Scienze Economiche e Commerciali, N. 28. 146 Since the 1960s, neo-classical approach to the link between growth and foreign debt has been widely used. At the end of the 1980s, Eaton, based on the Solow growth model40, simulated the case of a small economy in open economy that contribute to guarantee an optimal allocation of world capital in such a way that marginal productivity of domestic capital is equal to interest rate41. He identified five stages: 1. a first stage in which the country is experiencing a deficit in balance of trade and current account balance (in order to import capital stock), 2. a second stage in which a trade surplus coexists with a deficit in current account balance (due to the payments of debt service), 3. a third stage in which there is a surplus in both balance of trade and current account balance, but the country is still an indebted country, 4. a fourth stage, when the country becomes an international lender, even though income from interests are lower than the amount of provided loans, 5. a final fifth stage, when the country begins to receive net positive flows from abroad. Takagi42 criticised the gap-theory and stressed the importance of distinguishing the poor countries from the middle-income countries in analysing foreign debt, and debt to official lenders from debt to private creditors. These differences are relevant in order to define the stages of debt cycle. Poor countries owe their debt to official creditors, who increase their exposure when a debtor faces crisis (due to saving decrease or interest payments increase), and the direct consequence is the worsening of debt situation. Differently, middle income countries owe their debts to private creditors, who reduce their exposure when debt increases and GDP decreases, inducing recession in the short-term up to the point in which debt service decreases and growth can start again to grow. Bhandari, Haque and Turnovsky43 closed the decade with another analysis based on the neo-classical framework. They assumed, differently from other models, that there is not a perfect elasticity of money supply, and concluded that interest rate on debt increases as debt stock increases and that foreign capital supply is upward sloping, rather being constrained. A conclusion, which has been denied by recent history, when credit supply decreased becoming downward-sloping in presence of an international credit rationing. In the 1990s, Boone adopted a standard neo-classical growth model and found no effect of aid in the long run, because aid is consumed instead of invested44. The linkage between fiscal policy, foreign finance and growth is another issue, which received particular attention in empirical studies. In 1975 Dacy45 found a negative correlation between the amount of foreign finance and the impact of fiscal policy on growth. Given an increasing rate of growth, consumption increases much more than saving, which is due to tax revenue, because of the auxiliary role of foreign finance that is devoted to fund 40 R. Solow (1956), “A Contribution to the Theory of Economic growth”, Quarterly Journal of Economics, N. 70. 41 J. Eaton (1989), “Foreign Public Capital Flows”, H. Chenery and T. N. Srinivasan (eds.), Handbook of Development Economics, North Holland, Amsterdam. 42 Y. Takagi (1981), “Aid and Debt Problems in Less Developed Countries”, Oxford Economic Papers, N. 33, 2. 43 J. S. Bhandari, N. U. Haque and S. J. Turnovsky (1990), “Growth External Debt and Sovereign Risk in A Small Open Economy”, IMF Staff Papers, N. 37, Washington D.C. 44 P. Boone (1996), “Politics and the Effectiveness of Foreign Aid”, European Economic Review, Vol. 40. 45 D. C. Dacy (1975), “Foreign Aid, Government Consumption, Saving and Growth in Less Developed Countries”, Economic Journal, N. 85. 147 higher consumption. Ten years later, Singh46 estimated the relationship among the same variables, analysing two periods (1960-1970 and 1970-1980). Foreign aid seemed to have positive effect on growth, whereas government intervention (through fiscal policy) resulted negatively correlated to growth. But whenever the government intervention variable has been included, the fit of the model has been better (i.e. the multiple coefficient of determination, R2, has increased) and statistical significance of foreign aid has dramatically dropped. It means, from Singh’s point of view, that foreign aid is not particularly important in terms of growth. But the authors do not seem to agree on this issue. Mosley47 stressed that appears to be no statistically significant correlation in any post-war period, either positive or negative, between inflows of development aid and the growth rate of GNP in developing countries. White48 admitted that the combination of weak theory with poor econometric methodology makes it difficult to conclude anything about the relationship. Hansen and Tarp49 conclude that there is strong evidence for a positive effect of aid on investment and growth. Borensztein50 obtained different results, based on a sample of 30 countries, in terms of the correlation coefficient, r, as a measure of the degree of association between public expenditure and foreign debt. The results were quite ambiguous: in 13 cases the correlation was positive, in 12 cases it was negative. Edwards contribution51 was linked to the role of fiscal policy in controlling debt and promoting growth. More particularly, it was linked to the debate on the optimal borrowing tax, which emerged in the 1960s as a way to optimalise the usage of foreign debt in order to boost limited debt by means of tax. He emphasised the presence of externalities due to foreign debt, as demonstrated by the high significance of positive relation between the spread on LIBOR52 and foreign debt/GNP ratio. During last decade, much of the empirical literature on growth, following the early study by Barro53, has focused on estimating cross-country regressions in search of a set of stable relations among the various variables suggested by theory54. New theory of growth offers different and more complex analytical basis compared to previous works on saving and investment. Proxies of economic policy, human capital and the institutional environment are included directly in the growth regressions. GDPratet = f (St + ODAt + Inflt + ETDt+ Opent + ToTt + BDeft + initGDP) 46 (5.4) R. D. Singh (1985), “State Intervention, Foreign Economic Aid, Savings and Growth in LCDs: Some Recent Evidence”, in Kyklos, N. 38 (2). 47 P. Mosley (1987), Overseas Development Aid: Its Defence and Reform, Wheatsheaf, Brighton. 48 H. White (1992), “The Macroeconomic Impact of Development Aid: A Critical Survey”, The Journal of Development Studies, Vol. 28, N. 2. 49 H. Hansen and F. Tarp (2000), ibid. 50 E. R. Borensztein (1989), “Fiscal Policy and Foreign Debt”, Journal of International Economics, N. 26. 51 S. Edwards (1984), “Ldc Foreign Borrowing and Default Risk. An Empirical Investigation, 76-80”, American Economic Review, N. 9. 52 London Inter Bank Offered Rate (LIBOR) is the rate of interest in the short-term wholesale market in which banks offer to lend money to each other. It is the most significant interest rate for international banks, officially fixed at 11 a.m. each day by five major London banks. The spread on LIBOR is the difference between interest rate on foreign debt of developing countries and this benchmark; difference due to the risk of indebted countries. 53 R. J. Barro (1991), “Economic Growth in a Cross Section of Countries”, Quarterly Journal of Economics, N. 106, May. 54 R. Levine and D. Renelt (1992), “A Sensitivity Analysis of Cross-Country Growth Regression”, American Economic Review, N. 82, September. 148 The average annual growth rate of real per capita GDP [GDPratet] - rather than GNP, which seems to be more appropriate - is analysed in terms of: Ø exogenous factors (such as terms of trade [= ToTt]), Ø macroeconomic structure (such as saving [= St] or investment, inflation [=Inflt], public budget deficit [= BDeft]), Ø policies and structural reforms (such as the degree of openness [= Opent]), institutional context (through indexes of institutional quality), Ø human capital quality (through the level of basic education and health), Ø external relations (such as aid [= ODAt] or external debt [= ETDt] Ø and the dynamics of the dependent variable itself (such as the logarithm of initial GDP per capita [= initGDP], capturing the conditional convergence effect). These studies include a wider set of variables and, in econometric terms, they reflect the sophistication of techniques, which make it possible to include the endogeneity of growth theory. The endogeneity of economic policy and aid can be addressed explicitly by lagging most of the explanatory variables, including foreign aid, one year. The relationship can be explicitly seen as non-linear, and non-linearity can be captured through squared terms (such as aid [= ODA2t]) or interaction terms (such as between aid and economic policies [= Opent * ODA2t]). To account for country-wise heteroskedasticity, geographical dummies as unit effects often estimate the parameters. This solution is aimed to correct the limitations of the constant coefficient model, in which the relationship between Xs and Y is the same for all crosssections and time-points. More recently, studies have adopted weighted least squares o have a random coefficient or error component model. This literature suffers, however, from some severe methodological problems. From an econometric point of view, there is an inappropriate treatment of measurement and specification errors, and lack of appreciation of the potential for simultaneity bias55. The first problem is that the data necessary to adequately test the predictions of the models do not exist or are difficult to construct. In many cases, the quality of the data is inadequate. There are also considerable variations in data definitions across Ldcs, which implies different coverage, based on arbitrary conceptual definitions. As a result, the explanatory variables typically introduced in cross-sectional growth regressions suffer from measurement errors. Arcand and Dagenais56 argued that results derived from Barro-type regressions are fragile when such errors are properly accounted for in the econometric procedure. Persson and Tabellini57 have attempted to gauge the sensitivity to measurement errors of the parameter estimates obtained in their cross-section studies by using more appropriate econometric techniques, such as instrumental variables. Moreover, the basic approach used in many cross-country empirical studies of growth, which consists of regressing the time-averaged growth rate for a group of countries on a set of ad hoc explanatory variables, faces other difficulties: 55 P. R. Agénor (2000), The Economics of Adjustment and Growth, Academic Press, San Diego. J. L. Arcand and M. G. Dagenais (1995), “The Empirics of Economic Growth in a Cross Section of Countries: Do Errors in Variables Really Not Matter?”, CRDE Discussion Paper N. 4195, University of Montreal, October. 57 T. Persson and G. Tabellini (1994), “Is Inequality Harmful for Growth?”, American Economic Review, No. 84, June. 56 149 • • • • • • • • • • Heterogeneity between Ldcs regarding growth patterns may be such that it is inappropriate to perform cross-country regressions. The practice of using regional dummies in pooled cross-section time series analyses is based on the assumption that geographical factors may yield a homogeneous sample58. But appropriate tests for pooling are rarely applied and in chapter 4 we stressed the lack of homogeneity even within the same SSA region. The behaviour of the actual growth rate of output reflects both a trend (or long-term) component and a cyclical (transitional movements around the steady state) component. But trend component is often measured as averaging both dependent and explanatory variables over a relatively long period, which is a largely arbitrary length, because the frequency of cycles is not generally known, and it makes difficult to separate the trend component from the cyclical component. It should be also taken in account that cycles have different frequencies across countries and across variables; thus using a uniform averaging period is likely to distort the long-run relationship between variables. Explanatory variables are often a combination of time averages of flows and beginning period of stock variables59. But these variables have different time-series properties: mixing stationary and non-stationary variables in estimation can lead to spurious results60. Averaging implies that cross-country regressions do not represent typical behavioural equations and parameter estimates represent cross-country average, which may not be representative of any individual country. Linear models and OLS techniques may produce unreliable results, because OLS estimates are consistent only to the extent that the individual effects are uncorrelated with other explanatory variables. This is a condition that is unlikely to hold as well as the assumption that the error term is uncorrelated with the explanatory variables is likely to be violated due to the inherently dynamic nature of growth regressions. Linear models and OLS techniques may produce unreliable results, because of an endogeneity or simultaneity bias, resulting from the failure to account for the endogenous nature of some of the explanatory variables. If the relationship between the explanatory variables and dependent variable is nonlinear, then it is difficult to detect empirically a significant correlation. A considerable diversity in the country coverage, the periods examined, and the set of explanatory variables have made it difficult to generalise or to consider any particular study as more reliable than other studies. More generally, the lack of clear theoretical underpinnings for many empirical studies makes it difficult to interpret adequately the results. Most of the cross-country regression results seem to be fragile: any change in the list of explanatory variables (or countries) often destroys the property of the equation, that is a lack of robustness, which is related to the omitted-variable bias. 58 That is why Rodrik preferred to focus only on one specific region. D. Rodrik (1998), “Trade Policy and Economic Performance in Sub-Saharan Africa”, NBER Working Paper N. 6562, May. 59 M. H. Pesaran and R. Smith (1995), “Estimating Long-Run Relationships from Dynamic Heterogeneous Panels”, Journal of Econometrics, N. 68, July. 60 W. H. Greene (2000), Econometric Analysis, 4th edition, New York, Prenctice Hall, 150 5. Current debate on the impact of foreign capital inflows on growth Referring to the last limitation, the study by Burnside and Dollar (1997)61, which tries to deal with non linear effects of aid, address the endogeneity of aid and links the impact of aid to economic and institutional policies, has attracted particular attention and discussed extensively in a number of papers. The important step forward made by Burnside and Dollar is that they remove the idea that foreign capital inflow has the same (positive or negative) impact on growth rate of all the countries, after controlling for a specific set of other additional factors. The main result of this study is that the effectiveness of aid depends on economic policy and that aid has a positive impact on growth, but only in a good policy environment. In sum, the effectiveness of aid in the growth process is directly dependent on the quality of economic policies (good economic management - or good governance - and “strong” institutions). This is an answer to crucial question raised by Cassen in 198662: does aid work? Yes, aid works, when it is selectively allocated to those countries pursuing “good” policies. Burnside and Dollar’s preferred equation is: GDPratet = f (initGDP + PolInstt + InsQut + FMt + Polt+ Aidt + [Polt x Aidt]) (5.5) Where: Ø Initial GDP [initGDP] is assumed to have negative effects on growth; Ø Political instability [PolInstt], measured by ethnic fractionalisation, number of assassinations and the product of the two, is assumed to have negative effects on growth; Ø Institution quality [InsQut], measured by rule of law, bureaucracy, corruption, risk of repudiation and expropriation, is assumed to have positive effects on growth; Ø Financial market [FMQ t], measured by money (M2) relative to income (GDP), is assumed to have positive effects on growth; Ø economic policies [Polt], measured by an index of fiscal, monetary and trade policies (combining a weighted average of budget surplus, trade openness and low inflation), is assumed to have positive effects on growth; Ø aid to real GDP [Aidt] is assumed to have very limited effects or to be insignificant by itself; Ø an interaction term between foreign aid and economic policies [Polt x Aidt], which captures the non linear aid-growth relationship and makes aid dependent on economic policies, is assumed to have positive effects on growth. The model also includes time dummies and dummies for SSA and East Asia. Burnside and Dollar show that the results confirm their assumptions and have the expected signs, particularly the fact that aid marginal contribution to growth depends positively on good macroeconomic policies. But numerous scholars question the econometric work in this study. 61 C. Burnside and D. Dollar (1997), “Aid, Policies and Growth”, Policy Research Working Paper, N. 1777, World Bank, Washington D.C.; and World Bank (1998), Assessing Aid: What Works, What Doesn’t and Why, The World Bank, Washington D.C.; P. Collier and D. Dollar (1999), “Aid Allocation and Poverty Reduction”, Policy Research Working Paper, N. 2041, World Bank, Washington D.C.; C. Burnside and D. Dollar (2000), “Aid, Policies and Growth”, American Economic Review, Vol. 90, N. 4. 62 R. Cassen (1986), Does Aid Work?, Oxford University Press, Oxford. 151 The results are very fragile and data dependent. The coefficient to the interaction term between foreign aid and economic policies is significant and positive, but five observations, which are excluded in Burnside and Dollar regression as “big outliers”, have critical influences on the coefficient. These observations are Nicaragua (1986-89, 1990-93), Gambia (1986-89, 1990-93), and Guyana (1990-93). If they are included, the value of the parameter becomes small and insignificant. Dalgaard and Hansen63 investigate the nature of these observations’ influence on the coefficient estimates by regressing the model on a sample in which the single observation is excluded. They show cross-plots of the changes in the estimated coefficients for four regressors (Polt, Aidt, Polt x Aidt, initGDP), when observations are excluded one-by-one. The changes are plotted against the excluded observation, scaling the change in the estimated coefficient by the estimated standard error, using ± 2 / n as a cut-off point, n being the number of observations. The five observations are not the only ones outside the cut-off value and none of the scaled changes exceed one in absolute value. The 5 observations are not outliers in the sense of having extreme studentised residuals, and other observations in the sample have higher leverage values. Thus, the 5 observations should not be deleted from the sample other than for an ad hoc rule (to make the coefficient significant). Interestingly, by excluding five other observations (Gambia 1986-89, 1990-93, Nigeria 1970-73, 1990-93 and Nicaragua 1978-81) aid results to have high positive impact on growth. If the sample of developing countries is expanded, and we pass from 40 countries to 56 countries, then the coefficient to interaction term becomes insignificant. Thus, the Burnside and Dollar specification lacks of robustness. Hansen and Tarp64 stress the problem of misspecification of the Burnside and Dollar regression. A full non-linear model must include five aid-policy terms: aid, policy, aid squared, policy squared, and interaction term between aid and policy. In fact, only the five terms define a full, second-order, polynomial approximation of the unknown functional form of the relationship. The problem of absorptive capacity of SSA (that is the capacity to manage) and Dutch disease constraints can be based on a Cobb-Douglas function including foreign capital inflow, which replaces the Harrod-Domar and two-gap models, assuming decreasing marginal returns to increased aid. GDPratet = aZt + b1Polt+ b2Aidt + b3Polt2+ b4Aidt2 + b5 [Polt x Aidt] + ε t) (5.6) Where Zt is a set of controls and ε t is the error term. Burnside and Dollar implicitly set b3 = b4 = 0, without testing this hypothesis. Hansen and Tarp test this hypothesis and find statistical support for diminishing return (b3 = b5 = 0; b4 0), preferring the presence of squared terms rather than the interaction term, thus rejecting the selectivity model. Aid must be evaluated after having conditioned on good policies. Their results show that the insignificance of the interaction term is not due to collinearity problems between the three regressors with Aidt. Another important aspect is that endogeneity of aid (and interaction term) implies that OLS estimates are inconsistent and requires particular attention to instruments, in terms of 63 C. J. Dalgaard and H. Hansen (2001), “On Aid, Growth and Good Policies”, The Journal of Development Studies, Vol. 37, N. 6. 64 H. Hansen and F. Tarp (2001), “Aid and Growth Regressions”, Journal of Development Economics, Vol. 64. 152 variation and correlation with the endogenous regressors65. Burnside and Dollar look at a cross country correlation (having geographical dummies as instruments for aid and population, which changes slowly over time), but live time series variation in aid unexplained. Moreover, as we stressed in chapter 4, the assumption of homogeneity within SSA region, which Burnside and Dollar adopt, is strongly rejected by real data. The fact that most of the results - including Burnside and Dollar’s results - appear to be sensitive to the countries included in the sample confirms the importance of specific country contexts. Foreign capital inflows’ effects on growth depend on specific conditions in each recipient country. The quality of results improves when the empirical model has a better specification. And historical evidence suggests that development and the effectiveness of foreign capital inflows are crucially dependent on external and climatic factors. These components, such as terms of trade fluctuations, export instability, oil price, droughts are not included in the Burnside and Dollar’s equation. Consequently, their model is inadequate to capture the complexity of reality. Some recent studies have tried to correct for a few of these problems: some studies66 have provided a proper treatment of country-specific effects; some others67 have addressed the simultaneity bias, using instrumental variables estimation; others68 have addressed both of these problems. Guillaumont and Chauvet69 simply add two variables to the initial Burnside and Dollar’s regression. They include a vector of external and climatic environmental variables (normalised on a scale from 0 to 100) and the interaction term between this vector and the level of aid. They use a two-stage least square (TSLS) procedure, the growth regression being estimated with simultaneous instrumentation of aid and policy. Assumed that aid and economic policy (as well as their interactive term) are endogenously determined, the TSLS methodology is the way to have these variables instrumented. The authors conclude that the macroeconomic effectiveness of aid is crucially dependent on external factors, rather than on the economic policy. They reject the original conclusions of Burnside and Dollar’s paper, based on the selectivity of aid. But another important element they provide is the fact that one should be cautious regarding the robustness of results. In fact, in any case, as many studies, following the 1991 seminal work of Barro, have included a lot of variables to be correlated with economic growth, particular attention must 65 If ODA t may be assumed to be endogenous, then it is important to find instruments for ODA, which are uncorrelated with the error and highly correlated with the dependent variable. Then, the exogenous component of ODA t is extracted and it can be used to run the regression and to examine whether this exogenous component is correlated with the dependent variable. The Hausman test for endogeneity can be used to determine whether instrumenting regressor is necessary. First, we have to regress ODA t on all exogenous variables from the base regression, including some instruments variables. Then, we reestimate the base regression, inserting the fitted value for ODA t. If the fitted value of ODA t is insignificant as an additional regressor in the base regression, then ODA t may be considered exogenous. See C. Mukherjee, W. Howard and M. Wuyts (1998), Econometrics and Data Analysis for Developing Countries, Routledge, London. 66 M. Knight, N. Loayza and D. Villanueva (1993), “Testing the Neo-classical Theory of Economic Growth”, IMF Staff Papers, N. 40, September. 67 J. Aizenman and N. P. Marion (1993), “Policy Uncertainty, Persistence and Growth”, Review of International Economics, N. 1, June. 68 F. Caselli, G. Esquivel anf F. Lefort (1996), “Reopening the Convergence Debate: A New Look at Cross-Country Growth Empirics”, Journal of Economics Growth, N. 1, September. 69 P. Guillaumont and L. Chauvet (2001), “Aid and Performance: A Reassessment”, The Journal of Development Studies, Vol. 37, N. 6. 153 be put on the robustness of variables. In fact, Levine and Renelt70 use extreme bond analysis (EBA) and show that most of the variables are not robust, as their coefficients and significance change substantially, depending which other variables are included in the estimated regression71. Lensink and White72 have examined whether there is an aid Laffer curve in relation to the growth, of real GDP as dependent variable, running a pooled cross-section time series analysis, with a basic panel of 138 countries. GDPratet = b1initGDP + b2initSEt + b3Debtt+ b4[GNPt x Aidt] + b5 [GNPt x Aidt] 2 + b5Ds (5.7) Where initSEt is the initial secondary school enrolment, Debtt is the debt to GDP ratio and Ds are intercept dummies for different regions and sub-periods. The idea of decreasing returns on aid, that is the existence of an aid Laffer curve can be modelled by using a non-linear relationship between aid and growth. Non-linearity indicates that beyond a certain threshold, additional aid can not be productively absorbed by the economy. A squared term for aid is introduced to model the non-linear link, which is determined endogeneously by the data and not in an ad hoc way. The results give an insignificant estimate for [GNPt x Aidt] 2 as they are sensitive to some outliers and the residuals are not normally distributed. Thus, having re-estimated the equation without some outliers, the existence of an aid Laffer curve is confirmed, as the coefficient of the quadratic term has a negative sign. To test the reliability of their results, the authors include some other regressors and estimate 455 regressions, based on all the combinations of the set of 15 independent variables. Then, they run the EBA procedure. They find some evidence for an aid Laffer curve, but the results –including the sign of the quadratic term – seem not robust, as they are very sensitive to the exact specification of the model. Other recent studies have attempted to develop other econometric techniques. Some studies73 analysed the evolution of the entire distribution, considering inadequate to focus on cross-section averages over long period of time. Some others74 preferred to analyse time70 R. Levine and D. Renelt (1992), “A Sensitivity Analysis of Cross-Country Growth Regressions”, American Economic Review, Vol. 82, N. 4. 71 For each regression we find an estimate of the coefficient of the variables of interest and a corresponding standard deviation. The lower extreme bond is the lowest value of the difference between the given coefficient and two times its standard deviation. The upper extreme bond is the highest value of the sum of the given coefficient and two times its standard deviation. The variable is considered not robust if the upper extreme bond is positive and the lower extreme bond is negative (the sign of the coefficient changes). Sala-i-Martin provides an alternative stability analysis, looking at the entire distribution of the coefficients, instead of a zero-one (robust-fragile) decision. He assumes that the distribution of the estimates of the coefficients is normal, then he computes the point-estimates and the standard deviation, the mean and the average standard deviation of the assumed normal distribution. Then he computes which fraction of the cumulative normal distribution is on the right or left side of zero. If the largest of the two areas exceeds 0.95, then the variable is said to have a robust effect on dependent variable. See X. Sala-i-Martin (1997), “I Just Ran Two Million Regressions”, American Economic Review, Vol. 87, N. 2. 72 R. Lensink and H. White (2001), “Are There Negative Returns to Aid?”, The Journal of Development Studies, Vol. 37, N. 6. 73 D. T. Quah (1996), “Empirics for Economic Growth and Convergence”, European Economic Review, N. 40, June. 74 A. B. Bernard and S. N. Durlauf (1996), “Interpreting Tests of the Convergence Hypothesis”, Journal of Econometrics, N. 71, March. 154 series regressions for individual countries, as time-series technique allows to analyse the possibility of bi-directional causality, whereas averaging out variables over long periods of time makes cross-country variations difficult to be interpreted. Other researchers75, even if using a limited number of observations, used a cointegration approach to analyse the determinants of growth in an individual-country context, an approach that makes it difficult to capture changes in the steady state itself. Pooled time series or panel data econometrics seems to be a promising framework for integrating the cross-country and time-series approaches76. 6. The links between macroeconomic policies and growth The important Burnside and Dollar’s paper has led to frequent debates on both methodological and political implications. It clearly assumes that macroeconomic policies determine growth. At this regard, some studies had previously tried to test this hypothesis, including the analysis on the role of foreign borrowing. Stanley Fischer (1991) aimed to establish that macroeconomic policies matter for growth. By macroeconomic policies he means, monetary, fiscal and exchange rate policies that help to determine the rate of inflation, the budget deficit and the balance of payments. This is exactly the set of policies, which have been adopted by Burnside and Dollar. Fisher’s cross sectional evidence suggests that macroeconomic factors and policy affects economic performance. GY = 1.38 - 0.52 RGDP70 + 2.51 PRIM70 + 11.16 INV - 4.75 INF + 0.17 SUR 0.33 DEBT80 - 2.02 SSA - 1.98 LAC where: GY=per capita real growth over the period 1970-85 RGDP70=real income in 1970 PRIM70=enrolment rate for primary school 1970 INV=average share of investment in GDP over 70-85 INF=average inflation rate over 70-85 SUR=ratio of budget surplus to GNP over 75-80 DEBT80=Debt/GNP in 1980 SSA=Sub Saharan Africa dummy LAC=LACs and the Caribbean dummy The evidence (data and regressions) supports the view that quality macroeconomic management (reflected in the inflation rate, external debt ratio and the budget surplus) matters for growth. Macroeconomic management might affect economic growth because it might affect investment and thus the rate of change in capital. It also might affect the efficiency with which the factors are used. Fischer estimates cross section investment (average share of investment in GNP) over 1970-85. The independent variables are GY, RGDP70, PRIM70, INF, SUR7580, DEBT80, BLAV (average black market premium), PINV (relative price of investment goods), SSA and LAC (dummy variables to control for SSA and LAC). Although, none of the set of alternate regressors can provide a 75 P. Arestis and P. Demetriades (1997), “Financial Development and Economic Growth: Assessing the Evidence”, Economic Journal, N. 107, May. 76 J. Temple (1999), “The New Growth Evidence”, Journal of Economic Literature, N. 37, March. 155 (5.8) satisfactory account of the determinants of investment, Fischer argues that at least BLAV and INF affect investment. Fischer concludes those macroeconomic indicators (inflation rate, external debt, government deficit) and hence, macroeconomic policies matter for growth. The results are less clear on the mechanisms through which macroeconomic policy affects growth. The separate role of macroeconomic variables in the growth regressions implies the existence of other channels, which need investigation. X.Sala-I-Martin (1991) stressed that the relation between inflation and growth, and budget deficits and investment lack foundations. In the absence of theories, the correlation between inflation, budget deficit and growth might be spurious or the direction of causation is reversed. For example, models that relate steady state growth with inflation predict no relation (superneutrality of money) or a positive relation along the transitional path towards a higher steady state level of income. If the latter is the case, then cross country regressions have little to say about the steady-state growth rate. Also, either budget deficits have no effect on growth (Barro, 1974) or have a negative effect on income through high real interest rate and the crowding out of investment (Blanchard, 1985). Fischer finds a negative relation between government surplus and investment (or positive relation between budget deficit and investment) and this needs a good theoretical explanation. Sala-I-Martin worries about the endogeneity of the macroeconomic variables, specially the budget deficit in Fischer’s regressions. He replicates Fischer's growth and investment regression extending the sample size and excluding the foreign debt variable because it was never significant. In the growth regression, INF was not significant. In the investment regression, neither SUR nor INF was found significant. Only BLAV and PINV were significantly negative. Even when assuming a non linear relation between I and BLAV, the macro variables remain insignificant. Therefore, the relation between BLAV and INF and the growth and investment rates are not as clear as suggested by Fischer. Sala-I-Martin concludes that the empirical evidence presented by Fischer is weak. However, he argues that this does not mean that we should not worry about macroeconomics. The main contribution of Fischer’s paper is to highlight the need for a theory that could explain the relation between short-run macroeconomic management and long-run growth. Michael Bleany (1996) also worries about the effects of macroeconomic instability on investment and growth. He takes similar approach to that of Fischer (1991) but incorporates robust regressors for his growth and investment equations and then, adds variables, which capture macroeconomic stability. He tested the following cross-section regressions for 41 countries over 1980-1990 : PCGR = a0 - a1 LGR + a2 INV + a3 XYGR - a4 LYPC79 +a5 BS - a6 SDERER - a7 CPINFL - a8 DEBT79(or HIC) + u1 INV = b0 + b1 XYGR + b2 LXY79 +b3 RERDOL +b4 BS - b5 SDERER - b6 CPINFL b7 DEBT79(or HIC) + u2 where the regressors found significant by previous research are: PCGR= average annual growth rate of per capita output (in logs, 1980-1990). LGR= average population growth (in logs, 1980-1990). 156 (5.9) (5.10) INV= investment / GDP (average 1980-1990). XYGR= average annual growth rate of the exports/GDP ratio (in logs, 1980-1990). LYPC79= log of the US dollar per capita GDP in 1979. LXY79= 1979 exports/1979 GDP (in logs). RERDOL= index of real exchange rate distortion calculated by Dollar for 1976-85. And the policy-induced macroeconomic instability variables are: BS= govt budget surplus (% of GDP). SDERER= standard deviation of the log of the real exchange rate over 1980-1990, i.e. exchange rate volatility. CPINFL= average annual consumer price inflation over 1980-90 and set to 100% if it exceeded that level. DEBT79= end 1979 foreign debt/1979 exports revenue HIC= dummy variable taking the value of 1 if country was classify as highly indebted by the World Bank and 0 otherwise. He found that, for a given INV, BS and SDERER were significantly correlated with growth. INF did not significantly affect growth negatively. The policy-induced variables appeared no to affect investment significantly, although they had the expected signs. He concluded that macroeconomic instability affects growth negatively but that it was not clear that it affected the volume of investment. In contrast to Fischer (1991) and Bleany (1996), Corden (1990) does not take an econometric approach but analyses the experiences of 17 developing countries so as to draw some lessons from experience and pinpoint policies which promote growth. He arrives at the following lessons: - From the 1970s public spending booms caused by the ready availability of funds from the world capital market: a) Sound and profitable spending, hence the need for a cost-benefit approach. b) Beware of euphoria (arising say from liberalisation reforms) since it might lead to over-borrowing by private investors or the government. - From de 1980s crisis and adjustment: c) Unfavourable and surprise shocks can hardly be avoided. Therefore, countries should aim to establish favourable initial conditions (e.g. low debt ratios, high reserves, avoid spending booms) so as to make their economies and policy reactions as flexible as possible. d) If the country has a chance, it should plan gradual adjustment but initiate a comprehensive adjustment program promptly. - From the effects of inflation on growth: e) Countries should avoid inflation since it is hard to reduce it without severe both economic and political costs. - From the effects of exchange rate policy on growth: f) It is not necessary to use the exchange rate as a nominal anchor to make a noninflationary monetary policy commitment. The use of fiscal restraint is usually more important. g) Real exchange rate misalignment and variability should be avoided through appropriate nominal exchange rate adjustments. h) Devaluation will be ineffective if it is not part of a policy package involving monetary and fiscal policies. 157 Sound macroeconomic policies seem to promote growth in developing countries. This becomes evident when studying the experiences of developing countries since the 1970s. However, both the relation between macroeconomic stability and growth and how it affects growth, becomes less evident in econometric studies. Dornbusch (1985) explored the role of disequilibrium exchange rates and budget deficits in promoting external indebtedness and the 1980s debt crisis. He analyses the period 197882 for Argentina, Chile and Brazil. Oil, US interest rates and the 1981-82 world recession are often mentioned as the causes of the crisis. However, these factors only made clearer the underlying disequilibrium in which exchange rate overvaluation and budget deficits were perpetuated by a continuing and excessive recourse to the world capital market. Dornbusch proposes a simple framework to analyse the causes of the debt crisis: GNP = GDP + NFP (5.11) Where: GDP = final output produced by residents (nationals and foreigners) of the country within the territory of the country, NFP = payments received by residents for factor services (K,L) rendered abroad minus payments made to other countries for factor services rendered by non residents. For most LDCs, NFP<0. CA = GNP + NTR – E (5.12) Where: CA = Total receipts (= Total income received by residents + transfers) - total payment (= Expenditure on goods and services + transfers). NTR = net transfers. For most LDCs, NTR>0 E = expenditure. CA = (S - I) + (T - G) since GNP = C + T + S and E = G + C + I (5.13) These two last identities are crucial to the understanding of external balance issues: - Current account (CA) deficits reflect and excess of expenditure over income - Improvement of the CA can be brought only if savings rises relatively to investment or if the government surplus improves - There is a direct link between budget surplus (T-G) and the external balance Let NFA = stock of net foreign assets i.e. all claims by residents on the rest of the world less all claims by foreigners on domestic residents. Then the change in NFA is equal to the current account and just tells that all bills must be paid. If CA>0, then income>spending, so we make claims on the ROW. If CA<0, then income<spending and thus we are borrowing or selling assets. From the way the current account deficit (CADEF) is financed we can derive CADEF = ExtDebt + DFI – R – KF (5.14) 158 or ExtDebt = CADEF – DFI + R + KF (5.15) that is, an increase in gross external debt can have three broad sources: current account deficits not financed by long-term capital inflows (CADEF-DFI), borrowing to finance official reserves build-up (R), or private capital flight (KF). There is a link between CADEF, budget deficit and excess of private investment over private savings. Dornbusch concludes that the experience of these countries show that the world debt problem was not only caused by world recession, dollar appreciation and unexpected increase in the world interest rates but also by overvalued exchange rates, import sprees or capital flight financed by external borrowing. 7. The links between ODA and its determinants Another model specification, which has been extensively studied during the past years, is to consider foreign capital inflow, in terms of ODA, as dependent variable. Theories and the use of econometric methods to estimate ODA determinants and allocation can be grouped into three approaches (Pesenti, 1987). The first approach, developed by Mosley (1985) and, not long before, by Beenstock (1980), explains the ODA flows trend through a time series analysis, based on the aid demand from NGOs and private enterprises of donor countries and aid supply from donor countries. The empirical results of this kind of analysis cannot be considered adequate, because of the difficulties in translating the social welfare, demand and supply functions into variables, in finding an aid quality indicator (proposed by Mosley). Beenstock achieved a multiple correlation coefficient close to one, but together with a Durbin Watson value of 0.97, which can be interpreted as the presence of a residuals positive autocorrelation or specification bias of the equation. The second approach, proposed by Edelman-Chenery (1977), analysed the geographical bias of aid allocation, considering the impact of such variables as population and GDP per capita of recipient countries, through the use of cross-country regression analysis. The empirical results are very partial, because this approach considers just a part of the main determinants of aid allocation. The third area, the most prolific and complete, was developed by Dudley and Montmarquette (1976), and received the major interest in Italy. It is based on the simultaneous analysis of different determinants of aid allocation, from one or more donor countries to some developing countries, in a given year. This approach makes it possible to study the interaction between different interests of recipient and donor countries, and to suggest some patterns of donor’s behaviour. The multiple regression analysis, based by Isenman (1976) on the ordinary least squares (OLS) method to estimate the coefficients, through the step-wise technique, has become the most used method in this third area of research. Just in two cases, a different analysis was used: the TSLS method by Mosley (1981) and a non-recursive structural equations system by Roeder (1985). In most cases, cross-sectional analyses have been used; recently, some dummies have been introduced, and just once, pooled analysis was specified to take in account time (Dowling and Hiemenz, 1985). 159 Usually, the period of time to be investigated has been limited within a decade, and it has never been adopted a time series analysis. Thus, past studies on aid allocation preferred the use of comparisons between specific years, estimating the political changes over the time and using the weighted mean of the considered years. These analyses have examined aid allocation over the period 1960-1980 (and once beginning with the 1940s). In terms of recipient countries to be examined, there have always been many developing countries in the sample, sometimes considering particular groups of Ldcs - Henderson (1971) excluded the former colonies - or geographical areas. A sample of fewer than 17 Ldcs (Chambas 1986), or more than 93 Ldcs (Dudley and Montmarquette, 1976) has never been considered. Referring to the donor countries, the studies ranged from 1 to 15 (Dudley and Montmarquette, 1976) and 16 (Henderson, 1971) case studies. In actual fact, the prevailing idea is that, given the absence of a strong homogeneity among the different political motivations leading donors’ strategies, it is better to limit the analysis to one donor country. The United States have been the most studied, whereas those who prefer comparisons - McKinley and Little (1979), and Mosley (1981) - have compared the main donors (France, Germany, United Kingdom, Japon, besides the United States). Usually, the dependent variable is a proxy of aid volume. Current or constant US$, total aid or standardised values of per capita aid or aid as a percentage of GNP, import or export, bilateral or multilateral aid, ODA or total net financial flows, annual commitments or, in particular, disbursement. Sometimes, aid flows in logarithmic form, to correct the biased effect of the outliers. In terms of independent variables, given the different approaches we have briefly described above, we have: (a) Ldcs population to measure the level of distribution equality between developing countries; (b) Ldcs income level (or growth rates) to measure the correlation between allocation and people’s needs in recipient countries. On the basis of this variable, the main findings of empirical studies confirm the hypothesis of a biased allocation, benefiting middle income countries and the small countries. Some “control” variables, referred to the development and economic integration of recipient countries, have been introduced. The main regression analyses have included the following variables: (a) the volume of merchandise exports to the given donor country, (b) the volume of merchandise imports from the given donor country, (c) the flow of foreign direct investment, (d) gross international reserves, (e) balance of payments, (f) direct investment from the donor country, (g) the number of transnational corporations operating on that country, (h) availability of raw materials, (i) the structure of production measured by the sectoral percentage of GDP in current US dollars. Moreover, some other “control” variables, referred to political aspects, have been considered. In particular: (a) instability, (b) alignment referred to political blocks, (c) expenditure on arms and trade, (d) length of independence, 160 (e) security treaties and links, (f) a dummy variable of the strategic interest in the country, (g) the number of US army bases. Some other variables, which have been considered, are those referred to demographic and social aspect, as: (a) life expectancy at birth, (b) infant mortality rate, (c) protein consumption per capita, (d) adult illiteracy and education enrolment. Just once (Wittkopf, 1972), the volume of multilateral aid and bilateral aid - apart from the given donor - have been included and treated as a proxy of the trend prevailing at international level, which can be described as a “towing-effect” of the main world aid strategies and international division of labour. In such a case, a positive coefficient can indicate the “run” component, that is an imitative and co-operative attitude among donors; whereas, a negative coefficient can indicate the prevailing effect of the competition or the division of labour among donors rather than a coherence between the interventions. In all, the independent variables which have been considered in the analyses vary from six (Mosley, 1981) to twenty-five (McKinlay e Little, 1979), ten-eleven being the average number of variables. The main objective, given this complex set of independent variables, has been to test the prevalence of egoist motivations of donors - strategic, political, security, trade, historical, cultural - over the needs and reliability of the recipient countries. Wittkopf (1972), considering the use of dependent and independent variables, underlines the fact that empirical results vary a lot in accordance with different measures of the dependent variable, “aid” (current US dollars or constant, for example). He also points out the fact that one year is the most appropriate lag to be applied to independent variables compared to the dependent one. Most of these considerations seem pertinent to the analysis of external debt, too. 8. The measurement of debt servicing capacity and country risk of default The empirical research on debt capacity reflects the nature of the theoretical foundations, and their conceptual problems. The use of external borrowing is not limited to augmenting investment (first gap) and imports (second gap), but it can be used to fund consumption, in presence of income fluctuations. McDonald77 stressed an important element to be considered: debt can be said sustainable from both the borrowing country and lenders’ perspective. A debt situation may be theoretically sustainable from a debtor’s point of view, if it is consistent with its intertemporal budget constraint, but it may be unsustainable from the lenders’ point of view, because of international credit rationing. Thus research discovered he importance of the creditor perspective. From the creditor perspective, there is a main concern for the willingness of the debtor to sustain repayment of debt, that is called creditworthiness or country risk. Since the beginning of the 1970s, several commercial banks introduced their checklists, based on a long battery of different indicators, to control the country-risk associated with 77 D. C. McDonald (1982), “Debt Capacity and Developing Country Borrowing: A Survey of the Literature”, IMF Staff Papers, Vol. 29, N. 4. 161 their exposure78. The increase in lending to governments of developing countries by private commercial banks raised questions of country credit-worthiness, that is the capacity of the borrowing country to service the debt (to pay interest and amortisation in foreign currency). Most of these techniques lacked of any robust theoretical foundation to select indicators and used arbitrary criteria to weight different indicators79. They failed to distinguish between liquidity and solvency problems, too. The problem is that there is no formula or one definitive approach to determine the borrower’s ability to generate sufficient foreign exchange to meet debt service obligations. The main challenge is to understand development process and the way in which the course of development affects the balance of payments. To determine the capacity to service external debt, some analysts engage in ratio analysis, examining such ratios as debt outstanding/GNP, debt service/exports, debt service/debt outstanding, net transfer/imports, imports/international reserves, imports/exports, rate of growth of debt/rate of growth of exports. If the numerator in the fraction increases, debt crisis might be indicated. Unfortunately, no one ratio captures the risk of a sharp fall in any kind of foreign exchange inflow and a rise in import needs, compared with the ability to offset such risks by compressing imports and using international reserves. Beyond ratios, country risk analysis monitors some key performance indicators, indicating how national economic management is affecting the growth of the economy. They are a raising ratio of savings to national income, a raising ratio of taxes to income, a decreasing ratio of incremental capital to output, a decreasing current account deficit, high rate of growth in employment relative to the rate of growth in output, more equitable income distribution. In 1992, Ngassam80 looked at the empirical determinants of lending to SSA and he presented a defensive lending model in which an increase in the riskiness of a country promotes more lending to protect previous loans. The regression consists of 256 observations, eight annual observation per country from 1982 to 1989 for 32 countries. The dependent variable is the value of the loans disbursed to each country in each year. The independent variables are the amount of debt outstanding to private creditors, trade ratio to the GNP, variability of reserves, total debt service paid, real per capita GDP, the number of deaths from political violence, the length of present political system. The results provide strong support to the defensive lending model, in opposition to the literature that concludes that anything that makes the country riskier reduces lending. Lending is strongly correlated with the amount of debt outstanding to private creditors, more so for commercial banks than for official creditors, and with debt service payments. As the size of the debt service payment rises, lending rises. Countries with high trade are lent less since the fear of trade penalties is sufficient to prevent default and additional lending is unnecessary. The political pressure variables tend to be less significant than the others. The implications of this study are quite interesting. As the empirical results support the defensive lending model, debt forgiveness may be more harmful than helpful for SSA. In fact, the crucial question is 78 Within the context of pure theory of country risk, there are three crucial aspects of the problem. Enforcement (the lack of any international collector of debt service), moral hazard (the incentive to cheat in the absence of penalties for cheat) and adverse selection (the incentive to conceal information about one’s true nature). 79 P. Nagy (1978), “Quantitative Country Risk: A System Developed by Economist at the Bank of Montreal”, Columbia Journal of World Business, N. 13. 80 C. Ngassam (1992), “The Empirical Determinants of Lending to Sub-Saharan Africa”, Dept. of Finance, University of Delaware, Newark, mimeo. 162 whether the creditor governments will continue to make future loans to Africa once the former loans have been forgiven. Apart from studies of country risk, an interesting area of studies emerged as concerned with debt servicing capacity problems, causes and likelihood of rescheduling and arrears. Because a clear manifestation of debt servicing problems occurs when a debtor country requests for debt renegotiating, literature ha considered the incidence of debt reschedulings as a proxy measure for debt capacity problems. Given the dichotomous nature of the problem (rescheduling is a “yes” or “not” type variable), in this area of studies the classical regression models are not appropriate. Thus, scholars have implemented other techniques. (a) Principal components analysis. Dhonte81 used this technique, based on the linear combination of observed variables, possessing properties such as being orthogonal to each other, and the first principal component representing the largest amount of variance in the data, to identify a limited group of dimensions, which capture the main information. Unfortunately, the economic significance to be given to the components resulted difficult and the picture remained unclear. (b) Discriminant analysis. More appropriate than the previous technique to address the problem of dichotomous variables, this technique was used by Abassi and Taffler, among others, to study 95 countries from 1967 to 1978, involving 42 variables82. This method allow you to discriminate between two group, in these case being those countries having renegotiated debt from one side, and those countries having not being in another side. Even though the huge amount of variables makes it difficult to synthesise results, nevertheless the outcome is quite clear: financial component of this problem (debt/export ratio, inflation rate, domestic credit/GNP) is the bulk of any serious country risk and liquidity. Also Frank and Cline83, who used discriminant analysis to try to identify the characteristics of borrowing countries that presage rescheduling of their debt, found the same basically financial variables to be positively associated with rescheduling. (c) Logit analysis. This analysis seems to be more appropriate. It takes the logarithm of the odds ratio in favour of the option of rescheduling (the ratio of the probability that a country will reschedule to the probability that it will not reschedule), in order to guarantee that the estimated probabilities lie in the 0-1 range84. Feder and Just85 apply logit analysis to 41 countries in the 1965-1972 period. They found some financial variables to be positively associated with rescheduling (foreign capital flows/debt service ratio, debt service/export earnings ratio, import/international reserves ratio, export growth, maturity of debt) as well as the main economic indicator (per capita income). 81 P. Dhonte (1975), “Describing External Debt Situation: a Roll Over Approach”, IMF Staff Papers, N. 22. B. Abassi and R. J. Taffler (1982), “Country Risk: A Model of Economic Performance Related to Debt Servicing Capacity”, Working Paper N. 36, City University Business School, London. 83 C. Frank Jr. and W. R. Cline (1971), “Measurement of Debt Servicing Capacity: An Application of Discriminant Analysis”, Journal of International Economics, N.1. 84 A main statistical problem with logit and probit models arrives when the estimated probabilities differs from 0 or 1. In this case, it could be useful to fix a cut-off point, considering equal to 1 all the probabilities > .5 and equal to 0 all the probabilities ≤ .5. This solution can fail to reach the objective of minimising the errors, due to the low percentage of rescheduling cases in the sample, which should suggest having a cut-off point at a lower level than .5. 85 G. Feder and R. Just (1977), “A Study of Debt Servicing Capacity Applying Logit Analysis”, Journal of Development Economics, N. 4. 82 163 Mayo and Berrett86, applying the same analysis to 48 countries in the 1960-1975 period, confirmed the same results, finding positively associated with rescheduling both financial (debt/export ratio, import/international reserves ratio, inflation rate) and economic (gross investment/GDP ratio, Import/GDP ratio) variables. Edwards87 found that the spread between interest rates on debt payments and LIBOR seemed to be highly correlated with external debt/GNP ratio, implying the existence of externalities and adverse selection in the indebtedness process. Also Cline 88 used the logit analysis and he found the importance of the same variables from the demand side, including the current account balance/exports ratio square, which implies a weighted measure89 of non linear relationship90, and some variable from the supply side. In particular, international credit rationing, within a context of asymmetric information, seems to explain the rescheduling phenomenon. Bresolin91 applied logit analysis to 15 countries in the 19711986 period and he confirmed the importance of domestic financial and economic variables (particularly, per capita GNP growth rate). (d) Probit analysis. This model is related to the logit, the chief difference being that the normal, or probit, curve is used in place of the logistic cumulative distribution function of logit analysis to model regressions where the response variable is dichotomous, taking 0-1 values. Kharas92 applied probit analysis to the rescheduling of 43 countries in the 1965-1976 period, and he confirmed the importance of both economic and financial domestic aspects. Thus, most of these different studies found that variables which are positively associated with rescheduling are debt service obligations at the time of rescheduling, ratio of imports to foreign reserves, stock of debt as a ratio of GNP, and stock of debt as a ratio to exports. In order to classify the set of variables involved in these studies, it is possible to identify some main groups: Ø Debt burden indicators: relating the stock of debt or debt service payments to exports, imports or GDP. Ø Other Balance of Payments indicators: rate of growth of exports, the current account deficit, the stock of international reserves. Ø Development indicators: GDP growth rate, per capita income, domestic investment on GDP ratio, inflation. In a cross-section statistical model, higher income inequality was found to be a significant predictor of a higher probability of debt rescheduling. This is because political management becomes more difficult in economies with extreme inequalities93. 86 A. Mayo and A. Berrett (1978), “An Early Warning Model for Assessing Development Country Risk”, S. Goodman (ed.), Financing and Risk in Developing Countries, New York. 87 S. Edwards (1984), op. Cit. 88 W. Cline (1984), “International Debt. Systemic Risk and Policy Response”, Institute for International Ecoomics, Washington D.C. 89 Exports act as the weights, taking in account the absolute weight of the economy. 90 The presence of the square ratio implies a non-linear function. 91 F. Bresolin (1990), “Il rimborso del debito estero dei paesi in via di sviluppo: un’analisi del periodo 1971-1986”, Economia Internazionale, N. 43, 2-3. 92 H. Kharas (1984), “The Long-Run Creditworthness of Developing Countries: Theory and Practice”, Quarterly Journal of Economics, N. 99, 3. 93 A. Berg and J. Sachs (1988), “The Debt Crisis. Structural Explanations of Country Performance”, Journal of Development Economics, N. 29. 164 Other recent techniques involve simulation estimation analysis94 and vector autoregressive methodology95. 9. The measurement of debt servicing capacity applied to SSA countries Among the recent studies applied in the specific context of Sub-Saharan African, two studies appear important. They have included the ratio of debt service to exports as the dependent variable, considering it as a proxy for a debt crisis, rather than more correctly as a proxy of the relative burden of debt on the economy measuring the proportion of foreign exchange which is not free to purchase imports. The first96 used an ordinary least square regression applied to 17 countries and included real GNP growth, decline in GNP due to terms of trade losses, maturity of debt as significant explanatory variables. The second 97 used a TSLS method applied to 11 countries and included level of development (measured by the ratio of real per capita GDP/average ratio of per capita GDP of industrialised countries), interest rates, average maturity of debt, degree of openness and the ratio of domestic prices index/international price index as significant explanatory variables. Three other studies have applied the logit analysis, all of them assuming the probability of debt rescheduling as the dependent variable. The first98, applied to 8 countries, involved debt service/GNP ratio, amortisation/debt ratio, revolutions, elections and purges as significant explanatory economic and political variables. The second 99, applied to 45 countries, involved debt service/export ratio, international reserve/imports ratio, debt service/capital inflows ratio, real GDP growth, domestic inflation and net government deficit/GDP ratio as significant explanatory variables. The third100, applied on 39 countries, included a large number of independent variables. They are Debt/GNP ratio, Debt/Exports ratio, Debt service/GNP ratio, Debt Service/Exports ratio, Investment/GNP ratio, International reserves/debt ratio, International reserves/Imports ratio, payment of interest rate on debt/debt ratio, real GDP growth, per capita income, domestic inflation, share of agriculture in output, terms of trade, export and import unit value, capital inflow/GNP ratio, real export growth, import/GDP ratio, trade deficit/GDP ratio, IMF debt/Imports ratio. 94 V. A. Haijvassiliou (1993), A Simulation Estimation Analysis of the External Debt Crises of Developing Countries, Cowles Foundation Discussion Paper, N. 1057, New Haven, Yale University. 95 A. Beltratti (1989), “Empirical Estimates of the Capacity to Repay a Foreign Debt: A Vector Autoregressive Methodology”, The European Journal of Development Research, Vol.1, N. 2. 96 S. Lall and G. Perasso (1989), “Determinants of the Debt Problem in eastern and Southern Africa: A Statistical Analysis”, Rivista Internazionale di Scienze Economiche e Commerciali, Vol. 36, N. 10-11. 97 I. O. Taiwo (1991), “Commodity Prices and Debt Crisis in Sub-Saharan Africa”, Eastern Africa Economic Review, Vol. 7, N. 2. 98 A. M. Assiri, R. A. Parsons and N. Perdikis (1990), “A Comparative Analysis of Debt Rescheduling in Latin America and Sub-Saharan Africa”, Scandinavian Journal of Development Alternatives, Vol. 9, N. 23. 99 C. Ngassam (1991), “Factors Affecting the External Debt-Servicing Capacity of African Nations: An Empirical Investigation”, Review of Black Political Economy, Vol. 20, N. 2. 100 M. O. Odedokun (1995), “Analysis of Probability of External Debt Rescheduling in Sub-Saharan Africa”, Scottish Journal of Political Economy, Vol. 42, N. 1. 165 Another study101, applied to 39 countries, used the ordinary least square regression combined with random effect technique to explain the amount of debt rescheduling and interest arrears. The independent variables are changes in interest rate on new lending, long term debt/total debt ratio, international reserves/debt ratio, international reserves/imports, investment/GNP ratio, real GDP growth, per capita income, domestic inflation, share of agriculture in output, terms of trade, export and import unit value, capital inflow/GNP ratio, real export growth, imports/GDP ratio, trade deficit/GDP ratio, IMF debt/imports ratio. These studies showed some contradictory results: imports/GNP ratio is statistically significant, but in the study by Taiwo the sign is positive, whereas the study by Odedokun has a negative relation. This difference may be due to the period under investigation: the 1970s and 1980s in the first case, when borrowing surged, compared to the 1980s in the second case, when imports were already compressed. Another result is the fact that the higher is the relative size of agricultural output, the higher the probability of debt rescheduling, confirming the idea of agriculture as a proxy of backwardness. The fiscal deficits/GDP ratio has statistically significant positive effect in increasing the probability of debt rescheduling. It confirms what the majority of empirical literature has neglected: external debt in Africa is predominantly a public sector liability, which can not be reduced to a problem of only external insolvency, and the fiscal weakness of government is a key proxy of debt crisis. 10. The links between public sector fiscal behaviour and foreign capital inflows The state in a developing country is limited in its ability to play an activist role. The available resources are scarce since tax bases are small and tax administration weak, much of tax revenue comes from distortionary indirect taxes such as excise duties. With diminishing official aid and poor private equity flows, external financing of the fiscal deficit has to rely increasingly on external debt. In developing countries, where the main borrowing agent is the central government, the fiscal source of deepening external debt problems is evident if the tax base is not expanded commensurately with maturing debt service obligations102. As the benefits of increased investment and consumption linked to external borrowing used to finance deficit mainly accrue to private agents, where the costs of repayment remain with the government, the problem of sustainability of internal and external debts can be described and analysed in terms of the “twin deficits”103. But, when we look at empirical investigations, we have to consider the problem of measurement of fiscal deficit. Tanzi104 stressed that the conventional measure of the deficit fails to recognise that different tax and expenditure categories have different effects on aggregate demand105. Another relevant question is the problem of arrears’ accumulation: the 101 M. O. Odedokun (1993), “Econometric Analysis of External Debt Burdens of African Countries: Debt Rescheduling and Arrears of Interest”, African Development Review, Vol. 5, N. 2. 102 H. Kharas (1984), “The Long-Run Creditworthiness of Developing Countries”, The Quarterly Journal of Economics, Vol. 99,N.3. 103 R. Jha (2001), “Macroeconomics of Fiscal Policy in Developing Countries”, The Australian National University, Camberra, mimeo. 104 V. Tanzi (1993), “Fiscal Deficit Measurement: Basic Issues”, in M. Blejer and A. Cheasty (eds.), How To Measure the Fiscal Deficit, IMF, Washington D.C. 105 Expenditure on the infrastructure has a different impact than expenditure due to consumption subsidies, and within expenditure on the infrastructure there are some expenses productive in nature 166 interest payments have specific nature and effects, as public debt is rapidly raising, with new debt being issued to meet interest payments (the so-called “Ponzi” game). Moreover, tax revenues are not exogenous of expenditures, as the level of public expenditure determines national income, which then determines tax revenue, at least in part. Finally, there is a problem of different sources of financing the deficit106. The difficulties in measuring and interpreting the deficit notwithstanding, it is quite evident that SSA countries have considerable difficulties in meeting internal and external deficit sustainability conditions. Some studies107 have examined the relationship between government expenditures and revenues in developing countries. Others have studied the determinants of government expenditures108. McGillivray and Morrissey (2001a) provide a review of numerous studies on the linkages between aid and public sector fiscal behaviour. The so-called fiscal response literature was originated by the study of Heller (1975). Current literature is on the effect of aid on various categories of public sector revenue and expenditure. Some studies are interested on the fungibility problem of aid (Pack and Pack, 1990, 1993; World Bank, 1998; Feyzioglu et al., 1998). Other studies simultaneously analyse interactions between aid, taxation and expenditure decisions. The public sector borrowing requirement net of aid loans, being a component of the public sector budget constraint, receive marginal attention. It is simply assumed to be a substitute for aid. When aid increases, borrowing decreases. Borrowing can be considered financing of last resort, as it covers a gap between expenditure and revenue, which is not covered by aid. But, a survey109 of the empirical results of literature on aid and public sector fiscal behaviour finds that the results of a number of studies are consistent with aid leading to increases in this borrowing. Further investigation, in the form of econometric analysis of panel data, also points to this outcome. Aid may lead to an increase in external borrowing. 11. The econometric evidence of fiscal dimension of external debt Different studies have analysed the links between fiscal deficit and external debt accumulation. Tanzi110 found that, during the first years of the 1980s, external debt financed more than half of fiscal deficit in a sample of 30 developing countries. He found that 64 developing countries had used, during the 1980-1983 period, external borrowing to finance and others wasteful in nature. Thus, we should distinguuish between revenue or current deficit and capital deficit 106 In 1990, de Hann and Zelhorst found a positive correlation between inflation and the fiscal deficit in developing countries only when inflation rate is high and there is a clear seignorage motive to get additional revenue from money creation. See J. de Hann and D. Zelhorst (1990), “The Impact of Government Deficits on Money Growth in Developing Countries”, Journal of International Money and Finance, Vol. 9, N. 3. 107 M. Bleaney, N. Gemmel and D. Greenaway (1995), “Tax Revenue Instability, with Particular Reference to Sub-Saharan Africa”, Journal of Development Studies, N. 31. 108 D. Fielding (1997), “Modelling the Determinants of Government Expenditure in Sub-Saharan Africa”, Journal of African Economies, N. 6. 109 S. Feeny and M. McGillivray (2000), Aid, Public Sector Fiscal Behaviour and Developing Country Debt 110 V. Tanzi (1985), «Fiscal Management and External Debt Problems», H. Mehran (ed.), External Debt Management, IMF, Washington D.C. 167 more than half of the deficit in half the cases. This relationship between fiscal deficit and external debt seems to be very strong, even though a clear causality is far from evident. In fact, the link works in the opposite direction too: the abundance of international capital at very low cost pushed developing countries to maintain and expand public consumption and investment. This conclusion was confirmed by Zaidi, as referred to 20 developing countries111. Dornbusch112 explored the role of disequilibrium exchange rates and budget deficits in promoting external indebtedness and the 1980s debt crisis. He analysed the period 78-82 for Argentina, Chile and Brazil, studying the underlying disequilibrium in which exchange rate overvaluation and budget deficits were perpetuated by a continuing and excessive recourse to the world capital market. The increase in external debt in Brazil is largely attributed to failure to adjust the budget to the external shocks of high world interest rate and increase in real oil prices. Government subsidised the price of oil to maintain it low and also the government borrowed to finance the increased debt service. The budget deficit absorbed all the shocks. Tanzi and Blejer113 found that fiscal deficits had the strongest effect determining changes in external debt - particularly referred to debt from official sources - in a sample of 15 developing countries. Also Schmidt-Hebbel114 found a high correlation coefficient (-0.62) between fiscal deficit and external debt in a sample of 49 countries. Ojo115 found a statistically significant and positive relationship between external borrowing and budget deficit, analysing 39 Sub-Saharan African countries during the 1978-1981 period. The same result, referred to the Tanzanian case and referred to the 1970-1991 period, was showed by Kilindo116. Berthélemy117, referring to the poor heavily indebted countries, has tried to stress the importance of fiscal dimension of external debt. The idea is to demonstrate that the most adequate indicator of debt sustainability is the indicator that measures the cost of debt in terms of public revenues. The three debt sustainability indicators analyzed are (a) the debt-to-export ratio [DEXP], (b) the debt-to-GNP ratio [DGNP], and (c) the debt-to-government revenue ratio [DREV]. In algebraic terms, the authors estimate the following three regressions: INVFIPR = a0 + b1log(INTLEN) + b2HESLAG + b3log(DEXP) INVFIPR = a0 + b1log(INTLEN) + b2HESLAG + b3log(DGNP) 111 (5.17) (5.18) I.M. Zaidi (1985), “Saving, Investment, Fiscal Deficits, and the External Indebtedness of Developing Countries”, in World Development, Vol. 13, No. 5, May. 112 R. Dornbusch (1985), “External Debt, Budget Deficits, and Disequilibrium Exchange Rates”, in G. W. Smith and J. T. Cuddington (eds.), International Debt and Developing Countries, World Bank, Washington D.C. 113 V. Tanzi and M. I. Blejer (1988), “Public Debt and Fiscal Policy in Developing Countries”, K. J. Arrow and M. J. Boskin (eds.), The Economics of Public Debt, St. Martin’s Press, New York. 114 K. Schmidt-Hebbel (1995), Fiscal Adjustment and Growth: In and Out of Africa, Special Paper, N. 19, Afrfican Economic Research Consortium, Nairobi. 115 K. O. Ojo (1989), “Debt Capacity Model of Sub-Saharan Africa: Economic Issues and Perspectives”, Development Policy Review, Vol. 7. 116 A. A. L. Kilindo (1993), “Budget Performance and Foreign Indebtedness in Tanzania”, M. S. D. Bagachwa and A. V. Y. Mbelle (eds.), Economic Policy under a Multiparty System in Tanzania, University of Dar es Salaam Press, Dar es Salaam. 117 J. C. Berthélemy (2001), “HIPC Debt Relief and Policy Reform Incentives”, Unu/Wider Conference On Debt Relief: Helsinki, Finland 17-18 August. 168 INVFIPR = a0 + b1log(INTLEN) + b2HESLAG + b3log(DREV) (5.19) Where INVFIPR = the ratio of fixed private domestic investment to GDP, INTLEN = the nominal lending interest rate, HESLAG = the lagged real growth rate (based on PPP adjusted GDP). It turns out that - among the three debt sustainability indicators - the debt-to government revenue ratio (DREV) is the most significant determinant for private fixed investment, followed by the debt-to-GNP (DGNP) and debt-to-exports (DEXP). The author has also tested the robustness of the relative significance of the three debt sustainability indicators by substituting the three debt ratios (DEXP, DGNP, and DREV) with three corresponding debt service ratios: (a) the debt service-to-export ratio [DSEXP], (b) the debt service-to-GNP ratio [DSGNP], and (c) the debt service-to-government revenue ratio [DSREV]. Hjertholm, in analysing the fiscal dimension of SSA countries’ external debt, provides two models118. The first model is an error component specification of pooled time-series cross-sectional analysis applied to 23 countries and 5 years. The regression parameters are estimated using generalised least square (GLS) method, whereas the Fuller and Battese estimation method is used to add the individual country and time-specific effects to the random disturbances (error term [uit] is equal to a cross sectional component [vi], a time-series error component [et] and a random error component [ε it]). This model has the debt service performance, represented by the level of exceptional financing (corresponding to the discrepancy between actual and scheduled debt service payments) relative to the stock of total external debt, as dependent variable (EF). A group of independent variables is expected to have positive sign: the debt-to-government revenue ratio (DR), the debt-to-export ratio (DE), agriculture-toGDP ratio (AG). Another group of variables is expected to have negative sign: the international reserves-to-debt ratio (RD), the level of GNP per capita (GNP). Two variables are expected to have positive and negative signs as possible: import-to-GDP ratio (IG) and the share of concessional debt in total debt (CD). EFit = a0 + b1DRit + b2DEit + b3AGit + b4RDit + b5GNPit + b6IGit + b7CDit + vi + et + ε it The results confirm the hypothesis that the budgetary burden of debt contributes to the explanation of the debt service performance of SSA. In fact, DR has a low positive sign (+ 0.005) and is significant, whereas the conventional indicator DE is not relevant when DR is included. The correlation matrix of the estimated coefficients confirms the existence of a highly correlated relationship (-0.80) between DR and DE. Thus, external debt servicing capacity can not be separated from the issue of the government budget constraint, which is more important than export earnings. The second model tests the effects on gross private investment in percentage of GDP (GPI) of four groups of independent variables, in a debt overhang context. In fact, debt overhang 118 P. Hjertholm (1997), “An Inquiry Into the Fiscal Dimension of External Debt: The Case of Sub-Saharan Africa”, Ph.D. Thesis, N. 68, Institute of Economics, University of Copenhagen. 169 (5.20) theory assumes that investment (and thus growth) is hampered and discouraged by public debt (and, thus, debt relief can improve growth of debtor country). Public debt variables represent the first group, with expected negative signs: debt-to-government revenue ratio (DR), debt service-to-government revenue ratio (DSR). A development indicator represents the second group, with expected positive sign: the level of GNP per capita (GNP). Public investment indicator represents the third group, with expected positive sign: gross public investment-to-GDP ratio (PIn). Some macroeconomic instability indicators represent the fourth group, with expected negative signs: the standard deviation of annual inflation rates over the preceding four-years period (SDI), the standard deviation of annual changes in the real effective exchange rate over the preceding four-years period (SDE), the average annual ratio of exceptional financing in percentage of long-term public debt (EF). GPIit+1 = b1DRit + b2DSRit + b3GNPit + b4PInit + b5SDIi + b6SDEi + b7EFi + vi + et + ε it The model is conducted in two-steps. First, the model is run with GPI as dependent variable and various combinations of explanatory variables (and one or two-years lagged ratios), in order to determine the direct effects of public debt on investment. Then, the model is run with those (non-debt related) variables found to be significant determinants of GPI in the first step as dependent variables, and various combinations of debt and other variables to determine the extent of influence of public debt on these variables. This second step determines the indirect effects of public debt on investment. From the first step, two years lagged PIn is determinant of GPI, together with GNP (with limited magnitude) and combinations of one-year lagged DR and DSR. The influence of the variability of inflation is not relevant, whereas the variability in the real effective exchange rate is a relevant determinant of investment. Uncertainty about future debt servicing, derived from a huge amount of unpaid debt service obligations, seems to discourage investment. This step confirms that public debt burden has a direct negative effect on investment. In the second step, public debt is always a relevant determinant of GNP as the dependent variable, however it is more confused when we include both DR and DSR as regressors. The cash-flow effect (=DSR) and its overhang effect (=DR) are not cumulative effect on GNP. One-year lagged DSR is clearly (and more clearly than DR) associated with a crowding effect on public investment (=Pin), as a dependent variable. In such a way, public debt indirectly reduces private investment. With regard to two-years lagged public debt data, the results are not clear. High DR clearly leads to high level of unpaid debt service (=EF), as dependent variable, but again the effect of DSR tends to dominate over the DR effect, as reflected by the larger sized of the estimated coefficients. Also the variability of inflation (SDI) results to determine the level of exceptional financing and it is itself affected (when it is assumed as dependent variable) by the level of public debt. Finally DR has a significant influence on SDE, as dependent variable, whereas DSR is not so relevant. From the second step, external public debt seems to deter private investment also indirectly, by aggravating investment disincentives, provided that debt influences non-debt investment determinants. What this analysis is not able to demonstrate is the relative importance of the crowding-out effect, which can be addressed only in a country-specific context. 170 (5.21) 12. A basic model on the relationship between aid, growth, government expenditure and debt In 1999, Grosskurth proposed a very simple model to investigate the relationship between aid, growth, external debt and government expenditure119. Its simple structure notwithstanding, the model is useful because it tries to investigate the relationship between four dimensions of development, which have been rarely taken in account simultaneously as the only factors to be tested. In fact, the author examines four important characteristics of recipients of development aid and debt. The dependent variable of a model is the amount of ODA received as a share of its GNP in per cent. Explanatory variables include GNP per capita and the Human Development Index (HDI) as measures of poverty and development, and they are expected to have a negative sign. Per capita level of external debt as a share of GNP and of military spending as shares of GNP are independent variables to test the arguments of the critics of ODA, and they are expected to have a negative sign. The model uses cross-sectional data, to produce marginal extra precision (none of the variables involved are subject to high volatility over a few years across a larger set of countries). The data may suffer from a selection bias as it could be argued that countries that receive development aid are often tied to report national data to international agencies. A higher dependency on aid would then result in a higher likelihood of inclusion in the study. The set of data obtained also suffers from multicollinearity. The exogenous variables ‘HDI’ and ‘GNP’ are highly correlated (Pearson correlation: 0.748). This is partly due to the fact that income is a relevant component of the HDI. As a result of this the significance of the coefficients for ‘HDI’ and ‘GNP’ are likely to be underestimated. The model is: aidt = α + β 1 HDIt + β 2 GNPt + β 3 debtt + β 4 militaryt + ut (5.22) The null hypothesis is: β 1 = β 2 = β 3 = β 4 = 0 The alternative hypothesis is that the null hypothesis is not true, and therefore the model has explanatory power. After a test-run, Sierra Leone was excluded from the sample. Its high level of aid received (‘aid’=164.4% of GNP) strongly biased the regression. The exclusion of Sierra Leone significantly increased the explanatory power of the model, as captured by higher values of the ‘R-squared’ (adjusted) and ‘F’ statistics. The F-value more than doubled. No sign changes took place. Table 1 summarises the adjusted regression output. The null hypothesis may be rejected at the 0.1% level (F= 22.41). The explanatory power of the model is moderate (adjusted Rsquared= 55.7%). The coefficients of the variables ”HDI”, ”debt” and ”military” exhibit the expected orientation. Table 1. Summary of Main Regression Output Aid = 22.1 – 41.0 hdi + 0.00174 gnp + 4.58 debt + 1.24 military Predictor Coefficient t-Ratio (p value) 119 J. Grosskurth, (1999), “Where does development aid go?”, Student Economic Review 1999, Trinity College, Dublin. 171 Constant HDI GNP Debt Military F = 22.41 22.066 -40.995 0.0017369 4.575 1.2408 R-sq. = 58.3% 5.33 (0.000) -5.42 (0.000) 1.76 (0.084) 4.55 (0.000) 2.13 (0.037) R-sq. (adj.) = 55.7 The intercept value of 22.1 suggests a high level of aid being granted. However, the negative coefficient of ‘HDI’ (-41.0) is relatively large. The debt coefficient of 4.5 has to be seen in relation to an average level of debt of 1.02 times GNP. This makes the impact of debt on the level of aid received rather small compared to the HDI. The average level of military expenditure (2.67) has on average even less influence, with a coefficient of 1.2. Against prior expectations the coefficient for ‘GNP’ is positive. It is, however, very small (0.0017). Even the high average GNP of 1300.9 does not help GNP to overcome the status of the least influential variable with an average impact of 2.6 percentage points on aid. With a t-value of 1.76 (p=0.084) it is also the least significant variable. The variables ‘HDI’ and ‘debt’ are significant at the 0.1% level (tHDI= -5.42; tdebt= 4.55). The variable ‘military’ is significant at the 5% level with a t-value of 2.13. Applying White’s general heteroskedasticity test, the null hypothesis of no heteroskedasticity may be rejected at the 0.05% level. This significantly exceeds the relevant critical value of 31.32 at p=0.005. The Spearman rank correlation test suggests heteroskedasticity in the variables ‘HDI’ (t-value of the Spearman rank test = -2.97), ‘GNP’ (-3.96) and ‘debt’ (4.59). Thus, the value for the standard deviation and consequently, the t-ratios and the significance levels of the respective variables are biased. The size and the direction of the bias depend mainly on the exact relationship between the true values of the observed variable and the true variance. There are several potential causes of the heteroskedasticity. The low R2 of model suggests that an important variable might be omitted. Possible omitted variables include variables capturing the political situation, the occurrence of natural disasters and the degree of dependency on foreign trade. Another potential cause is outliers that strongly bias the regression. Mozambique is a likely candidate. However, the exclusion of Mozambique from the sample would only marginally reduce the level of heteroskedasticity. A closer look at the suspected relationship between the observed variable and its residual may help to transform the data in order to reduce the level of heteroskedasticity. In order to further examine the properties of the four explanatory variables each of them was individually used as regressor against ‘aid’. The coefficient of ‘HDI’ has very high explanatory power (R2(adj)= 39.5%). The ‘HDI’-model is very significant with an impressive F-value of 45.4%. Its coefficient slightly decreases, but it is still the single most useful variable to explain the amount of aid received. The coefficient for ‘GNP’ changes its orientation and is now negative. This makes ‘gnp’ the most unreliable variable of the model. The R2(adj) of 17.1% is disappointing, even though the model and the coefficient are significant at the 1% level. The influence of “GNP” on ”aid” might be more significant below a threshold level of roughly $1800 per capita. In order to check this, a Chow test was applied to test the 172 alternative hypothesis that there is a structural change in the aid-GNP relationship at a level of $1800. The null hypothesis that there is no structural change may be rejected at the 1% level. The F-value of 9.87 is significantly larger than the critical value of 4.98. For the set of 16 rich countries ‘gnp’ is not significant at the 10% level (p=0.158). For the 53 countries with a GNP smaller than $1800 the variable ”gnp” is significant at the 0.1% level. Its coefficient increases significantly to a value of -0.015. The rationale behind this characteristic might be that richer countries only get aid under exceptional circumstances. These might for example be political reasons or the occurrence of natural disasters like earthquakes and hurricanes. The amount of military expenditure does not have large explanatory power (R2(adj)=5.3%). However, both the restricted model and the coefficient of the variable ‘military’ are significant at the 5% level. The low explanatory power might be due to the low number of strategically important LDCs. It would be interesting to check the parameters of this variable during the height of the cold war. The regressor ”debt” is again highly significant while exhibiting an R2(adj) of 28.0%. The level of external debt seems to be an important factor. After all, debt is the one problem that can definitely be solved with money. The altruistic argument for ODA may very well be questioned. It is in this light that ongoing campaigns for debt relief as a means to achieve long-term economic development should be viewed with some sympathy. However, this variable should be interpreted with care. As ODA consists mostly of loans it would be logical to assume that the relationship is two-fold. A Hausmann specification test was applied to test the null hypothesis that there is no simultaneity problem concerning ‘aid’ and ‘debt’. The null hypothesis may be rejected at the 0.1% significance level. The consequence of this is that the estimated parameters are biased and not consistent. A simultaneous equation method of estimation would be more appropriate to estimate the relationships at hand. 173 Bibliography B. Abassi and R. J. Taffler (1982), “Country Risk: A Model of Economic Performance Related to Debt Servicing Capacity”, Working Paper N. 36, City University Business School, London. P. R. Agénor (2000), The Economics of Adjustment and Growth, Academic Press, San Diego. J. Aizenman and N. P. Marion (1993), “Policy Uncertainty, Persistence and Growth”, Review of International Economics, N. 1, June. J. L. Arcand and M. G. Dagenais (1995), “The Empirics of Economic Growth in a Cross Section of Countries: Do Errors in Variables Really Not Matter?”, CRDE Discussion Paper N. 4195, University of Montreal, October. P. Arestis and P. Demetriades (1997), “Financial Development and Economic Growth: Assessing the Evidence”, Economic Journal, N. 107, May. A. M. Assiri, R. A. Parsons and N. Perdikis (1990), “A Comparative Analysis of Debt Rescheduling in Latin America and Sub-Saharan Africa”, Scandinavian Journal of Development Alternatives, Vol. 9, N. 2-3. R. J. Barro (1991), “Economic Growth in a Cross Section of Countries”, Quarterly Journal of Economics, N. 106, May. R. Barro and X. Sala-i-Martin (1995), Economic Growth, McGraw-Hill, New York. M. Beenstock (1980), “Political Economy of Official Development Assistance”, in World Development, n. 1. A. Beltratti (1989), “Empirical Estimates of the Capacity to Repay a Foreign Debt: A Vector Autoregressive Methodology”, The European Journal of Development Research, Vol.1, N. 2. A. B. Bernard and S. N. Durlauf (1996), “Interpreting Tests of the Convergence Hypothesis”, Journal of Econometrics, N. 71, March. A. Berg and J. Sachs (1988), “The Debt Crisis. Structural Explanations of Country Performance”, Journal of Development Economics, N. 29. J. C. Berthélemy (2001), “HIPC Debt Relief and Policy Reform Incentives”, Unu/Wider Conference On Debt Relief: Helsinki, Finland 17-18 August. A. Bhaduri (1987), “Dependent and Self-Reliant Growth with Foreign Borrowing”, Cambridge Journal of Economics, N. 11. J. S. Bhandari, N. U. Haque and S. J. Turnovsky (1990), “Growth External Debt and Sovereign Risk in A Small Open Economy”, IMF Staff Papers, N. 37, Washington D.C. M. Bleaney, N. Gemmell, and D. Greenaway (1995), “Tax Revenue Instability, with particular Reference to Sub-Saharan Africa”, The Journal of Development Studies, Vol. 31, N. 6. P. Boone (1996), “Politics and the Effectiveness of Foreign Aid”, European Economic Review, Vol. 40. 174 E. Borensztein (1990), Debt Overhang, Debt Reduction and Investment: The Case of the Philippines, IMF Working Paper WP/90/77. E. R. Borensztein (1989), “Fiscal Policy and Foreign Debt”, Journal of International Economics, N. 26. E. Borensztein, J. De Gregorio, and J. W. Lee (1998), “How Does Foreign Direct Investment Affect Growth?”, in Journal of International Economics, N. 45. B. Bosworth and S. M. Collins (1999), “Capital Flows o Developing Economies: Implications for Saving and Investment”, The Brookings Institution, Meeting Draft, March 15. F. Bresolin (1990), “Il rimborso del debito estero dei paesi in via di sviluppo: un’analisi del periodo 1971-1986”, Economia Internazionale, N. 43, 2-3. C. Burnside and D. Dollar (2000), “Aid, Policies and Growth”, American Economic Review, Vol. 90, N. 4. C. Burnside and D. Dollar (1997), “Aid, Policies and Growth”, Policy Research Working Paper, N. 1777, World Bank, Washington D.C. F. Caselli, G. Esquivel anf F. Lefort (1996), “Reopening the Convergence Debate: A New Look at Cross-Country Growth Empirics”, Journal of Economics Growth, N. 1, September. P. Cashel-Cordo and S. G. Craig (1990), “The Public Sector Impact on International Resource Transfers”, Journal of Development Economics, N. 32. R. Cassen (1986), Does Aid Work?, Oxford University Press, Oxford. H. B. Chenery and P. Eckstein (1970), “Development Alternatives for Latin America”, Journal of Political Economy, N. 78, 4. G. Chambas (1986), “Les facteurs explicatifs de la Repartition de l’Aide en Afrique. Le cas Francais et Japanais”, Monde en Developpement, n. 53. W. Cline (1984), “International Debt. Systemic Risk and Policy Response”, Institute for International Ecoomics, Washington D.C. P. Collier and D. Dollar (1999), “Aid Allocation and Poverty Reducion”, Policy Research Working Paper, N. 2041, World Bank, Washington D.C. D. C. Dacy (1975), “Foreign Aid, Government Consumption, Saving and Growth in Less Developed Countries”, Economic Journal, N. 85. C. J. Dalgaard and H. Hansen (2001), “On Aid, Growth and Good Policies”, The Journal of Development Studies, Vol. 37, N. 6. J. de Hann and D. Zelhorst (1990), “The Impact of Government Deficits on Money Growth in Developing Countries”, Journal of International Money and Finance, Vol. 9, N. 3. S. Devarajan., A, S. Rajkumar and V. Swaroop (1998), “What Does Aid to Africa Finance?”, African Economic Research Consortium (AERC), Nairobi, mimeo. P. Dhonte (1975), “Describing External Debt Situation: a Roll Over Approach”, IMF Staff Papers, N. 22. 175 R. Dornbusch (1985), “External Debt, Budget Deficits, and Disequilibrium Exchange Rates”, in G. W. Smith and J. T. Cuddington (eds.), International Debt and Developing Countries, World Bank, Washington D.C. R. Dornbush (1983), “Real Interest Rates, Home Goods and Optimal External Borrowing”, Journal of Political Economy, N. 91, 1. J. M. Dowling and U. Hiemenz (1985), “Biases in the Allocation of Foreign Aid: Some New Evidence”, World Development, n. 4. L. Dudley and C. Montmarquette (1976), “A model of the supply of Bilateral Foreign Aid”, American Economic Review, n.1. W. Easterly (2001), “How Did Highly Indebted Poor Countries Become Highly Indebted? Reviewing Two Decades of Debt Relief”, Unpublished Paper, The World Bank, Washington D.C. J. Eaton (1989), “Foreign Public Capital Flows”, H. Chenery and T. N. Srinivasan (eds.), Handbook of Development Economics, North Holland, Amsterdam. J. A. Edelman and H. B. Chenery (1977), “Aid and Income distribution”, in J. M. Bhagwati (ed.), The New International Economic Order: The North-South Debate, MIT Press, Cambridge. S. Edwards (1984), “Ldc Foreign Borrowing and Default Risk. An Empirical Investigation, 76-80”, American Economic Review, N. 9. G. Feder and R. Just (1977), “A Study of Debt Servicing Capacity Applying Logit Analysis”, Journal of Development Economics, N. 4. S. Feeny and M. McGillivray (2000), Aid, Public Sector Fiscal Behaviour and Developing Country Debt, mimeo. T. Feyzioglu, V. Swaroop and M. Zhu (1998), “A Panel Data Analysis of the Fungibility of Foreign Aid”, World Bank Economic Review, N. 12. D. Fielding (1997), “Modelling the Determinants of Government Expenditure in SubSaharan Africa”, Journal of African Economies, N. 6. A. K. Fosu (1999), “The External Debt Burden and Economic Growth in the 1980s: Evidence from Sub-Saharan Africa”, Canadian Journal of Development Studies, Vol. 20, N. 2. C. Frank Jr. and W. R. Cline (1971), “Measurement of Debt Servicing Capacity: An Application of Discriminant Analysis”, Journal of International Economics, N.1. J. Frankel and D. Romer (1999), “Does Trade Cause Growth?”, American Economic Review, Vol. 89, N. 3. J. Frankel and A. Rose (1996), “Currency Crashes in Emerging Markets: Empirical Indicators”, NBER Working Paper, N. 5437, Cambridge. I. N. Gang and H. A. Khan (1991), “Foreign Aid, Taxes and Public Investment”, Journal of Development Economics, N. 24. 176 O. Garavello (1981), “Processi di sviluppo e dipendenza dai flussi esterni di capitale”, Rivista Internazionale di Scienze Economiche e Commerciali, N. 28. N. Gemmell and M. McGillivray (1998), Aid and Tax Instability and the Government Budget Constraint in Developing Countries, CREDIT Research Paper 98/1, University of Nottingham, Nottingham. W. H. Greene (2000), Econometric Analysis, 4th edition, New York, Prenctice Hall. K. B. Griffin and J. L. Enos (1970), “Foreign Assistance: Objectives and Consequences”, Economic Development Cultural Change, N. 18, 3. E. Grinols and J. Bhagwati (1976), “Foreign Capital, Savings and Dependence”, Review of Economic Statistics, N. 58. J. Grosskurth, (1999), “Where does development aid go?”, Student Economic Review 1999, Trinity College, Dublin. G. M. Grossman and E. Helpman (1991), Innovation and Growth in the Global Economy, MIT Press, Cambridge. P. Guillaumont and L. Chauvet (2001), “Aid and Performance: A Reassessment”, The Journal of Development Studies, Vol. 37, N. 6. H. Hansen and F. Tarp (2001), “Aid and Growth Regressions”, Journal of Development Economics, Vol. 64. H. Hansen and F. Tarp (2000), “Aid Effectiveness Disputed”, in F. Tarp (ed.), Foreign Aid and Development: Lessons Learnt and Directions for the Future, Routledge, London. V. A. Haijvassiliou (1993), A Simulation Estimation Analysis of the External Debt Crises of Developing Countries, Cowles Foundation Discussion Paper, N. 1057, New Haven, Yale University. P. D. Henderson (1971), “The distribuiton of Offical Development Assistance committments by recipient countries and by sources”, Bulletin of the Oxford University Institute of Economics and Statistics, n. 1. P. Hjertholm (1997), “An Inquiry Into the Fiscal Dimension of External Debt: The Case of Sub-Saharan Africa”, Ph.D. Thesis, N. 68, Institute of Economics, University of Copenhagen. P. Isenman (1976), “Biases in Aid Allocation Against Poorer and Larger Countries”, World Development, n. 8. R. Jha (2001), “Macroeconomics of Fiscal Policy in Developing Countries”, The Australian National University, Camberra, mimeo. H. Kharas (1984), “The Long-Run Creditworthiness of Developing Countries”, The Quarterly Journal of Economics, Vol. 99, N.3. A. A. L. Kilindo (1993), “Budget Performance and Foreign Indebtedness in Tanzania”, M. S. D. Bagachwa and A. V. Y. Mbelle (eds.), Economic Policy under a Multiparty System in Tanzania, University of Dar es Salaam Press, Dar es Salaam. 177 B. King (1968), “Notes on the Mechanics of Growth on Debt”, World Bank Staff Occasional Papers, N. 6, J. Hopkins Press, Baltimora. M. Knight, N. Loayza and D. Villanueva (1993), “Testing the Neo-classical Theory of Economic Growth”, IMF Staff Papers, N. 40, September. A. O. Krueger (1987), “Debt, Capital Flows, and LDC Growth”, American Economic Review, Vol. 77, N. 2, May. S. Lall and G. Perasso (1989), “Determinants of the Debt Problem in eastern and Southern Africa: A Statistical Analysis”, Rivista Internazionale di Scienze Economiche e Commerciali, Vol. 36, N. 10-11. C. Lancaster (1991), “African Economic Reform: The External Dimension”, Institute of International Economics, Washington, DC. L. Landau (1971), “Savings function for Latin America”, H. B. Chenery (ed.), Studies in Developmental Planning, Harvard University Press, Cambridge. R. Lensink and O. Morrissey (2000), “ Aid Instability as a Measure of Uncertainty and the Positive Impact of Aid on Growth”, The Journal of Development Studies, Vol. 36, N. 3. R. Lensink and H. White (2001), “Are There Negative Returns to Aid?”, The Journal of Development Studies, Vol. 37, N. 6. R. Levine and D. Renelt (1992), “A Sensitivity Analysis of Cross-Country Growth Regressions”, American Economic Review, Vol. 82, N. 4. A. Mayo and A. Berrett (1978), “An Early Warning Model for Assessing Development Country Risk”, S. Goodman (ed.), Financing and Risk in Developing Countries, New York. D. C. McDonald (1982), “Debt Capacity and Developing Country Borrowing: A Survey of the Literature”, IMF Staff Papers, Vol. 29, N. 4. R. D. McKinlay and R. Little (1979), “The Us Aid Relationship: a test of the recipient need and the donor interest models”, Political Studies. M. Michaely (1981), “Foreign Aid, Economic Structure and Dependence”, Journal of Development Economics, N. 9. J. Morisset (1989), “The Impact of Foreign Capital Inflows on Domestic Savings Reexamined: The Case of Argentina”, World Development, N. 17, 11. P. Mosley (1987), Overseas Development Aid: Its Defence and Reform, Wheatsheaf, Brighton. P. Mosley (1985), “Towards a Predictive Model of Overseas Aid Expenditures”, Scottish Journal of Political Economy, February. P. Mosley (1981), “Models of Aid Allocation Process. A comment on McKinlay and Little”, Political Studies, n.2. C. Mukherjee, W. Howard and M. Wuyts (1998), Econometrics and Data Analysis for Developing Countries, Routledge, London. P. Nagy (1978), “Quantitative Country Risk: A System Developed by Economist at the Bank of Montreal”, Columbia Journal of World Business, N. 13. 178 W. T. Newlyn (1977), The Financing of Economic Development, Clarendon Press, Oxford. C. Ngassam (1992), “The Empirical Determinants of Lending to Sub-Saharan Africa”, Dept. of Finance, University of Delaware, Newark, mimeo. C. Ngassam (1991), “Factors Affecting the External Debt-Servicing Capacity of African Nations: An Empirical Investigation”, Review of Black Political Economy, Vol. 20, N. 2. M. Obstfeld (1998), “Foreign Resource Inflows, Saving, and Growth”, in K. SchmidtHebbel and L. Servén (eds.), The Economics of Saving and Growth: Theory, Evidence and Implications for Policy, the World Bank and Cambridge University Press, Cambridge. M. O. Odedokun (1995), “Analysis of Probability of External Debt Rescheduling in SubSaharan Africa”, Scottish Journal of Political Economy, Vol. 42, N. 1. M. O. Odedokun (1993), “Econometric Analysis of External Debt Burdens of African Countries: Debt Rescheduling and Arrears of Interest”, African Development Review, Vol. 5, N. 2. K. O. Ojo (1989), “Debt Capacity Model of Sub-Saharan Africa: Economic Issues and Perspectives”, Development Policy Review, Vol. 7. D. Oks and S. van Wijnbergen (1995), “Mexico After the Debt Crisis: Is Growth Sustainable?”, Journal of Development Economics, Vol. 47. M. H. Pesaran and R. Smith (1995), “Estimating Long-Run Relationships from Dynamic Heterogeneous Panels”, Journal of Econometrics, N. 68, July. T. Persson and G. Tabellini (1994), “Is Inequality Harmful for Growth?”, American Economic Review, No. 84, June. P. Pesenti (1987), L’aiuto Pubblico allo Sviluppo: teorie e verifiche econometriche dei criteri di determinazione e di allocazione, CESPRI, Working Paper n. 5; Milan. D. T. Quah (1996), “Empirics for Economic Growth and Convergence”, European Economic Review, N. 40, June. H. Reisen (1999), “Sustainable and Excessive Current Account Deficits”, in J. Gacs, R. Holzman and M. Wyzan (eds.), The Mixed Blessing of Financial Flows, Edward Elgar Publ., Cheltenham. D. Rodrik (1998), “Trade Policy and Economic Performance in Sub-Saharan Africa”, NBER Working Paper N. 6562, May. P. G. Roeder (1985), “The ties that bind: aid, trade and political compliance in Soviet-Third World relations”, International Studies Quarterly, n. 29. J. Sachs, K. Botchwey, M. Cuchra, and S. Sievers (1999), “Implementing Debt Relief for the HIPCs”, Center for International Development, Harvard University, Cambridge, Mimeo. X. Sala-i-Martin (1997), “I Just Ran Two Million Regressions”, American Economic Review, Vol. 87, N. 2. A. Sengupta (1968), “Foreign Capital Requirements for Economic Development”, Oxford Economic Papers. 179 K. Schmidt-Hebbel (1995), Fiscal Adjustment and Growth: In and Out of Africa, Special Paper, N. 19, Afrfican Economic Research Consortium, Nairobi. R. D. Singh (1985), “State Intervention, Foreign Economic Aid, Savings and Growth in LCDs: Some Recent Evidence”, in Kyklos, N. 38 (2). R. Solomon (1977), “A Perspective on the Debt of Developing Countries”, Brookings Paper Economic Activities. R. Solow (1956), “A Contribution to the Theory of Economic growth”, Quarterly Journal of Economics, N. 70. M. Soto (2000), Capital Flows and Growth in Developing Countries: Recent Empirical Evidence, OECD Technical Paper, N. 160, Paris. I. O. Taiwo (1991), “Commodity Prices and Debt Crisis in Sub-Saharan Africa”, Eastern Africa Economic Review, Vol. 7, N. 2. Y. Takagi (1981), “Aid and Debt Problems in Less Developed Countries”, Oxford Economic Papers, N. 33, 2. V. Tanzi (1993), “Fiscal Deficit Measurement: Basic Issues”, in M. Blejer and A. Cheasty (eds.), How To Measure the Fiscal Deficit, IMF, Washington D.C. V. Tanzi (1985), “Fiscal Management and External Debt Problems”, H. Mehran (ed.), External Debt Management, IMF, Washington D.C. V. Tanzi and M. I. Blejer (1988), “Public Debt and Fiscal Policy in Developing Countries”, K. J. Arrow and M. J. Boskin (eds.), The Economics of Public Debt, St. Martin’s Press, New York. J. Temple (1999), “The New Growth Evidence”, Journal of Economic Literature, N. 37, March. M. P. Todaro (1989), Economic Development in the Third World, Longman, New York. B. Wasow (1979), “Saving and Dependence with Externally Financed Growth”, Review of Economic Statistics, N. 61. T. E. Weisskopf (1972), “The Impact of Foreign Capital Inflow on Domestic Savings in Uncerdeveloped Countries”, Journal of International Economics. H. White (1992), “The Macroeconomic Impact of Development Aid: A Critical Survey”, The Journal of Development Studies, Vol. 28, N. 2. E. R. Wittkopf (1972), “Western bilateral aid allocations: a comparative study of recipient state attributes and aid received”, AAVV, Sage Professional papers in international studies, SAGE, London. World Bank (1998), Assessing Aid: What Works, What Doesn’t and Why, The World Bank, Washington D.C. I. M. Zaidi (1985), “Saving, Investment, Fiscal Deficits, and the External Indebtedness of Developing Countries”, in World Development, Vol. 13, No. 5, May. 180 6. Exploring the complex relationship between foreign capital inflows and development in Africa 1. Introduction. Hypotheses to be tested Our basic idea is that the relationship between different channels of foreign capital inflows and development is complex by itself, and the existence of some interactions between these channels is part of the complexity, too. Moreover, Sub-Saharan African countries do not represent a homogeneous aggregate. Thus, heterogeneity across countries and conditional hypotheses of the effect of aid, debt and foreign investment are the restricted focus of our analysis. We think that this issue is very interesting, even though the translation of these ideas into empirical analysis is not easy and it was not adequately investigated. Most work in the growth literature relies simply on cross-country regressions. The reason is that many effects are long-term and cannot be found in the short-run effects of the time-series framework. Furthermore, the data for Sub-Saharan Africa might simply not be good enough for sophisticated econometrics. If we embark on the time series approach, we may simply analyze “what the World Bank random number generator has produced”. However, we don’t adopt cross-country regression as a reliable method of empirical argumentation relating to African development. We agree what Bhagwati and Srinivasan recommended1: the most compelling evidence on these issues can come from careful and patient in-depth case studies of individual countries. Weak theoretical foundation and specification of functional forms for the relationship, poor quality of data bases and errors and biases of measurement, inappropriate econometric methodologies are common problems. In the context of relationships that have both a temporal and cross-sectional dimensions, there is the problem that the estimated impact from a cross-section need not be the same as that from time-series data. Regressions and their conclusions are strongly dependent on the period, sample of countries, and variables chosen. Thus, we adopt two different techniques, in order investigate the heterogeneity across countries and the interactions across channels of capital inflows in terms of development. The first is an ideogram statistical technique. Another area of interest is pooled time-series. The proliferation of cross-country samples for econometric analyses is partly due to the scarcity of time series data available for adequate analysis, especially for the poorest African countries. Moreover, cross-country data allow us to identify aggregate relationships, or correlations in the data that appear to hold ‘on average’ over a wide sample. As we mentioned, the weakness of this approach is that what appears to hold on average is rarely an adequate explanation of what is happening in a particular country, given the importance of country specific context. Furthermore, the macroeconomic relationships of interest are typically dynamic in nature, so that we are interested in what is happening in countries over time. Thus, the pooled time-series and cross section technique provides us with a useful approach to treat these two dimension in parallel. Our basic idea is that the relationship between external debt – and, in general, foreign capital flows – and economic development is complex and non-linear. It does clearly depend on the specific country’s context (a cross-section component) and on the specific past history (a time-series component). Moreover, and we think that it does represent the bulk of this work and the implicit mainstreaming of all the chapters, we disagree with the conventional wisdom, which attributes to some kind of foreign capital inflows to be good or bad in itself for development. An implicit recommendation of this conventional wisdom is that a large 1 T. N. Srinivasan and J. Bhagwati (1999), “Outward-Orientation and Development: are Revisionists Right?”, September, mimeo. 181 share of FDI in total capital inflows is a measure of something good happening in the economy, as well as a large share of debt is bad. We think that the real effects are associated with the nature, size and composition of each capital flow, in addition to the cross-section and time-series components, and with the interaction component among different kind of foreign capital inflows. We believe that foreign capital inflows enter multiplicatively rather than additively into a development equation. It has direct policy implications. First, it is hard to argue that the rise in the share of external debt is an indication of good health. But this does not mean that the rise in debt is bad in itself. Hence, there is no reason to say that, in the next future, debt must be totally replaced by grants – as it was recently and authoritatively affirmed2 - or that in perspective FDI should be the bulk of foreign capital inflows, for the sake of private sector promotion. We discovered that, fortunately, two very recent papers share the same idea and strongly confirmed it econometrically, using common cross-countries regressions. These are the work of Ricardo Hausmann3, from the Harvard University, and of Henrik Hansen4, from the University of Copenhagen. Thus, we would like to enlarge the set of financial flows, which can be multiplicatively linked, in order to include the main international financial flows to SSA (debt, grants and Ide). We agree that reasonable levels of external debt and aid inflows are expected to have a positive effect on growth. In traditional neoclassical models, there is an incentive for capital-scarce countries to borrow and invest since the marginal product of capital is above the world interest rate. And it is true that experience showed that over-borrowing becomes a negative spiral and, if the costs of high taxes to service the debt are not internalized, both aid and new loans serve debt repayment without any positive impact on growth. Moreover, if the expectations are that future debt will be larger than the country’s repayment ability, the expected debt service will impose a high (and distortionary) marginal tax on investment and output level as well as the introduction of inflation-tax, the implication being that new domestic and foreign investment will be discouraged, even if we don’t consider the crowding out effect (debt overhang theory). It is very reliable that, in highly uncertain environments, investors will continue to exercise their option of waiting and will not realize long-term, high-risk, irreversible investment. High levels of external debt (and high levels of aid required to pay debt back) are expected to have a negative effect on growth, through the reduced efficiency of capital accumulation (i.e. investment). Thus, we can assume that there is a range of values after which the impact of foreign capital inflows on investment become negative. Given this premise, it should be possible to identify the optimal level of foreign capital inflow: a level after which the marginal impact of further capital accumulation becomes negative (the so called “wrong” side of the Laffer curve, in case of external debt, where increases in the face value of debt will lower expected debt repayment). Given these premises, it is clear that we want toto test the impact of aid and debt on growth through their impact on investment, analyzing how much they are level dependent and how complex is their interplay. This approach implies that foreign capital inflows may have nonlinear effects on economic development. 2 A. Meltzer (2000), International Financial Institution Advisory Commission Report, US Congress, Washington D.C., March. 3 Development Centre Seminars (2001), Foreign Direct Investment Versus Other Flows to Latin America, Oecd- Iadb, Paris. 4 H. Hansen (2001), “The Impact of Aid and External Debt on Growth and Investment: Insights From Cross-Country Regression Analysis”, WIDER Conference on Debt Relief, Helsinki, August. 182 In the present chapter the focus is first on the usage of an original ideogram statistical technique to describe the SSA sub-regional heterogeneity (that is the importance of the cross section component, which cannot be adequately captured by a Sub-Saharan African dummy). Then the focus is on the application of pooled time series and cross section model to show how external debt and foreign capital inflows affect development (investment and growth) in SSA. 2. An ideogram statistical technique. The “flowers” of SSA regional development 2.1 - The Flower ideogram of regional heterogeneity Through the usage of an ideogram, based on the flower symbol, we want to analyse and graphically describe the SSA sub-regional lack of homogeneity, in terms of economic variables, which are represented by the shape of petals and pistils. It aims at providing synthetic information on quantitative economic dimensions of some fundamental variables referred to some SSA sub-regions, compared to the EU area. It is a way to investigate and present data through pictorial representations, for immediate and easy reading. 2.2 - Considered SSA sub-regions We compare the benchmark of 15-members EU area with four African sub-regions (UEMOA – Union Economique et Monetaire Ouest Africaine -, CEMAC - Communauté Economique et Monetaire en Afrique Centrale -, EAC – Eastern African Community -5, SADC - Southern African Development Community -), which have been explicitly mentioned during the post-Lome negotiations as possible counterparts for the future EU-ACP trade agreements6. Current Cotonou agreement, agreed in 2000, explicitly mentions the importance of these sub-regions. Tab. 1 –SSA sub-regions defined by the EU-ACP Cotonou agreement LLDCs no-LLDCs UEMOA CEMAC EAC SADC Benin Burkina Faso Guinea Bissau Mali Niger Senegal Togo Cote d’Ivoire Chad Equat. Guinea Centr. Afr. Rep. Tanzania Uganda Cameroon Congo, Rep. Gabon Kenya Angola Lesotho Malawi Mozambico Congo, Dem. Rep. Tanzania Zambia Botswana Mauritius Namibia Seychelles South Africa Swaziland Zimbabwe 5 The EAC integration, launched in 1967, was aborted in 1977 after 10 years. Efforts to revive the community began in 1993 with the heads of state signing an agreement to establish a commission for East African cooperation and in November 1999 the principles for economic, monetary and political union were set out. In January 2001, the three East African heads of state inaugurated the East African Community. 6 Tanzania is the only country being in two regions: EAC e SADC. Ghana and Nigeria are the non-LLDCc excluded by these regions. And Nigeria is the main African trade partner for EU, which import basically oil. Sixteen LLDCs are excluded by this sub-regions: Horn of Africa, Rwanda, Burundi, Madagascar, Mauritania and some small countries. In Africa, the sub-regional architecture, apart from the EU pressure, is considered a way to gradually implement the Lagos Action Plan (1980), in order to create an African common market. 183 A first proxy of development level in these regions is provided by the Human development index (HDI), wich has been proposed, since the first Human Development Report, published in 1990, by the UNDP. This is a composite index based on three indicators: (i) longevity, as measured by life expectancy at birth; (ii) educational attainment, as measured by a combination of adult literacy (two.third weight) and the combined gross primary, secondary and tertiary enrolment ratio (one-third weight); (iii) standard of living, as measured by GDP per capita (PPP US$). With normalization of the values of the variables that make up the HDI, its value ranges from 0 to 1 – the highest possible value – and also allows for inter-country comparisons. There are countries in the high HDI category (“H”: with a value equal to or more than 0.800), in the medium HDI category (“M”: 0.55-0.799), in the low category (“L”: less than 0.500). Tab. 2 – HDI in SSA sub-regions UEMOA CEMAC SADC EAC Eight countries. Mean: 0.297 All the members are in “L” group Six countries. Mean: 0.439 Two countries are in “M”, four are in “L” group Fourteen countries. Mean: 0.516. Two countries are in “H”, five in “M”, seven in “L” group Three countries. Mean: 0.383. All the members are in “L” group From a commercial point of view, intra-African trade is still only about 7% of Africa’s total trade. That said, there are important sub-regional variations in the progress towards integration, as is well summarised in the report of a recent study by ECA and others7. In the east and south, SADC has undergone significant institutional changes since the 1990s, with potential positive effects. South Africa has joined SADC, and there has been a series of successful negotiating sessions aimed at reaching agreement on a SADC Free Trade Area following the ratification of the SADC Trade Protocol in 1998. Thus, The most dynamic are is the SADC, where almost all of the intra-area trade is due to the South AfricaZimbabwe bilateral flows. Moreover, in the SADC, the growing role of intra-area trade is direct result of a “political” change in the South African regime. Tab. 3 – Intra-area trade in SSA sub-regions (percent, 1995) UEMOA CEMAC EAC intra-area 8.7 3.2 7.9 With EU-15 49.6 52.0 36.3 With others 41.7 44.7 55.9 total 100 100 100 Source: IMF (1999), Direction of Trade Statistics, Washington D.C. 7 SADC 11.5 36.4 52.1 100 World Bank, African Development Bank, ECA, OAU and African Economic Research Consortium (2000), Can Africa claim the 21st Century, Addis Ababa. 184 2.3 - Considered variables We take in account ten variables, referred to the 1995-97 period, which was the period when European Commission negotiated with the ACP the future of Lome agreements, and we consider their arithmetic mean. The variables are: 1. population, 2. per capita GNP (corrent US$), 3. Annual percentage growth rate of per capita GNP, 4. Annual percentage growth rate of inflation, 5. Gross domestic savings as a share of PPP GDP, 6. Gross domestic investment as a share of PPP GDP, 7. Foreign direct investment net flow as a share of PPP GDP, 8. Trade in goods as a share of PPP GDP, 9. Budget deficit as a share of PPP GDP, 10. External debt as as a share of GNP. 2.4 - Comparison between pistils The first variables (population and per capita GNP) are in absolute terms (million of people) and in comparable relative wealth (per capita US$) and they make it possible to define market potentialities of sub-regions: population as a proxy of consumers and per capita GNP as a proxy of wealth. Figure 1 is a comparison between the weighted economic profile of different areas. We have data easily represented by circles (population is the grey circle, whereas per capita GNP is the black one). Figure 1 - The absolute weight of sub-regions Population Per capita GNP Cemac Sadc Eac EU Uemoa 185 The UE market is the most developed, richest and biggest market, and it is the natural benchmark for other sub-regions, which look at EU as the model to be reproduced. Thus, through normalisation of the values of the variables, we have considered the EU as the unit of measure. SADC seems to be the only SSA subregion with a potential significant market. 2.5 - Comparison between petals As a second and very different step we consider the other eight variables, represented by petals. For every variable (=vi) we get: i) The arithmetic mean of 1995-1997 period for each member country, ii) The weighted Arithmetic Mean (µ = Σ vi GDPi / Σ GDPi) of the sub-region, iii) The weighted standard deviation Statistics (σ = S-1/2 = Σ [vi - µ]2 GDPi / Σ GDPi) of the subregion. The usage of weighted values makes it possible to correct the summary descriptive statistics of mean and standard deviation through the effective weight of single economies within their regions. In fact, to consider the dynamics of German domestic investment equal to that of Greece makes no sense. The problem is simply that the relative percentages in such a context do not represent fractions of the same total, so they can't be added, unless we assign the correct “weight”, linked – for example – to the GDP. For every couple of the variables, we associate weighted arithmetic mean and standard deviation to get a diagram of Cartesian co-ordinates (or rectangular co-ordinates), given by two real numbers. These two numbers indicate the length of the perpendicular projections from the central point to two fixed, perpendicular, graduated lines, the x-axis and the y-axis, provided by weighted arithmetic mean and standard deviation. Then, we proceed to the parametric transposition of the ellipse, the dimensions of which are the weighted arithmetic mean and standard deviation, in order to get a petal. We find a position of the eight ellipses angles to the plane through a rotation of 45° one from another, in order to cover a circle. All the different flowers, representing different sub-regions, keep the same order of petals, in order to facilitate the visual comparison. The horizontal and vertical axes of flowers are defined by four domestic macroeconomic variables (n. 3, 4, 5, 6), whereas the diagonals are defined by budget deficit (n. 9) and three external variables (n. 7, 8, 10). Looking at the petal dimensions, its length measures the weighted mean of the sub-region, whereas its width measures the heterogeneity within the sub-region (i.e. the weighted standard deviation). Excepted the “bad” petals of external debt and inflation, which are dark petals, the more the petals are long and narrow (i.e. more homogeneous) the more the flower is “nice”. 186 Figure 2 - Petals of development Annual percentage growth rate of per capita GNP Trade in goods as a share of PPP GDP Foreign direct investment net flow as a share of PPP GDP Gross domestic savings as a share of PPP GDP per capita GNP Gross domestic investment as a share of PPP GDP Budget deficit as a share of PPP GDP External debt as as a share of GNP Annual percentage growth rate of inflation The symptoms of the flower illness are the dark petals as well as a withdrawn petal, which assumes a dark colour because it means that the region is experiencing a negative growth rate of the given variable. By assumption, normalisation of the values of the variables has been solved using the EU area data as the benchmark: the weighted mean of EU corresponds to 1 and its weighted standard deviation to 0.3. Moreover Thus the EU flower is by assumption the “nice” flower. The dark petal representing external debt is absent, as EU has zero external debt. We define a way to treat the “inflation” variable and to solve the problem of no-zero values for the EU (being equal to 2.5 percent) and very high values (numbers of two or three figures) for the SSA sub-regions, which should imply respectively the presence of “dark” petal for the EU flower and too much long petal for other regions. We don’t consider the level value, but the logarithm of the standardised value of this variable. Thus, we get no petal for the EU flower (as the standardized value of the level value of 2.5 percent is equal to 1, and the logarithm of 1 is zero) and treatable length of petal for other regions. 187 Figure 3 - The level of homogeneity of SSA sub-regions UE a n. a n. . . UEMOA EAC CEMAC SADC It is quite evident the heterogeneity of SSA sub-regions, compared to the EU shape. In order to make the comparison immediate, we have given the same radius to all the regions, provided that the variables are relative ones, but we remind that the absolute values are those in figure 1. Every petal results from two axes with their linear scales. By comparing EU and EAC, the EAC petal of “annual percentage growth rate of per capita GNP” is twice as long as the EU petal: it means that the EAC relative value of this variable is twice the EU value. Obviously, it does not mean that EAC economy is growing twice the amount of EU, as it does not incorporate the absolute dimension of economies. Moreover, the EAC sub-region is the region with the best result, in terms of width: the petals are quite narrow, and it means that the homogeneity is higher. But it is simply due to the fact that the number of countries in the EAC region is very small (only three countries), compared to the other regions, and we did not include a correction weight for the number of countries. Obviously, the degree of freedom is relevant in terms of probability of dispersion of values around the group mean. Looking at the SADC sub-region flower, we have to mention another contrivance we have used. We got some weighted standard deviation values that are too high to be reproduced in the picture. Thus, we have arbitrarily fixed a limit of the width of petals and we used a “de-framing” technique (the petals with the frame that has partially taken away mean “exceptional” values) for investment, savings, external debt and 188 inflation variables. Nevertheless, SADC confirms to be the sub-region with a potential market, as its shape is similar to a “nice” flower, but its necessary enlargement process – up to 14 countries - is at the cost of heterogeneity. In fact, all the petals are quite wide8. What seems to represent a common characteristic all over the sub-regions is the presence of a long dark petal of external debt. Thus, notwithstanding the clear heterogeneity of economic conditions – confirmed by the studies by Easterly and Levine (1997) and Temple (1998), who point to persistent correlations between macroeconomic policy indicators and country specific, cultural, and socioeconomic characteristics in developing countries -, all the SSA sub-regions and countries exhibit an overwhelming degree of similarity in terms of general external debt problem. 2. Pooled time-series cross-section approach The need for a pooled time-series cross-section model Most of the studies have attempted to identify changes in the development patterns within the structure of a cross-sectionally estimated growth model. They analysed differences in point-to-point (or averaged periodto-period) patterns as a result of different socioeconomic conditions existing among various countries at a particular point (or period) in time. Cross-sectional models do not provide insights into structural changes over time. Neither do they reflect sudden changes due to dynamic exogenous shocks. Cross-sectional analysis, dealing with only one time period and yielding different results in different periods can not lead to generalisations of the dynamics of development. Thus, the use of single cross-country regressions is inappropriate, when: (1) we have to consider dynamic dimension as well, (2) single cross-country regressions suffer from omitted variable bias, (3) one or more regressors may be endogenous. For these reasons, time-series analysis is appropriate for the estimation of the economic aspects of growth: it allows the tracing of changes over time in the underlying socioeconomic conditions. Moreover, time-series analysis make possible to include lagged dependent variables in the model equations. From the analytical point of view, including the lagged dependent variables in the model equations allows growth to be seen as an essential dynamic process involving lengthy responses lags where the transmission process is not instantaneous. From the aspect of model building, to include the lagged dependent variables in the equation provides a better model specification, as they pick up the effect of unmeasured factors (i.e. omitted variables) that may have affected development in the previous period and were not included in the model specification9. This point is critical. Usually, panels have large number of cross-sections, with each unit observed only a few times. In panel models, with small T there is no hope of saying anything about the time series structure 8 Southern Africa has joined the ranks of the world’s leading trading emergent areas. The Southern African Customs Union (SACU) - member countries of SACU are: Botswana, Lesotho, Namibia, South Africa and Swaziland - figures among the world’s top 15 exporters for five of the 14 sectors covered by the Trade Performance Index. The Trade Performance Index ranks the export performance of 184 countries in 14 export sectors, placing export sectors in all countries on a global competitiveness ladder. The Trade Performance Index, launched in February 2000 in Bangkok at UNCTAD X, is based on 1998 export performance, as well as shifts in export performance between 1994 and 1998. Transport equipment is one example of SACU performance, where it ranks ninth, with exports of US$ 1.4 billion and a high degree of product and market diversification. South Africa, SACU’s largest member, has experienced growing interest of transnational corporations to invest in the country, as shown by UNCTAD's survey on foreign direct investment in Africa. 9 We know that (permanent) unobserved country specific effects are likely to be correlated with some of the observed regressors, such as the initial level of GDP. As Hsiao (1986) shows, omitting unobserved time invariant country effects in a dynamic panel data model will cause OLS level estimator to be biased and inconsistent. And the lagged dipendent variable is positively correlated with the permanent effect. 189 of data and the units are just a sample extracted from a population. These assumptions can not be assumed in most of pooled analysis, when data have reasonable sized T and not very large N that are specific units (countries, in our case). Differently from panel data, the relevant issue of time (the within-country variability of data: how foreign capital inflows affect investment and growth over time) and heterogeneity of units (between-country variability of data: how countries with different levels of foreign capital inflows experience different investment and growth patterns) should be seriously taken in account in pooled time-series crosscountry analysis. Different Model specifications Our hypothesis to be tested is that heterogeneity across Sub-Saharan African countries is very important, as our graphical investigation shows, despite the fact that Africa is often considered an homogenous aggregate, and that the effects of aid, foreign debt and foreign investment on development are more complex than additive and linear. Results are very sensitive to the econometric approach used. Most of the studies rely heavily on the OLS method of estimation; all of the results reported in Assessing Aid (World Bank, 1998) are obtained using OLS. In the case of the basic pooled analysis, it is assumed that the parameters are constant, independently from both the countries and the years. For this reason, it is possible to combine the observations and to calculate only one regression for all the units in every year, rather than to estimate one given cross-sectional equation for every considered year or, in alternative, one time series for every country. The general framework for this regression model has the following form: yit = α i + βxit + uit for x =1,2,…, K; i =1,2,…,N; and t = 1,2,…,T (1) N and T are the cross section and time series dimensions respectively, α i is a scalar being a given constant term (differences across units are not captured in differences in the constant term), β is the (k * 1) parameter vector of xit, the vectors of k regressors (k explanatory exogenous variables). The vector of disturbance terms uit is assumed to be uncorrelated with the xit’s and the α i’s have zero mean and constant variance. This model restricts the coefficients on x to be common across i and t. The assumption made about α i has implications for the consistency and efficiency properties of estimates of β in equation (1). The model is transformed into a fixed effects (FE) model, or the so-called LSDV (Least Squares Dummy Variable), if we assume a variation of the intercept from section to section. The dummyvariable coefficients measure the change in the cross-section and time-series intercepts (with respect to the first individual in the first period of time) and account for the effects of the omitted variables that are specific to individual cross section units but stay constant over time, and the effects that are specific to each time period but are the same for all cross section units. The FE model is a first step to take into account heterogeneity, but it does not directly identify what causes the regression line to shift over time and over individuals. This lack of knowledge about the model can be described through the disturbance term, which is always the component that determines, with its statistical characteristics (in terms of its nature and structure), the performance of estimation procedures. The assumption is that the unit specific effects cannot be observed or measured and they must be treated as part of our “general ignorance”. What this means is that the large number of factors that affect the value of the dependent variable, but which are not explicitly accounted for in the model, can be summarized by a random disturbance. Heterogeneity across units can also be accounted for by treating the individual specific effects as random variables. If this assumption is made then 190 we have a random effects (RE) model, or the so-called Error Component Model (ECM)10. It attributes a random nature - absorbed by the residual term - to the variation in the largeness and direction of the sectional relations. The Error Component Method is used to allow for the effects to be random and to improve the efficiency of least squares estimates by accounting for the cross section and time series related disturbances. Under the error components specification, the uit disturbances takes the decomposed form: uit = v i + wt + zit (2) where the v i [∼ N(0, σv2)] are cross section specific error component, wit [∼ N(0, σw2)] are time-series specific error component and zit [∼ N(0, σz2)] is the combined error component. This formulation is derived from the FE model, by assuming that the mean effect of the random time-series and cross-section variables is included in the intercept term, and the random deviations about the mean are equated to the error components, v i and wt respectively. This model assumes that the pattern of shifting regression intercepts follows a normal distribution. Usually, we just do a simple transformation on the observations to make the resulting variance-covariance matrix of the errors satisfy the Gauss-Markov assumptions. Thus, Cochrane-Orcutt transformation to eliminate serial correlation is almost Generalised least squares (GLS)11, as is Weighted regression (WLS) to eliminate heteroscedasticity. The Seemingly Unrelated Regression (SUR) method, also known as the multivariate regression, or Zellner's method, estimates the parameters of the system, accounting for heteroskedasticity, and contemporaneous correlation in the errors across equations, which may affect our panel: SUR-GLS method is much more sensitive to individual contribution (cross-country heterogeneity), that is the critical assumption of pooling12. It is estimated using a variant of Generalised Least-Squares (GLS) regression. The estimation weights observations in inverse relationship to their variances and a twostage estimation process is used. In the first stage, OLS is run on the entire pooled sample. The OLS residuals are then used to calculate sample estimates of the variance components. The estimated variances are used in the second stage, in which the GLS parameter estimates are obtained. This method takes care of the twin problems of auto correlation among the time series disturbances and the heteroskedasticity of cross section disturbances. The idea of GLS is very simple: it is basically a procedure to weighting units by how well they fit the underlying regression, and so is simply downweighting those that fit poorly. We assign more weight (importance) to observation with smaller variability, that is more closely clustered around their mean values. Obviously, SUR method has some serious limitations. SUR covariance estimator is not consistent if any of the right-hand side variables are correlated with the residuals (that is they are endogenous). In this case, 3SLS is a system method that estimates all the coefficients of the model to obtain a consistent estimate derived from SUR. 3SLS is derived from Two stages least squares (TSLS), a special case of Instrumental variables regression. If we think that some Xs are endogenously determined with Y, then OLS estimates will be biased and inconsistent, if this were the case. Thus, TSLS specification means that we test using instrumental variables correlated with the suspect endogenous Xs, but not with the residuals of equation and then we run two OLS: (1) we regress the suspect endogenous Xs on the contemporaneous values of 10 G. S. Maddala (1971), “The use of variance components models in pooling cross section and time series data”, Econometrica, Vol. 39, No. 2, pp. 341-358. 11 As we don’t know the covariance matrix of the errors, we estimate it by residuals computed from OLS estimates of coefficients: this procedure is the Feasible GLS method. 12 A potential problem associated with SUR estimation is that the GLS procedure, which is used and is called Parks, provides wrong standard errors if T is very small. See Beck, N. and J. N. Katz (1995) “What to Do (and Not to Do) with Time-Series Cross-Section Data”, American Political Science Review, Vol. 89, 634–647. 191 exogenous Xs and on the lagged values of the endogenous regressors themselves (used as instruments)13; (2) we re-estimate Y fluctuation including the fitted values from the previous regressions. If OLS are consistent then the coefficients on the (1) residuals are not significantly different from zero. This is a way to solve contemporaneous correlation due to error of measurement (OLS estimators are biased and inconsistent) or simultaneous equation bias (we have endogenous variables – values of which are determined by the system we are observing – and exogenous variables)14. In order to control for heteroscedasticity and contemporaneous correlation (similar to SUR) as well as for endogeneity (similar to TSLS), 3SLS adopt the two stages of TSLS and, in the third stage, it applies feasible GLS to the equations in a manner analogous to the SUR-GLS estimator. From a broader point of view, we have to remind that single-equation models assume that a set of various explanatory factors determine a dependent variable, with no feedback relationship. That is to say: they hypothesize causality in one direction and explain neither the interdependencies that exist between the explanatory variables themselves nor how these variables are related to other variables. This is a problem when the parameter estimates of these models possess a simultaneous-equations bias that may vitiate the findings. Because of high interdependence between the causes and the consequences of growth, the endogeneity of many determinants should push research in the direction of a simultaneous-equations framework. But, reviewing literature, very few studies have controlled for the endogeneity of the regressors: for a dynamic panel data model OLS estimation are likely to suffer from biases due to endogeneity and unobserved country specific effects. Tab. 4 - The matrix of model specifications given different combinations of assumptions Gauss-Markov assumptions Heteroscedasticity Contemporaneous correlation Serial correlation Exogeneity OLS WLS SUR Endogeneity OLS-IV 2SLS 3SLS Differencing, AR, lagged Y, VAR, VECM The model specifications to be tested A preliminary analysis demonstrates that investment is highly significant in all the growth regressions15. The derived idea is to test that investment appears to be the main channel through which external capital promote development. Thus we run new regression with investment as dependent variables and foreign capital inflows as regressors. We remind that our hypothesis is that external financial flows are interlinked in their impact on development, and their effects are level dependent (in terms of current levels of different financial flows), temporal dependent as well as spatial dependent. Thus, in addition to simple linear regressions, we employ other specifications to investigate the non-linearity relationship between foreign capital inflows and investment. These specifications include a model with interaction terms and a quadratic specification. The quadratic specification would support a debt (or aid) inflows and investment Laffer curve relationship if the coefficient 13 We use the lagged endogenous variable as an instrument, because it is correlated with the endogenous variable and not with the residuals. 14 This way is called the Reduced form of the model, as we solve the system for endogenous variables, in order to know which disturbances influence endogenous variables. 15 The majority of current studies find that the most robust determinant of economic growth is investment. The new growth theory further emphasises the role of investment in the growth process. See D. Asterious and S. Price (2000), “Uncertainty, investment and economic growth: evidence from a dynamic panel”, City University, mimeo. 192 of debt (or aid) is positive and the coefficient of debt (or aid) squared is negative. In terms of problems, combining both temporal and space analysis, like the pooled analysis does, the different problems of these two econometric techniques are combined as well. In particular, we have to take care of: • the risk of serial correlation between the data of various periods of time; • the heteroschedasticity risk; • the bad specification of the model. (a) OLS technique Our basic estimation includes a OLS technique, which does some average of longitudinal and cross section specific effects (it averages the two dimensions: space and time). Thus, OLS estimation is between the between and the within estimation. When we keep the mean in using mean value between the variables, it doesn’t say anything about variation, and our pooled time series has a lot of variation. Thus, we lose a lot of information. If we correct for both cross-section (=countries) and time (=years) means, we get rid of a lot of information, both across time and space, simply by including time and period effect dummies. In these case, we expect to have little left to explain: total variation explained by model, including dummies, is extremely inflated. Moreover, OLS is optimal only if the model specification satisfies the Gauss-Markov assumptions. All errors must be independent and identically distributed. It assumes that all differences between countries are accounted for by differences in the independent variables (that is, non unmodeled heterogeneity), no effects of other countries on each other (no spatial effect), all countries obey same equation (pooling homogeneity), no dynamics (temporal dependence). The OLS errors are wrong if the errors show any of panel heteroscedasticity, contemporaneous correlation of the errors, serially correlated errors. Our hypothesis is that OLS specification is not appropriate and, as heterogeneity is the main characteristics of our analysis, other specifications are more appropriate. (b) Fixed Effect technique We extended the benchmark OLS model, adding country effects and also time effect (time dummies). With FE model we have different constant for each country, that is we subtract the “within” mean from each variable and estimate OLS using the transformed data. If intercepts vary by country, slope are invariant. With Fixed Effects we have some advantages, as we take in account parameter heterogeneity and we get rid of possibly distorting group effects, as we focus on within group variation. But we have some disadvantages, too: it eats degree of freedom and doesn’t involve theory in explaining phenomena. Obviously, to find substantive variables that explain the differences is much better16. FE model is useful if we have a handful of countries or years which have meaning by themselves in theory. We assume that in comparing the unrestricted model (with dummies) to OLS restricted model, there is some improvement, as the residual variation will be lower, which means that OLS did not account for the variation captured by some variables – it put them, as the country-specific intercept, equal to zero –. (c) Seemingly unrelated regression (SUR) technique Usually we try to correct the errors that show any of heteroscedasticity, contemporaneous or serial 16 That’s why Random error (RE) model is so attractive when one doesn’t know of any explanatory variable which explains the fixed effects: this model allows intercept terms to vary as well as he slope coefficients to vary from country to country. 193 correlation problems. As we mentioned, the Seemingly unrelated regression (SUR) is the feasible GLS specification correcting for both heteroscedastic and contemporaneously correlated residuals. This specification uses the OLS residuals to obtain a consistent estimate of the cross-section residual covariance matrix. In our context, we argue that the SUR method is more appropriate estimation method, than the previous ones. Moreover, concern about simultaneity bias in foreign capital inflows-growth regressions, caused by potential endogeneity of foreign capital, is another common feature in recent studies, as it is difficult to perceive of foreign capital as a lump-sum transfer, independent of the level of income. If foreign capital inflows depend on the level of income it cannot be exogenous with respect to growth as traditionally assumed. A two-way interaction, due to non-clear causality, creates an endogeneity that may lead to biased coefficient estimates when capital inflows are used as an explanatory variable. A domestic shock that raises the return to capital may increase both capital inflows and investment. This would tend to bias the coefficient on capital inflows in an investment equation upwards. In order to uncover the effect of capital inflows on investment, we use instrumental variables to isolate the flows that are related to exogenous factors. The problem is solved by using lagged observations of the regressors as instruments. If we assume significant dynamic effects of capital inflows in the growth relation, then we augment the growth specification and include lagged values of explanatory variables. But we need also to be attentive to the econometric issues that arise from the addition of the lagged dependent variable, in the right-hand side of the investment regression. This is referred to in the econometric literature as the dynamic panel data problem. In fact, on the lags’ length, Granger admonishes that “using data measured over intervals much wider than actual causal lags can destroy causal interpretation”17. It has been established that the lagged dependent variable is correlated with the error term by construction, rendering the OLS estimator biased and inconsistent. To tackle this problem, we follow the basic idea of Anderson and Hsiao18, who propose a solution that requires first-differencing all variables and using the lagged differences as time-varying set of instruments19. In fact, in order to control for the contemporaneous covariance between first differences (which eliminate the time-invariant individual effects) of regressors and residuals, they propose to instrument the offending terms by the second lagged level. Thus, their regressors (particularly, the second lag of the dependent variable) is not contemporaneously covariant with the equation error. As mentioned by Nat Beck20, we can easily model dynamics and correct for serial correlation with a lagged dependent variable, which is simpler to estimate and interpret than serially correlated errors - i.e. AR(1) -. Concerning the investment regression, we think that the model specification is not affected by the risk of 17 C. W. J. Granger, (1987), “Causal Inference,” The New Palgrave: Econometrics, W.W. Norton, New York, p. 49. 18 T. W. Anderson and C. Hsiao (1981), “Estimation of dynamic models with error components”, Journal of the American Statistical Association, Vol. 76. 19 Also timing in the regressions using differences has been addressed by time-series econometricians and for macroeconomic data the autoregressive distributed lags (ARDL) model, in which lags of the dependent variable as well as the explanatory variables are included in the regression, have proven to be quite useful. We have to experiment with lags of the growth rate and the explanatory variables in the regressions. Another similar estimator that produces consistent estimates, but also efficient estimates, is the General method of moments (GMM) estimator, used by Arellano and Bond, which uses all feasible lags of the dependent variable plus current and lagged values of exogenous variables as instruments. See M. Arellano and S. R. Bond (1991), “Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations”, Review of Economic Studies, Vol. 58. 20 Nathanial Beck (1998), “Taking Time and Space Seriously (particularly in Comparative Politics and International Relations): A Short Course on Time-Series-Cross-Section Data”, The Political Methodology Video Series, Richard J. Timpone, Series Director, The Ohio State University. 194 endogeneity, thus we don’t need to run a 3SLS specification, based on the use of some instrumental variables. 3. Pooled time-series cross-section data description and hypotheses The purpose of the present analysis is to take a closer look at the impact of foreign capital inflows (aid, debt and FDI) on economic development in SSA. We wish to make some general statements about the relation between the variables we are considering, looking at a limited set of countries provided by 30 SSA countries. Our time periods is determined by data availability from the main sources. The main data source are the World Bank datasets21. We work with data from 1971 to 1999. We consider a large set of variables. The ways in which the regressions are specified draw on inspiration from Burnside and Dollar (2000) and Hansen and Tarp (2001). The annual growth rate in real GDP per capita (= GDPgt) is chosen as the dependent variable, in the first set of regressions. It is a readily available output measure and we can draw on a well-established approach to conduct our analysis, as it (and the ratio of other variables to GDP) does correct for countries and economies lack of homogeneity. Referring to the independent variables, first we have the measures of foreign capital inflows, the set of variables of interest. Concerning the measures of external debt, related to GDP (in percent), we distinguish between the resource flow that leads to changes in total external debt, and the debt stock per se. Therefore the initial level of total external debt to GDP (= EDT1970/GDP1970) is used as a measure of the debt stock. The hypothesis is that it appears with a significant and negative impact on growth: the resulting impact of external debt on growth can be related to the standard inverse U-shape found in many studies of external debt problems and to the hypothesis of negative effect of high external debt stock on growth. The resource flow generated by external borrowing is calculated as the annual change in the stock: D(EDT/GDP)t= [(EDT/GDP)t+1–(EDT/GDP)t]. For countries experiencing negative growth the measure of external debt will increase even without increases in total external debt. The advantage of splitting total external debt into an initial debt stock and a flow is that the flow must be expected to be endogenous. Furthermore, there is no ambiguity regarding the expected impact of either of these series on the growth rate in GDP per capita as well as on investment. A positive flow of external funds should have a non-negative impact while the impact of initial external debt should be non-positive. We also include Total debt service as a percentage of GDP (= EDSt/GDPt). This is the sum of principal repayments and interest actually paid in foreign currency, goods, or services, and we expect a negative impact on growth and investment. For the aid flows we use Official Development Assistance in current US$ as a percentage of GDP (= ODAt/GDPt), and we expect a very limited positive impact on growth and investment. As we are considering SSA countries, we suspect that the well-known issue of impact of aid on growth is misleading. Referring to poor countries, most of aid is devoted to interventions for humanitarian actions and pro-poor safety nets and for external debt service repayment. We suspect that it is very unlikely that we can find any economic significant relationship between aid and GDP growth. In Africa, the majority of people depend on informal sector’s economy, without any immediate link to monetary market economy and GDP. When aid was targeted towards the poor, there was no clear effect on growth, being the macroeconomic 21 World Bank Africa Database, Global Development Finance Database, and World Development Indicators Database, which provide indicators covering SSA countries from 1965 to 1999. 195 environment positive or not. Aid to poor has been basically devoted to support consumption of the poor: only in recent years, the approach of introducing strategies of pro-poor-growth defined interventions to promote investment through a direct targeting of the poor. If we want to demonstrate how much aid did promote poverty reduction and growth in Africa during the last thirty years, we have to recognize that aid have had two different aims. Aid to reduce poverty, without positive investment effects; and aid to support investment and growth, without trickle-down effect on the poor and with very limited results. A third variable measuring an external resource inflow is foreign direct investment as a percentage of GDP (= FDIt/GDPt), and we expect a limited positive impact on growth and investment. As we mentioned, foreign investment represents very marginal capital inflows to SSA, very dependent on the natural endowment of countries (i.e. reflecting their heterogeneity). The majority of studies are not sufficiently informative about the interaction between different components of external capital inflows. We think that the effects are complex and varied, but that aid tends to be associated with other external capital inflows. Thus, the aid (ODAt/GDPt2) and debt (FLTDt/GDPt 2) squared terms (with negative coefficients) allow for the possibility that there may be diminishing marginal returns to aid and debt, i.e. that at some point the impact of additional aid and debt on economic development falls as the volume of aid and debt increases. A consequence of this formulation is that efficiency of aid and debt on development initially rises (other things held constant), up to a turning point at which the efficiency begins to fall. What this reflects are absorptive capacity constraints. The interactive term between aid and debt [(ODAt/GDPt)*(EDTt/GDPt)] needs to be interpreted with some caution, because it can mean both that the impact of aid on development increases with the quantity of debt, and that the impact of debt on development increases with the quantity of aid. Thus, the inclusion of an interaction term between foreign aid and economic policies and their squared terms represent the specification to capture non-linearity in the aid and debt-growth relationship22. Then, there is a vector of control variables that often appear in the regressions and are based on past empirical studies and economic theory. The first fixed variable, introduced by Barro and Sala-i-Martin (1992), is the logarithm of the initial level of per capita GDP (=logGDP1970). Usually, it is expected to have a significant negative influence on the growth rate, capturing the conditional convergence effect23. In our specific (SSA) and heterogeneous context, we suspect that this relationship is not so clear. Another variable chosen is the average population growth rate (= POPgt). It measures the demographic 22 The decreasing marginal returns can in empirical work be approximated by a second order polynomial in and debt in the regression. In general, quadratic terms and interactions represent separate terms in a second order approximation of what is really an unknown functional form. A more precise, second-order Taylor approximation of the theoretical growth equation leads to an empirical reduced form where quadratic terms as well as an interaction term (e.g., a cross product) are present in a linearisation (linearising converts the theoretical growth specification into an empirically manageable reduced forme quation) of a standard Solow model with convergence effects. Hadjimichael et al. and Durberry et al. studies include only the squared term, others only include the aid-policy interaction term to capture polynomial effects in the aid-growth relationship. There are, of course, more complex non-linearities that can be considered, other than quadratic form. 23 The conventional neoclassical growth model has most often been used in the study of growth until recently. The model suggests an inverse relationship between a country’s per capita growth in income and its initial level of per capita income. This result hinges on the assumption of diminishing returns to capital. Countries with low capital to labour ratios will therefore have higher marginal products of capital, enabling them to grow faster. Conditional convergence suggests that poor countries tend to grow faster than richer countries holding steady state per capita income constant. This motivated Lucas (1988) and Romer (1986), among others, to develop new growth models in which the sources of technological (and productivity) progress are endogenized rather than assumed to be exogenous. 196 constraint and it is expected to have a negative influence on the growth rate. Again, the SSA context induces us to doubt the reliability of this relationship. A very important variable is investment. Among others. De Long and Summers (1991) show that investment in equipment and machinery is the most important factor in influencing growth. In fact, the average investment rate is one of the variables that is most widely used in the literature (and one of the fixed variables in the Levine and Renelt paper). But investment has increasingly been seen as endogenous to growth, and therefore part of what needs to be explained rather than part of the explanation24. Moreover, if foreign capital inflows affect growth through investment, once investment is included as regressor, then the foreign capital inflows variables can become insignificant25 Thus, we include total gross domestic investment as a percentage of GDP (=INt/GDPt) and it is expected to have a significant positive influence on the growth rate, but we also estimate equations in which investment is excluded to control for the impact of foreign capital inflows and, above all, equations in which investment is the dependent variable and foreign capital inflows are the regressors. In such a specification, where INt/GDPt is the dependent variable, we can directly measure the impact of foreign capital inflows on the main determinant of growth. Linked to investment is savings rate. Notwithstanding the problems of measurement, as also savings figures must be considered tentative, because they are derived as a residual in the national accounts from expenditure and production data that are themselves quite unreliable, it is quite important as domestic source for investment. Thus, we include total domestic savings as a percentage of GDP (=St/GDPt) in the regressions with investment as dependent variable, and it is expected to have a significant positive influence on the growth rate, through investment. Then, there is a measure of human capital variable, given by the percentage of adult illiterate rate (=AIR). It is expected to have a negative influence on the growth rate and investment. 4. Pooled time-series cross-section preliminary analysis Data exploration and cross-correlations Our analysis refers to a set annual data, over the 1970-1999 period (due to the presence of missing data in the period 1960-69), of 30 Sub-Saharan African countries26. We excluded 18 out of the 48 SSA countries, due to missing data, geographical dimension (very small countries and islands, such as Comore and Sao Tomè) and anomalies (South Africa). Summary statistics (Appendix 1, Tabs. 1 and 2) show how limited is the capacity of the first moment (the expected values) to be a summary statistic of a probability distribution in an heterogeneous context, such as our phenomenon is. And other summary statistics (Appendix 1, Tab. 3) show that heterogeneity is particularly evident in the cross-country dimension, whereas time-series dimension indicates that, as in the majority of economic variables, a dynamic (non-linear, inverted-U shaped, in case of aid and debt service) trend seems to predominate, even though some interruptions due to exoghenous shocks occurred. Thus, we need more detailed investigation. 24 R. Barro (1997), pp. 32-3. 25 Evidence in support of this is provided in R. Lensink and O. Morrissey (2000), “Aid instability as a measure of uncertainty and the positive impact of aid on growth”, Credit Research paper, University of Nottingham. 26 Benin, Botswana, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Congo, Dem. Rep., Congo, Rep., Cote d'Ivoire, Gabon, Gambia, Ghana, Kenya, Lesotho, Madagascar, Malawi, Mali, Mauritania, Mauritius, Niger, Nigeria, Rwanda, Senegal, Sierra Leone, Sudan, Swaziland, Togo, Zambia, and Zimbabwe. The other 18 SSA countries have too many missing data to be included in the pooled analysis. 197 As a first step, divided the series into four sub-periods and considering the average values of each of these four periods, we explored the bivariate relationship between (i) growth and investment, (ii) growth and aid, (iii) growth and external debt, (iv) growth and debt service, (v) investment and aid, (vi) investment and external debt, (vii) investment and debt service, (viii) aid and external debt, (ix) aid and debt service, we plot the evolutions of the average values over time and cross-country. The figures show that growth appear positively correlated with investment and negatively correlated with foreign capital inflows during the 1980s and 1990s, after a positive correlation in the 1970s. The correlation matrix confirms this evidence (see Appendix 1, tab. 4). The positive correlation between growth and foreign debt in the 1970s, when the level of debt stock was much lower than the following decades can be used as a first confirmation of the fact that the impact of foreign capital inflows is level dependent: during the 1980s and 1990s, when external debt burden increased in a dramatic way the correlations become negative (level dependence as a way to interpret the time dependence). After having analysed bivariate correlations, which may be in part spurious, reflecting the effects of third factors, the next step is multivariate regression analysis. Thus, we tried some empirical analyses over the last 30 years that make use of cross-country multiple linear regressions in assessing the effectiveness of foreign capital inflows on investment, and of investment on growth. In fact, as we mentioned, from both theory and empirical literature27, a reasonably consistent pattern is that investment is the fundamental engine for growth and that foreign capital inflows are significant regressors on investment. Having annual data over the 1971-1999 period, we considered the average values of four different subperiods: Period I (from 1971 to 1978: the “illusion” stage of increasing indebtedness), Period II (from 1979 to 1985: the “crisis” stage of debt explosion), Period III (from 1986 to 1992: the “cure” stage of adjustment without reducing debt) and Period IV (from 1993 to 1999: the “disillusion” stage of lack of development without reducing debt). We use the mean values of the given periods for most of the variables. AIR EX_GDP FDI_GDP FLTD_GDP GGDPPC GPOP IM_GDP IN_GDP INF LTD_GDP ODA_GDP REM_GDP S_GDP SCR_GDP STE_GDP TDS_GDP Adult illiterate rate Export/GDP Foreign direct investment/GDP Foreign long term debt/GDP Annual GDP growth rate Annual population growth rate Import/GDP Investment (=Gross domestic capital formation)/GDP Annual inflation rate Long term debt stock/GDP ODA/GDP Remittances/GDP Domestic savings/GDP State current revenues/GDP State total expenditure/GDP Total debt service/GDP 27 Particularly, the 1990s new generation of aid effectiveness studies appeared focusing on the relations between aid, policies, and growth. 198 We considered the initial level of some other variables (1971, 1979, 1986, 1993): INGDPPC INPOP INTED_GDP Initial level of per capita GDP Initial level of population Initial value of total external debt A preliminary analysis of the correlation matrix of the variables shows some elements to be reminded. This analysis is useful also because very high (partial) correlation matrix coefficients (>.79) can help in detecting multicollinearity28, to be considered in time series and pooled analyses. In particular, as we expected, the highest link is between export and import. Concerning the GDP growth rate, theory of growth is partially confirmed. In fact, investment is correlated to growth, as expected since Harrod-Domar and Solow models and it is the more relevant findings for our purposes. The relationship between growth and human capital is not evident (the signs of the coefficients change from period to period), thus the endogenous growth theory, based on the importance of human capital, is not clearly confirmed in the SSA case: the negative correlation is confirmed in the first three periods, with a very low coefficient value, indicating the lack of economic significant relationship. Inflation is negatively related to growth in all the periods and the level of exports is positively related to growth, as expected by the Washington Consensus approach. Demographic pressure is negatively related to growth performance, at least in the last three periods. All the foreign capital inflows seem to be related to growth with different signs from period to period, however, the stock of foreign debt is negatively related to growth in the last three periods (when debt burden became unsustainable). In SSA countries’ case there is no conditional convergence of poor countries to higher growth. Concerning investment, there is a positive relation with domestic savings, as we expected, even though the reliability of this aggregate figure is quite low, because of the national account definitions. Investment is positively related to the openness of countries (expressed by imports and exports). But, given the persistence of primary sector in the export profile of African countries and the links between domestic investment and foreign direct investment, this results can be due to the effect of foreign investments concentrated in export-oriented raw materials. Investment are negatively related to inflation, confirming the importance of macroeconomic stable environment for attracting domestic investment. The positive link to human capital endowment seems to confirm the importance of investment in human capital to promote growth. A negative relation with demographic pressure emphasises the obstacle of demographic growth to development. No clear relation emerges between investment and aid or debt. Concerning external debt, there is no clear and permanent link to any other variables, as we expected, given the different impact of debt due to its high or low level, in absolute and relative terms. Moreover, we expect heterogeneity among different countries and, above all, a indebtedness spiral. We have to remind that the mean values (average values for different periods) force heterogeneity to cluster around summary statistics, which seem to be inadequate in our context. The moderate link between debt and aid derives from the presence of soft loans in the ODA aggregate and from the expected and complex relationship between these two flows. A moderate link which is highly dependent on the method of using average values for sub-periods. Concerning aid, it is negatively linked to per capita GDP and is positively linked to illiteracy, confirming the 28 The multicollinearity problem in regression analysis (indicated if the R-square is greater than .75 and only a few tvalues are significant) reflects the weakness of data. It simply means that we are unable to identify a statistical relationship that is insensitive to the conditioning set of information and we are not able to properly isolate the unique effect of each variable and the confidence with which we presume these effects to be true. 199 basic orientation towards the poor countries. The positive link to demographic pressure seems to confirm the same point. The negative relation to domestic savings could support the theory of aid devoted to close the savings-gap, but it is not confirmed by the link to investment, which is negative. It seems to demonstrate that aid went to African countries to help them to fight against famine, rather than promoting investment. The negative relation with openness can confirm this compensatory role of aid. There is no clear relation with other foreign capital inflows, due to the nature of data and inadequacy of summarising them through average values of sub-periods. Remittances data are not reliable, nevertheless, with caution we note the positive link between aid and remittances: both of them are attracted by poor contexts (low income, high illiteracy rates, low exports, low investment). Concerning foreign direct investment, the second and third periods shows the significant signs of the links, whereas the other two periods show no significant link. From 1978 to 1992, FDI seem to be linked to openness, initial GDP, low demographic pressure, higher human capital quality. FDI are positively linked to domestic investment and growth in Period II and III, but negatively in the other periods. This confirms the idea of lack of very strong relationship of the FDI flows with other structural variables of development. The assistance of sensitivity analysis During all the steps of our analysis, we explored whether the relationship between various measures of foreign capital inflows and economic development (both GDP growth and investment) are robust or fragile to small changes in the conditioning information set. Using Extreme bound analysis (EBA), we tried to show that most of the variables we included in our regressions are robust (i.e. their coefficients and significance don’t change substantially depending which other variables are included in the estimated equation). We used the version of Leamer's (1983) extreme bounds analysis as presented by Levine and Renelt29. The procedure of this version of EBA is as follows. We first select the basic variables to be always included in the regressions and the variables of our interest (M, in our case being the foreign capital inflows), and run a base regression that includes only these variables. Then we compute the regression results for all possible linear combinations of up to three other variables, which have a reasonable theoretical foundation, and identify the lowest and highest values for the coefficient of the M variables, β µ, that cannot be rejected at the 5 percent level of significance. For each regression j, we find an estimate β µj and a standard deviation σmj. The lower extreme bound is the lowest value of β mj - 2σmj, whereas the upper bound is β mj + 2σmj. If the upper extreme bound for variable M is positive and the lower extreme bound is negative (i.e. the sign of the coefficient β µj changes), then variable M is fragile. If the coefficient remains significant and it does not change sign, then the coefficient is robust and we can be confident in that partial correlation. The robustness of all different regressions have been examined and all the results we will present in the next paragraphs have passed the EBA robustness test, if no explicit mention is reported at this regard. Cross-country multiple linear regressions After this preliminary analysis, for each of the four periods, we run three different regressions. As these regressions don’t take in account the dynamics of the relationships (we are considering averaged values of some periods), they are basically instruments to confirm the presence of heterogeneity. The first regression tries to predict the value of the averaged annual growth rate of real GDP per capita (= 29 R. Levine and D. Renelt (1992), “A Sensitivity Analysis of Cross-Country Growth Regressions”, American Economic Review, Vol. 82. 200 GDPgt) from the values of: (1) the variable introduced by Barro and Sala-i-Martin (1992), that is the logarithm of the initial level of per capita GDP (=logInGDP), (2) total gross domestic investment as a percentage of GDP (=GDIt/GDPt), being referred – as the first regressor – to the Solow growth model, (3) a measure of human capital variable, derived from the endogenous growth theory and given by the percentage of adult illiterate ratio (AIR), (4) a measure of openness, linked to the Washington consensus approach and given by total exports and imports as a percentage of GDP (=EXt/GDPt+IMt/GDPt). Other variables, such as inflation (reflecting the Washington Consensus approach) and public budget (three-gaps approach) did not result relevant 30. The second regression adds simply the foreign long term debt inflows variable (=D(FLTD/GDP)t), aid and foreign direct investment to the first benchmark regression. The third regression tries to test the indirect effect of foreign capital inflows on growth, as total gross domestic investment as a percentage of GDP is the dependent variable, whereas the regressors are: (1) the average resource flow generated by total external borrowing, calculated as the annual change in the stock, (2) the average Official Development Assistance flow in current US$ as a percentage of GDP, (3) the foreign direct investment flows as a percentage of GDP, (4) total gross domestic savings as a percentage of GDP. This is the basic equation that has to be considered the benchmark also for pooled analysis, when we will compare it to other specifications, adding other foreign capital inflows variables, as (5) total debt service as a percentage of GDP, (6) the initial level of total external debt to GDP (= EDT1970/GDP1970), which is used as a measure of the debt stock, and including non-linear and additive components, as (7) the aid (ODAt/GDPt2) and debt (FLTDt/GDPt 2) squared terms, to allow for the possibility that there may be diminishing marginal returns to aid and debt, and (8) The interactive term between aid and debt (ODAt*FLTDt). First, it is extremely important to distinguish between statistical significance and economic significance. Investigation results depend on substantive not merely statistical significance: in our case, statistically insignificant results can mean important results, as they confirm the importance of heterogeneity among countries31. The first two equations32 show the see-saw of results, in statistical significance: the first and third periods confirm partially our expectations in terms of relationships among variables, whereas the other two periods have no statistical significance. The first periods confirm that the initial level of per capita GDP shows negative relationship with GDP growth (confirming conditional convergence), as well as in the case of initial population (demographic constraint), whereas the initial level of total external debt has a positive coefficient (debt as an engine for growth). External debt seems to have positive effect on GDP growth, whereas debt service and aid have negative effects; FDI is insignificant: however, these results are not statistically significant as the first moment (the mean) is not a reliable summary statistic (the standard error of the mean is always high). In the case of last period (1993-99), the multiple cross-countries linear regression with GGDPPC as 30 At this regard, we used the so called Extreme-bounds analysis (EBA), which will be presented in the next paragraph, as we extensively use it in the pooled time-series and cross-section analysis in order to test the robustness of coefficient estimates to small changes in the conditioning information set. 31 D. N. McCloskey and S. T. Ziliak (1996), “The standard error of regressions”, Journal of Economic Literature, Vol. XXXIV, March. 32 The second regression does not improve the general results: statistical significance of the regression is negatively affected by the added regressor, which shows an insignificant coefficient and denotes the lack of regularities in the foreign capital inflows and growth relationship. 201 dependent variable provides bad results, in terms of t-statistic: there is high probability of being wrong in concluding that there is a true association between the variables (i.e. the probability of falsely rejecting the null hypothesis, or committing a Type I error, based on t). In fact, ANOVA detected a significant difference of variability among the variables (F=14.9)33. Thus, the power of the regression to detect the observed relationship in the data is not strong in probability terms, and the findings must be interpreted cautiously. We calculated the variance inflation factor (VIF), as a measure of multicollinearity among the independent variables: the variables did not have any very severe multicollinearity problem. Data passed normality tests as well. It is quite clear that the presence of heterogeneity prevent results from demonstrating a clear relationship among the variables, and the heterogeneity implies also a dynamic dimension, as it differs consistently from period to period. The expected relationships between growth, investment and initial rate of growth occur before the Period II, that is before the second oil shock and financial crisis of African debtor countries, and less during Period III, after the introduction of structural adjustment programmes. During the big debt crisis (Period II) and disillusion on convergence (Period IV) the relationship disappeared. Other factors, and particularly the foreign debt spiral prevailed and made the traditional determinants of growth (physical and human capital) insignificant (see Appendix 1, tab. 5). Concerning the regression with investment as dependent variable, given the low statistical significance, the presence of heteroscedasticity (thus, in place of the standard OLS formula to compute the variances, we used the White heteroskedasticity consistent covariance matrix estimator which provides correct estimates of the coefficient covariances in the presence of heteroskedasticity of unknown form)34 and without regularity from period to period, there are some results to be stressed, which can be investigated through further steps of our analyses. • the forward stepwise regression confirmed that the independent variables that produces the best prediction of the dependent variable are FLTD/GDP and FDI/GDP, whereas other variables do not significantly improve our ability to predict the dependent variable. • The relationship with savings is positive, as we expected, and the partial coefficient is statistically significant in all the periods, included the Period IV, which presents high R2 and homoscedasticity. • There is no relation between investment and aid, confirming what the analysis of correlation matrix of all the variables showed and what was our hypothesis on SSA experience. The relationship between domestic investment and FDI is quite ambiguous: the sign of partial coefficient is positive but the probability of correctly rejecting the null hypothesis is high only in the Period III, when the regression is much better, in terms of ANOVA, F and t statistics, but also when heteroscedasticity is high as well. The probability is low only in the Period IV, when there is no heteroscedasticity (see Appendix 1, tab. 5). The African specificity compared to the prevailing literature on debt overhang effect was tested in the 1993-99 period (when FDI flows grew around the world, whereas the external debt problem was not solved): considering both FDI_GDP and IN_GDP as dependent variables and INTED_GDP and 33 We used all pairwise multiple comparison procedure to determine the differences between groups. Normality test of the estimated underlying population and equal variance test results showed that the data failed the test of the assumption that the underlying population or residuals were normally distributed and that the samples were drawn from populations with the same variance. These basic assumption of most of the parametric tests resulted to be violated. 34 H. White (1980), “A Heteroskedasticity-Consistent Covariance Matrix and a Direct Test for Heteroskedasticity”, Econometrica , Vol. 48. 202 LTD_GDP as regressors, no significant relationship emerged between these variables. There was no sign of particular preference of foreign investors for less indebted countries (and for favourable environment): in the SSA case the main motivations seem to derive from other considerations, namely the opportunity of runaways or of export-oriented enclaves. Cross-plot of external debt and aid We should emphasize at this point that the need of pooled analysis stems from the technical limitations of repeated cross-section analyses for sub-periods, in a heterogeneous phenomenon as our case is. As the pooled coefficients take on information from both cross sections and time series, we could assess the relative impact of these two dimensions and we should try to inspect the heterogeneous nature of the relationships over time and countries. The limitations of repeated cross-section analyses for sub-periods is graphically displayed in Figure 3, which provides us with a broad map of the correlation between the level of external debt and aid inflows in the total sample of observations (a point for every country in every year) under study. Differently from what the correlation matrix of variables for the four sub-periods (with average values) indicated, the data show a positive correlation between the level of external debt and aid, but the figure reveals that the linear relationship is very far from perfect. There is a substantial variation in the aid inflows, without any significant clustering around the mean and no significant regression with the scatter. In the more dense area of the Figure, that is when external debt to GDP ratio is lower than 100% and aid to GDP ratio is lower than 30%, the high level of dispersion of points is confirmed, too. Figure 3 – Cross-plot of external debt and aid ODA_GDP vs. LTD_GDP (1970-99, 30 countries) ODA_GDP vs. LTD_GDP (1970-99, 30 countries, LTD_GDP<1) 0.6 1.0 0.8 ODA_GDP ODA_GDP 0.4 0.6 0.4 0.2 0.2 0.0 0.0 0.0 0 1 2 3 0.2 0.4 0.6 0.8 1.0 LTD_GDP LTD_GDP The estimated joint levels of aid and external debt, at any level of investment rate (i.e. the cross-plot of external debt and aid at different iso-investment levels) gives the combination of debt and aid that correspond to the same investment rate. Within the plotted ranges of investment (see Appendix 1, tab. 6), even though investment depends on other country characteristics than aid and debt, there is an interesting basic result: the higher the investment rate, the narrower is the range of external debt ratio (in any case less than 100% of GDP, when investment rate is higher than 20% of GDP). Obviously, it is not possible any causal interpretation of the results, and it is important to be aware of the reality of aid and external debt in SSA. We think, and the Figure confirms it, that aid and external debt are not basically devoted to promote investment and growth: particularly in case of conflict-affected countries, aid goes to humanitarian assistance and to pay debt service, whereas external debt go to arms and to feed debt spiral. Thus the 203 assumption of any rigid dependence between foreign capital inflows and economic development risks to be misleading as well as any assessment on the impact of aid on growth and macroeconomic good environment remedy to attract FDI. Time-series model identification: equation in first differences As following step, we constructed our time series model, based on the variables’ underlying generating process. The most fundamental assumptions in time series analysis is that the time series must be stationary, and time series component of our panel, as most of time series data dealt with in economics and social sciences, are nonstationary35. When we visually examine the graph of the different series, which we are using, there are some key features: (1) most of these series (apart of INFL, SCR and STE) contain a clear trend (in particular, POP and OER) and don’t have a time-invariant mean; (2) they seem to go through sustained periods of upward and downward trend, with high persistence and no tendency to revert to a long-run mean, demonstrating a “random walk” behaviour, (3) some series (GDPpc, LTD/GDP, AIR) show a particular tendency to increase, whereas (4) other series (ODA/GDP, INV/GDP) show marked shocked declines followed by a resumption, around a given level, thus (5) the volatility of many series (IMP/GDP, INV/GDP, ODA/GDP, IDE/GNP) is not constant over time, showing a conditionally (or short-run) heteroskedastic behaviour around a mean. Some series share co-movements with other series. This confirms current consensus that most macroeconomic time series contain a stochastic trend. Thus, to begin with, following the Box and Jenkins procedure, we have to transform the time series employed into a stationary series in the wide sense in order to interpret correctly the autocorrelation function (ACF). The most common transformation is that of differencing, that is subtracting a past value of a variable from its current value. We must determine how many lags to be included in the specification, provided that a first order difference trasformation is sufficient to remove a linear trend. We use the Augmented DF test, which accounts for a more dynamic specification of the regression than the DF test36. In most of the series, the regression values of t are less than the critical value, thus the null hypothesis is accepted that the level series seem nonstationary and they possess a unit root37. We decided to estimate our equation in first differences, thus reducing the dataset to the 1971-1999 period. The first reason is to do with the non-stationarity of the time series component of our panel38. Given the large time dimension of the panels one cannot discard the time series properties of the individual donorrecipient pairs. Since not all the series for the donor-recipient pairs were found to be non-stationary taking the first differences means over-differencing the data. On the other hand combining the levels and the first differences of the data will introduce some methodological problems as well as considerable complexity in 35 Lloyd et al. (2000) established that most of the series for the donor-recipient pairs were non-stationary, i.e. integrated of order one. 36 This is a Unit Root test: residuals must be a White Noise process, which implies no significant autocorrelation at every lag. We test the residuals of OLS: if they are not White Noise, we re-do the analysis including some lags of Y in the right side as regressors. In our case, this test fails to reject test in level, but reject test in first differences. This means that our series contain one Unit root at I(1), as this test specifies the order of serial correlation to account for in the series. 37 A trasformation is based on the fact that wake stationarity requires that a process must have a constant variance along with a constant term. If nonstationarity enters through the variance of a process, then heteroskedasticity becomes a problem. In order to counteract this risk, we can perform a transformation on the sample time series. 38 Yt and Yt-1 are highly correlated in time-series, as series change only slowly over time, whereas differencing data (∆Yt and ∆Yt-1) are not highly correlated, as the change of growth is more erratic, without trend (∆Yt is stationary, whereas Yt has a long memory – i.e. a trend -, there is no fast drop, and autocorrelation is close to 1: it has a Unit Root) 204 interpretation (see Quah, 1994). Therefore, although there will be some loss of information as a result of differencing, for consistency and ease of interpretation we use the first differences of the entire time series component for all variables39. There is another reason to use first differences: it is a means of addressing the problem of lagged dependent variables in panel data analysis (the dynamic panel data problem we have mentioned). In fact levels are instrumented by first differences of the regressors, considering that in the presence of fixed effects, the results are biased by the presence of lagged Y among the regressors, as the coefficient on the lagged dependent variable is negatively correlated with the residuals40. A preliminary step of time-series analysis is to compare time series of the 30 countries analyzed. This is an indicator that more detailed analysis of the data is useful, as well as also repeated period averaged crosssection analysis demonstrated. In our case, this step confirmed that the adequate way of representing dynamic relationships is to analyze the whole relationship in differences, but in terms of adequacy of model specification for all the countries the results stressed the need of caution: within country dynamic variations of the relationships between variables of interest were frequent, showing the presence of heterogeneity in terms of relationships between countries and within countries. Thus, considering that we have to use weighted least squares to account for country-wise heteroscedasticity, and we attempt to address the potential endogeneity of foreign capital inflows by lagging most of the explanatory variables one year, this further step should be implemented to demonstrate that the SUR model improves the quality of results compared to fixed effects and basic pooled time-series models. 5. Pooled time-series cross-section analysis, estimation and results In order to combine both the dynamic and the cross-country comparison dimensions, which have been separately analyzed in the preceding section, we used the pooled analysis. Under evident conditions of heterogeneity, in all the pooled analyses we ran the SUR estimation, which is much sensitive to individual contributions, proved better than simple OLS regression, Fixed (including country and year dummy variables) and Random Effects models. Even though no significant difference appeared in the results of regressions which used annual data compared to regressions with centered moving averages (each year being replaced by the average of itself and surrounding observations, the span – i.e. the number of observations – being three), we preferred to present the pooled analyses with three-year span moving averages, to net out the effects of short run fluctuations. We also have used seven-year averages to check the robustness of our findings, which was confirmed. Pooled GDP growth regressions The first specification of the models included growth rate of per capita GDP as dependent variable and run a a complete level regression. Table 5 presents the various basic specifications (corrected OLS and GLS), assuming a static model in levels and test whether the structure of the error term is adequately captured. 39 The use of first differences of all variables means to work in terms of rate of change of variables. In such a way the problems are to ignore information on the levels of variables and to focus on short term relationship between Xs and Y. If we look at the standard deviations, we can check the variation across countries (level values) compared to across time (i.e. the variation in the differences). If we find much more variation in the differences, this is an example of divergent information in the cross-section dimension compared to the time-series dimension. However, as we expect a non-linear relation between foreign capital inflows and development, it is not correct to infer too much from the pair-wise correlations. 40 The system-GMM method yields unbiased estimates and simultaneously address the endogeneity issue for some Xs. 205 The starting point of our analysis was that we tested the series and found that the process is stationary. Then, confirming the fact mentioned by Greene that panel data typically exhibit serial correlation, crosssectional correlation, and groupwise heteroskedasticity (Greene, 2000), we found such a structure in the residuals. The result of modified Wald test for groupwise heteroskedasticity in cross-sectional time-series FGLS regression model (H0: sigma(i)^2 = sigma^2 for all i, chi2 (28) = 161.01,Prob>chi2 =0.0000) showed the presence of heteroscedasticity. And the result of Breusch-Pagan LM test of independence (chi2(378) = 632.155, Pr = 0.0000) confirmed that there is cross-sectional contemporaneous correlation. However, the model showed non multicollinearity problem (VIF always smaller than 2.7). In order to detect serial correlation problem we ran Durbin’s M test on the basic specification and we found the presence of low serial correlation41. However, we decided to include a lagged dependent variable – which is based on theoretical foundations – and the serial correlation resulteds completely removed. The GLS corrected for heteroscedasticity and contemporaneous correlation, with lagged dependent variable to correct for serial correlation confirmed to be a well specified model (even though it may be overconfident,as the PCSE S.E. values show). Tab. 5 –pooled analysis: specifications of growth as dependent Variable (GGDPPC?) PCSE-I lggdppc fdi_gdp tds_gdp fltd_gdp2 oda_gdp2 oda_fltd ex_gdp in_gdp air _cons L1* (P>t) rho -10.48871 -36.53754 -33.69861 -58.23711 66.63817 2.543563 22.22635 .0653652 -2.403198 PCSE-II P>z 0,72 -10.337 0,076 -36.063 0,737 0,609 -31.925 0 -41.204 0,619 0,415 56.295 0,655 1.779 0,482 0,054 19.855 0,212 .05955 0,076 0,533 -1.8188 0,645 0,075 0,009 0,753 0,243 GLS -13,80218 -62,457 -3,000615 -60,907 75,56013 6,77033 25,7593 0,0681 -3,548513 P>z GLS-AR 0,403 0 0,948 0 0,169 0,051 0 0,014 0,144 0,001 -14,46546 -58,543 -12,10513 -72,719 81,26202 6,059671 25,0965 0,07382 -3,40922 P>z 0,399 0 0,798 0 0,148 0,126 0,001 0,022 0,225 0 GLS corr. 0,12761 -11,08647 -54,64 -2,361696 -58,864 79,21858 6,29674 20,9024 0,05738 -2,867855 P>z 0,001 0,494 0 0,959 0,001 0,149 0,067 0,002 0,038 0,237 0,543 0,15253 GLS corr. = GLS corrected for heteroscedasticity with lagged Y PCSE-I = Beck-Katz OLS regression, correlated panels corrected standard errors (PCSEs), with autocorrelation correction via PraisWinstein (common rho) PCSE-II = Beck-Katz OLS regression, correlated panels corrected standard errors (PCSEs), with autocorrelation correction via PraisWinstein (unit-specific rho). * M Durbin test (P>t) Rho = autocorrelation coefficient Then, we included both country and time effects and the residuals did indeed reveal the improvement of panel structure analysis (Table 6). But dummies resulted multicollinear with Xs, implying that the coefficients of Xs were not good, as there is a sort of trade off between technical-statistical and substantive specifications. We ran F-Test for the inclusion of country dummies (country effects), of year dummies (time effects); and for the inclusion of both country and year dummies (country & time effects). It resulted useful to add dummies as they are significantly different from zero, but if we included both time and unit dummies, as they probably interact, they become not significant. Thus, we accepted the less efficient results (serial correlated) as serial correlation is low, and rather than GLS with lagged dependent variable, we used GLS heteroscedastic panels with AR and unit and period FE (X coefficients resulted less disturbed by FE than by lagged variable and the full specification of GLS-AR with both dummies removed the low level of serial correlation, as M Durbin test shows). Moreover, using FE, net improvement occurs in R squared (= 41 The lagged residual coefficient is very low (.14) to suspect the presence of Unit Root. Thus, even though the serial correlation persists it does capture a very limited part and we don’t need to be worried about it. 206 regression coefficient), which shifts from 0.08 to 0.11 (unit dummies) or 0.42 (period dummies) or 0.45 (both). Tab. 6 –pooled analysis: Fixed Effects specifications of growth as dependent Variable (GGDPPC?) FE (within) lggdppc fdi_gdp tds_gdp fltd_gdp2 oda_gdp2 oda_gdp oda_fltd ex_gdp in_gdp air L1* (P>t) P>t 0,059541 0,131 7,531318 0,722 -37,95206 0,02 -26,04567 0,591 -42,37224 0,01 62,84712 0,324 2,096748 0,821 10,63131 0,288 0 0,34403 twoway FE P>t -0,085136 0,031 -9,886468 0,58 -38,87485 0,006 -97,89157 0,018 -35,90956 0,01 2,511523 0,711899 16,36762 0,112291 0,963 0,927 0,054 0,572 FE GLS-AR P>t -1,151699 -58,87261 -16,63864 -2,176088 -59,45625 72,99343 1,827937 16,34562 0,353753 FE2 AR 0,948 0,002 0,734 0,913 0,061 0,206 0,829 0,091 0 GLS- P>t -21,49748 -36,76116 -87,06568 5,549322 -45,34141 19,72739 7,357569 21,15774 0,012851 FE3 AR 0,132 0,001 0,021 0,617 0,04 0,695 0,009 0 0,614 GLS- P>t -11,67419 -58,79213 -97,5296 4,371541 -44,58846 27,36396 6,92655 22,47253 0,064733 0,443 0 0,015 0,787 0,076 0,6 0,299 0,002 0,693 0,153 FE (within) = with unit effects. sigma_u=7,2649; sigma_e=14,5907; Rho (fraction of variance due to u_i)= .1986; F test that all u_i=0 has Prob > F = 0.2252 Twoway FE = Including both unit and period effects FE GLS-AR = with unit dummies; FE2 GLS-AR = with period dummies; FE2 GLS-AR= with both unit and period effects * M Durbin test (P>t) Further improvement is due to the inclusion of dynamic relationship, considering the lagged Xs (apart from tds_gdp, oda_fltd, in_gdp). Tab. 7 –pooled analysis: Fixed Effects specifications of growth as dependent Variable (GGDPPC?) Dynamic regression P>t lggdppc l.fdi_gdp tds_gdp l.fltd_gdp -6,91009 0,739 -70,1412 0 47,67193 0 l.oda_gdp -18,0502 0,148 l.fltd_gdp2 l.oda_gdp2 oda_fltd l.ex_gdp in_gdp 4,64073 0,918 80,935 0 -98,1657 0,02 13,4769 0 15,72424 0,02 4 0,054298 0,08 3 l.air Dynamic twoway FE P>t -16,788 0,357 -52,1302 0 18,09404 0,09 1 52,00182 0,00 2 -21,5772 0,586 0,633768 0,978 -152,836 0 39,11913 0 11,12123 0,173 0,022923 0,909 Then, we improved our specification by assuming that intercepts and slopes were heterogeneous. In fact, we ran unrestricted (UR) model, that is unrelated regressions (OLS with cross-section specific slopes). We estimated cross sectionand period-specific intercepts and slope coefficients and we compared UR against FE model through Nested F-test and we found that UR could not be rejected. Finally, we ran a SUR model and we used use the Breusch-Pagan Lagrange Multiplier statistic as a test for 207 cross-sectional residual correlation, and it supported the idea to use SUR rather than unrestricted OLS 42. However, both of these models stress the importance of heterogeneity, which is clearly demonstrated by the fact that when we ran the same model specification by codcount 43, then we got very different values of adjusted R squared (from 0. 0.7288 to 0.0074). The signs of coefficients, which are statistically significant, are what we expected (also showed by Appendix, tables 14-22): negative effects of debt service, aid*debt interaction, marginal aid and marginal debt; positive effects of export and investment; ambiguous effects of debt and aid. 8.2 Pooled first difference GDP regressions The second specification of the models included the first differences of per capita GDP as dependent variable. We used this variable, rather than using per capita GDP growth, in order to analyze a variable that is more homogeneous to the independent variables, which are level variables (even though in terms of GDP). GDP growth rate is more meaningful from an economic point of view but the results do not change and, as we did not standardize data before performing the regression44 and we are using differences, the results are complex in interpretation in any case. This specification can include both current values of independent variables or lagged values, without significant differences in terms of results: the growth specification that includes lagged values of regressors seem theoretically better founded to take in account the dynamic pattern in the impact of GDP- determinants, which make the contemporaneous effect less relevant (see Appendix 1, tab. 7, compared to the following Tab.8 in this paragraph) 45. As part of the relationships may be spurious, reflecting the effects of third factors, we included a set of conditioning variables, such as the usual determinants of growth (investment, human capital, openness variables) and, in certain specifications, time and country effects. The pooled specifications were estimated using simple OLS, instrumental variables to correct for endogeneity (through a two-stage least squares, using as instruments the lagged values of the endogenous regressors), F.E. model to allow countries to have different intercepts and with and without time dummies to ensure that the results were not driven by time specific effects, and GLS-II (or GLS-SUR). The main differences between specifications appeared in terms of robustness and statistical significance: as we assumed, OLS resulted inadequate, whereas FE model showed the existence of some specific years and countries importance. The GLS-II estimation method resulted the most robust and provided the most significant coefficients. Whenever we added fixed effects to the given specification, we dramatically reduced the residuals variance, due to the importance of heterogeneity distortions in our dataset. In order to investigate the impact of external capital inflows on growth we augmented the specification by adding different foreign capital variables (debt, aid, FDI). To the extent that we controlled for investment (using an alternate specification that excluded investment), the results were very partially biased toward a smaller effect on foreign capital inflows on growth. It demonstrates that what is suggested by theory, that is part of the effect of foreign capital inflows on growth occur through the investment channel rather than directly, is not so true in the SSA case. Tab. 8 –pooled analysis: dynamic specifications of growth as dependent variable: D(GDPPC?) 42 chi2(190) = 282.691, Pr = 0.0000. 43 by codcount: reg ggdppc fdi_gdp tds_gdp fltd_gdp oda_gdp fltd_gdp2 oda_gdp2 oda_fltd ex_gdp in_gdp air. 44 As we are considering a linear relationship, the mean values of independent variables are very low (differences of percentage values) compared to the dependent variable: the difference in the magnitude of the values xi variation compared to the corresponding value change for y implies the regression coefficients may have very high values. 45 To save space neither the country fixed effects not the period effects are reported in our tables. 208 Y = clse D(GDPPC?) OLS FE RE-GLS GLS GLS II D(GDPPC?(-1)) 0.785 0.689 0.947 0.708 0.678 D(IN_GDP? (-1)) 1.9 88.1 130.4 123.9 91.4 D(AIR? (-1)) 0.489 -4.2 1.2 0.732 -4.4 D(EX_GDP? (-1))+ D(IM_GDP? (-1)) 108.3 97.8 36.8 46.5 98.8 D(FLTD_GDP? (-1)) 208.9 172.6 312.8 89.6 145.5 D(ODA_GDP? (-1)) 34.4 33.4 59.8 34.0 43.5 D(FDI_GDP? (-1)) 399.4 409.4 305.5 203.0 367.4 D(TDS_GDP? (-1)) -47.2 -76.3 -66.4 -17.7 -76.6 Adjusted R-squared 0.619 0.623 0.544 0.529 0.621 Durbin-Watson stat 1.468 1.460 1.456 1.503 1.441 Note: RE-GLS is based on Variance Components, GLS on Cross Section Weights. Underscored and italics values are non significant, bold values are significant at 1%, bold and italics at 5%, bold and underscored at 10%. The EBA method confirmed the robustness of the statistical significant coefficients of (IN_GDP) and D(EX_GDP)+D(IM_GDP), the other coefficients – including D(TDS_GDP) and D(FLTD_GDP) – resulted less robust, whereas D(AIR), D(ODA_GDP) and D(FDI_GDP) resulted fragile. We included the initial values of population, GDP and external debt: only INGDPPC was not statistically significant. In all the specifications, the regression showed the expected signs concerning investment (positive) and external debt service (negative), whereas aid coefficient confirmed the difficulty to define a comparable positive relationship between GDP growth and ODA in the SSA case. The external debt variable showed a statistically significant coefficient, which may imply the presence of structural relationship between growth and external loans. Concerning the problem of underestimating timing in the regressions using differences, for macroeconomic data the Autoregressive distributed lags (ARDL) model, in which lags of the dependent variable as well as the explanatory variables are included in the regression, have proven to be a useful step to be followed. We have therefore experimented with lags of the growth rate and the explanatory variables in the regressions. No particular difference emerged from dynamic specifications of the models (useful to have an endogeneity/simultaneity correction), compared to the models with current values of independent variables, which seems better theoretically founded. An important difference is due to the inclusion of the lagged dependent variable 46, which dramatically improves the results, in terms of R-squared and D-W statistic. Thus, the best estimation method was GLS-SUR, by using an estimation technique for dynamic pooled models with country specific effects. Pooled Investment regressions A third group of pooled regressions investigated the investment (as a share of GDP) dependency on some regressors, following the GDP growth relations. The basic idea was to analyze in detail the dynamic effects of foreign capital inflows on investment. In this case we investigate the complex relationship, including interaction effect and squared values of aid and debt. We adopted the same set of estimators used in the preceding specification. Tab. 9 – pooled analysis: complex effects on investment. Dependent Variable: D(IN_GDP?) Y = D(IN_GDP?) D(S_GDP?) D(FLTD_GDP?) OLS 0.235 0.170 FE 0.242 0.166 46 RE-GLS 0.220 0.182 GLS 0.259 0.109 SUR 0.244 0.165 We did not include the lagged dependent variable among the regressors in the presence of fixed effects, which (negatively) bias the results introducing a correlation between the lagged dependent variable and the residuals. 209 D(ODA_GDP?) 0.127 0.133 0.118 0.108 D(FDI_GDP?) 0.043 0.024 0.192 0.124 D(AIR?) -0.001 -0.003 0.000 0.000 D(FLTD_GDP?)^2 -0.572 -0.547 -0.635 -0.024 D(ODA_GDP?)^2 0.165 0.216 0.086 0.101 (LTD_GDP?)*(ODA_GDP?) 0.000 -0.014 0.008 -0.004 D(IN_GDP?(-1)) 0.447 0.435 0.529 0.458 Adjusted R-squared 0.387 0.371 0.332 0.370 Durbin-Watson stat 1.817 1.818 1.818 1.829 Note: RE-GLSis based on Variance Components, GLS on Cross Section Weights Underscored and italics values are non significant, bold values are significant at 1%. 0.128 0.058 0.001 -0.531 0.153 -0.003 0.445 0.386 1.812 Once again, statistical significance is very low: as a consequence of the heterogeneity of countries, the SUR specification results the most adequate. From the economic point of view, our analysis confirms that in SSA countries changes domestic savings is the main engine of investment, follower by external debt. External debt has a positive impact on growth, but with diminishing returns (the squared term). A simple explanation of the decreasing marginal impact is that the debt effect on investment depends on the level of debt. This is probably the result of vicious circles of indebtedness, which tied external loans to the repayment of preceding loans, without any direct link with economic development. In terms of debt relief this means that one should not expect the same effect on growth in countries with an EDT/GDP ratio of 100 percent as in countries with a ratio of 50 percent. The effect will, on average, be smaller in the highly indebted countries. Unfortunately, the presence of heterogeneity and the resulting impact of external debt on growth differs from a unique threshold level (assumed by the HIPC Initiative) or a standard inverse U-shape found in many studies of external debt problems. In any case, external debt contains usefulinformation about the growth process and this is very important. Less important, but significant – both in statistical and economic terms – and positively related to investment is ODA flow. We remind that a high percentage of humanitarian aid to Africa implies a limited impact on economic investment, even though it may be very positive in terms of emergency support and poverty reduction. Finally, FDI seems to have limited influence of investment. This supports the idea of an ambiguous nature of FDI in Africa and it is important to criticize the prevailing rhetoric on the positive impact of FDI everywhere and in any case. The openness is not clearly related to investment. The complex interplay between the level of external debt and aid flows seem to be confirmed in terms of statistical significance, but the effect is very small. This is due to the fact that the consequence of this interplay between aid and debt can not be expected to lead to the same responses (positive or negative) in terms of the impact on investment from one country to another and from one period to another as it depends on the levels. We used an alternate specification that included import, in order to analyze the effects of aid on investment, controlling for import. If we include Import to the regression, the aid inflows coefficients become much smaller, this suggests that the impact of aid on investment is also through import, that is the most effective interventions of development co-operation, in terms of impact on investment, is in Africa through the Balance of Payment support, rather than through projects. It is due to the fact that aid project are basically to support consumption and to alleviate poverty rather than promoting investment. AIR is not important (the coefficient around zero) and it is due to the fact that the education proxy of human capital shows a positive trend, which is common everywhere (differently from the health component of human capital, due to the epidemic contagious AIDS in Southern Africa) and is not correlated with other variables (without 210 similar linear trend). 6. Conclusions Lack of homogeneity Conditional convergence means that all countries, with a common technology, have equal per capita growth rates in equilibrium. This implies that, in the long run, economic growth is independent of government policies and social institutions and it is one of the basic messages of the neo-classical growth theory. The neoclassical argument is that a simple variant of the Solow (1956) model can adequately explain long run, cross-country economy performance. In fact, one of the outcomes of Solow’s neoclassical growth model is that, given the same parameters, poorer countries, those with lower initial capital stock and output levels, will grow faster than richer countries. That is, countries with the same technology, savings rate, and population growth rate, etc. (i.e. having the same steady-state) should show evidence of the poorer countries catching up to the richer countries. This issue has been extensively investigated by Paul Evans47 and the empirical evidence of the existence of this convergence has been mixed. As Mankiw, Romer, & Weil (1992) put it, "given the inevitable imperfections in this sort of cross-country data, we consider the fit of this simple model to be remarkable. It appears that the augmented Solow model provides an almost complete explanation of why some countries are rich and other countries are poor". Many studies, most famously Barro & Sala-I-Martin (1992)48, Mankiw, Romer & Weil (1992)49, and Islam (1995)50, find evidence of conditional convergence when technology, savings rate and population growth rate are controlled for. Recently, Kevin B. Grier51 showed that studies finding conditional convergence with a common technology suffer from significant unmodelled country heterogeneity. Specifically, key results are taken from regressions using data that are not described by a single data generating process. The samples used are inappropriately pooled: when observations that cannot be combined in a single regression are treated separately through disaggregated regressions, their claimed empirical support for convergence vanishes, and estimated production function coefficients are not stable across sub-samples of countries in a large panel. We confirm this results, because a small group of apparently homogeneous countries from an ex-ante perspective, namely 30 Sub-Saharan African countries, show significant growth rate divergence. The African non-convergence finding is a strong counter example. Large samples cannot legitimately be pooled together without taking account of the heterogeneity and the disaggregated results do not support the augmented Solow model, whereas provide strong support for more specific endogenous growth models. Pooling countries with different underlying time series properties (different inherent causality) is inappropriate and gives misleading results. Moreover, the direction of causality between the variables of primary interest may differ among countries in the sample: it is therefore an example of what can be termed 47 Evans, P. (1996), “Using cross-sountry variances to evaluate growth theories”, Journal of Economic Dynamics and Control, vol. 20, p. 1027-1049; Evans, P. (1997), “How fast do economies convergence?”, Review of Economics and Statistics, vol. 79, p.219-225; Evans, P. (1998), “Using panel data to evaluate growth theories”, International Economic Review, vol. 39, p. 295-306. 48 Barro, R. and X. Sala-I-Martin (1992), “Convergence”, Journal of Political Economy, vol 100, p. 223-251. 49 Mankiw, N. G., Romer, D. and D. Weil (1992), “A contribution to the empirics of economic growth”, Quarterly Journal of Economics, vol. 107, p. 407-437. 50 Islam, N. (1995), “Growth empirics: a panel data approach”, Quarterly Journal of Economics, vol. 110, p. 1127-70. 51 K. B. Grier (1999), “Convergence: what does it mean, how can we measure it , and where is it found?”, Department of Economics, University of Oklahoma, mimeo, April. 211 time-series heterogeneity (the underlying causal relationship is not the same for all countries). The problem to be addressed is if the efficiency gains of pooling the data outweigh the losses from the bias induced by heterogeneity. In other words large samples are not necessarily the best samples: to identify aggregate relationships, or correlations in the data that appear to hold ‘on average’ over a wide sample is not the best procedure to be followed. What appears to hold on average is rarely an adequate explanation of what is happening in a particular country. Despite numerous studies and enormous research effort, the results are disappointing: as Kenny and Williams wrote “the universal failure to produce robust, causally secure relations predicted by models might suggest … that country growth experiences have been extremely heterogeneous” (, 2001: 12). The classical way of assessing homogeneity is an F-test, which compares the sum of squared residuals (SSE) of pooled equation to the SSE of the unpooled regressions, which are essentially the sums of squared errors of the N separate unit regressions 52. Both equations are estimated by OLS. Some countries are unlike others and some variables fail homogeneity and, obviously, there is not enough extra explanatory power in the unpooled model if we keep the same specification. But, if we repeat the test, allowing only some “variables of interest” to have separate coefficients by countries, the results seem to improve.Another test of homogeneity of all the countries is cross validation: we left out one country at a time and then predicted the dependent variable for that country. Countries appeared different, as the model did not perform equally for all countries. We also tested for whether we need fixed effects by an F-test comparing the SSE of the two pooled regressions (OLS and FE). FE technique improves the results, but fixed effects are not good explanatory variables, as they simply confirm the existence of heterogeneity. Moreover, fixed effects are not only perfectly collinear with independent variables that are time invariant, but also highly correlated with independent structural variables, which change little (such as AIR). Thus, fixed effects make it appear that AIR has no impact, as the effect of this variable is controlled for the fixed effects, which change the coefficient of AIR enormously53. In our case, the best estimation method was SUR, by using an estimation technique for dynamic pooled models with country specific effects. Using this approach has several advantages over the cross-section analysis widely used. In particular, this approach controls for country-specific, time-invariant (over the sample period) country characteristics. Thus, rather than looking for correlations between external debt and investment across countries (which could be due to any number of unobservable country characteristics rather than any causal relationship between the two variables of interest), we are instead asking whether a change in external debt within country is systematically associated with changes in investment. Spurious correlation could of course still occur, but the use of multiple time periods for each country and fixed effects reduces the she risk for these kinds of errors. In addition, this approach can give us some freedom to explore certain dynamic questions such as whether external debt in one period is linked to changes in investment in later periods, a useful piece of information for assessing the likelihood of causal relationship. Moreover, the presence of heterogeneity across periods and countries can be so evident to reduce the possibility to identify common trends in terms of aid (debt and FDI) and investment or growth relationship all over the large group of SSA countries. The identification of the heterogeneity and the complexity of these relationship is itself the main result, which induces us to proceed to country-case studies. 52 There are other ways to choose between models: the Schwarz criterion, used in time-series analysis, judges models by their sum of squared residuals plus a penalty for lack of parsimony. 53 It is clear that a time invariant variable tells us little about how the dependent variable in a country changes over time. 212 The importance of more homogenous sub-samples To investigate whether country effects are sufficient to capture the heterogeneity in these data, we split the sample geographically into 3 and 4 groups (within Sub-Saharan Africa)54 and estimated a separate regression for each sample. In this way, we relaxed the implicit constraint that the coefficients are the same in all the countries. We could test whether using one set of coefficients fits the data as well as using 3 or 4 sets. We confirmed there is not a unique set of production function coefficients that can explain the African economic performance, even allowing for country fixed effects. Further, we cannot place any confidence in these estimates either. Undoubtedly, each of our sub-regression would still fail tests for coefficient homogeneity. The point is simply that, because of pervasive heterogeneity, we are not going to learn much about the existence of clear relationships from regression analysis using a large number of countries. The assumption of homogeneity of data and relationships between variables, across both countries and years, is irrealistic, the promising approach is to attempt to model heterogeneity. Further investigation could be to identify (through the cluster analysis or multi-way analysis) some countries with different dynamic evolutions55 and then to analyse and compare in details their history. Otherwise, from a lender/donor perspective, we can deeply investigate the inter-relationship between debt relief and aid policies (see: Appendix 2). We also tried to analyze how those observations (countries in a given year) which have had investment values at least equal to 15% of GDP (see: Appendix 1, tab. 6), compare to other observations with different characteristics56. Or, we adopted a more conventional approach, creating sub-groups based on the openness degree, or on other characteristics derived from the available dataset, but the results were not so self-explanatory. In sum, the relationship between the variables is complex and dependent on the heterogeneity. Is heterogeneity the problem or the solution? Heterogeneity of African countries is the main result of our analysis. This is what we expected. Not only does it confirm the fact that, across countries and over time, there is no unique relationship between foreign capital inflows and economic growth, but it also lead to question the appropriateness of the units of investigation. Concerning the absence of a common relationship, it basically reflects the complex nature of the relationship: the impact of foreign capital inflows on economic growth is level-dependent, and it also depends on the interaction with other flows as well as on the specific national environment. We can assume that all the African countries experienced the same problems and faced the same challenges, but it can not 54 We used both Principal Component Analysis and Hierarchical Cluster Analysis. Principal Component Analysis attempts to identify underlying variables, or factors, that explain the pattern of correlations within the set of observed variables. Factor analysis was used in data reduction to identify a small number of factors that explain most of the variance observed in a much larger number of manifest variables. We also used factor analysis at the beginning of our research to generate hypotheses regarding causal mechanisms or to screen variables for subsequent analysis (for example, to identify collinearity prior to performing a linear regression analysis). Principal Component Analysis is a procedure that attempts to identify relatively homogeneous groups of countries based on selected characteristics among the variables we considered, using an algorithm that starts with each country in a separate cluster and combines clusters until only one is left. Distance or similarity measures are generated by the Proximities procedure. Then we used dendrograms to plot the results, to assess the cohesiveness of the clusters formed and to have information about the appropriate number of clusters to keep. 55 For example, a country being in the “good performers” sub-group in the 1980s and in the “bad performers” sub-group in the 1990s versus another country that remained in the same “good-performers” sub-group. 56 We defined an investment weighted dummy which equals to 1 if IN_GDP ≥ 0.15, and 0 otherwise. 213 mask the fact that factors endowment, historical heritage, patterns of specialization do matter. And all these factors clearly interact each other, in a non-linear relationship. If we are interested in investigating the nature of the debt Laffer curve, the sustainability thresholds of debt burden, the marginal impact of aid on development more deeply, we are obliged to adopt a country-case approach. There is no common rule of thumb, such as the HIPC criteria imply. Concerning the appropriateness of the units of investigation, we would like to simply mention that countries are not necessarily the best units to analyze economic growth, particularly in the new context of globalization. The importance of both the local development and territorial context from one side, and delocalization of economies emphasize other dimensions and borders than national ones. Particularly in Africa, we can use concepts as enclaves, export-oriented and cash-crops sectors. We have also to remind the importance of informal sector and dual economies. Local strategies for surviving, which are not registered in national accounts, involve the majority of people, particularly in rural areas. We know that it is impossible to rely on the fact that African people live with less than one dollar a day: these facts are simply misleading. The importance of informal sector, the high percentage of rural people and non monetary markets underline that countries and GDP are not the most appropriate concepts to investigate the lack of homogeneity. Heterogeneity within countries is at least as important as between countries. References Arellano, M. (1989). "A Note on the Anderson-Hsiao Estimator for Panel Data," Economic Letters 31, 337-341. Arellano, M. and Bover, O. (1995). Another look at the instrumental variables estimation of errorcomponents models. Journal of Econometrics, 68, 29-51. Baltagi, B.H. (1995), Econometric Analysis of Panel Data, John Wiley and Sons, New York. Barro, R. J. (1991). Economic growth in a cross-section of countries, Quarterly Journal of Economics 106, 2 (May), 407-33. Barro, R. J. (1997). Determinants of Economic Growth: A Cross-Country Empirical Study, Cambridge, Mass.: MIT Press. Barro, R. J. and X. X. Sala-I-Martin (1995). Economic Growth. McGraw-Hill, Inc. Beynon, J. (2001). Policy Implications for Aid Allocations of Recent Research on Aid Effectiveness and Selectivity, Paper presented at the Joint Development Centre/DAC Experts Seminar on “Aid Effectiveness, Selectivity and Poor Performers”, OECD, Paris, 17 January. Bleaney, M. and A. Nishiyama (2000). Explaining Growth: A Contest between Models. CREDIT Research Paper, No. 00/11, University of Nottingham. Boone, P. (1996). Politics and the effectiveness of foreign aid. European Economic Review, 40, 289329. Bosworth, B. and Collins, S. M.(1999), Capital Flows to Developing Economies: Implications for Saving and Investment, The Brookings Institution, mimeo. Burnside, C. and Dollar, D. (2000). Aid, policies, and growth. American Economic Review, 90, 847868. Buse, A (1973), “Goodness of fit in Generalized Least Squares Estimation”, American Statistician, Vol. 27, pp 106-108. Collier, P. and Dollar, D. (2001). Development Effectiveness: What have we learnt? Mimeo, World Bank, Development Research Group, Washington, DC. 214 Dalgaard, C.-J. and Hansen, H. (2001). On aid, growth and good policies. Journal of Development Studies, N.. Easterly, W. (2000). How did highly indebted poor countries become highly indebted? - Reviewing two decades of debt relief. World Bank, Development Research Group, Washington, DC. Feyzioglu, T., Swaroop, V, and Zhu, M. (1998). A panel data analysis of the fungibility of foreign aid. World Bank Economic Review, 12, 29-58. Granger, C.W.J. (1988) "Some Recent Developments in a Concept of Causality", Journal of Econometrics, 39, 199-211. Granger, C.W. J. (1987). “Causal Inference,” The New Palgrave: Econometrics, W.W. Norton, New York. Granger, C.W.J. (1986) "Developments in the Study of Co-Integrated Economic Variables", Oxford Bulletin of Economics and Statistics, 48, 213-228. Greene, W.H. (1993), Econometric Analysis, Macmillan, 3rd edition, Chapter 14. Guillaumont, P. and L. Chauvet (1999), `Aid and Performance: an Assessment’, paper presented at the 11th ABCDE Conference, World Bank. Gunning J. W. (2000), Rethinking Aid, paper presented at the 12th Annual Bank Conference on Development Economics (ABCDE), World Bank, Washington, D.C. Hamilton, J. D. (1994). Time-Series Analysis, Princeton University Press, Princeton. Hansen, B. (1999), “Threshold Effects in Non-Dynamic panels: Estimating, Testing and Inference”, Journal of Econometrics, Vol. 93, N. 2, pp. 345-68. Hansen, H and Tarp, F. (2001). Aid and growth regressions. Journal of Development Economics, 64, 547-570. Hansen, H and Tarp, F. (2000). Aid effectiveness disputed. Journal of International Development, 12:3, 375-398. Hansen, H., and F. Tarp (1999). "The Effectiveness of Foreign Aid". Mimeo, Development Economics Research Group, University of Copenhagen. Hansen, H (2001). The impact of aid and external debt on growth and investment: insights from crosscountry regression analysis, Hausman, J. A. (1978). “Specification Tests in Econometrics.” Econometrica 46:1251–72. Hausman, J. A. and Taylor, W. E. (1981), “Panel Data and Unobservable Individual Effects,” Econometrica, November, 1377-1398. Hsiao, C. (1986), Analysis of panel Data, Cambridge University Press, Cambridge. Johnston, J. and Di Nardo (1997), Econometric Methods, 4th edition, chapter 12. Knight, M., Loayza, N. and Villanueva, D. (1992), “Testing the Neoclassical Theory of Economic Growth: A Panel Data Approach,” IMF Staff Papers, September, 512-541. Lee, Frank C. (1996), Economic growth of OECD countries: focusing on Canada, International Economic Journal, Vol. 10, N. 2, Summer. Lensink, R. and White, H. (1999), “Is there an Aid Laffer Curve?”, mimeo. Lensink, R. and White, H. (2000a), “Aid allocation, poverty reduction and the Assessing Aid report”, Journal of International Development, Vol.12 pp.399-412. Lloyd, T., Morrissey, O. and R. Osei (2001), Problems with Pooling in Panel Data Analysis for Developing Countries: The Case of Aid and Trade Relationships. CREDIT Research Paper, No. 01/14, University of Nottingham. 215 Mankiw, N. G., Romer, D. and Weil, D. N. (1991), “A Contribution to the Empirics of Economic Growth,” Quarterly Journal of Economics, May, 407-437. McGillivray, M. and Oliver Morrissey (2001). A Review of Evidence on the Fiscal Effects of Aid. CREDIT Research Paper, No. 01/13, University of Nottingham. Parke, R.W. (1967) “Efficient Estimation of a system of Regression equations when Disturbances are both serially correlated and contemporaneously correlated”, Journal of American Statistical Association, Vol. 62, pp. 500-509. Renelt, D. (1991), “Economic Growth: a Review of the Theoretical and Empirical Literature”, Policy Research Working Paper 687, World Bank, Washington, D.C. Sala-i-Martin, X. X. (1997). "I just ran two million regressions". American Economic Review 87, 178183. White, H. (1992a), "What Do We Know About Aid’s Macroeconomic Impact? An Overview of the Aid Effectiveness Debate", Journal of International Development, 4(2), 121-137. White, H. (1992b). "The macroeconomic impact of development aid: A critical survey". Journal of Development Studies, Vol. 28, N. 2. White, H. (1999c), “Global poverty reduction: are we heading in the right direction?”, Journal of International Development, Vol.11, pp. 503-519. World Bank (1998). Assessing Aid. What Works, What Doesn’t, and Why. World Bank Policy Research Report. Oxford University Press. 216 Appendix 1 – Data results Tab. 1 – Summary statistics of the set of base variables, levels and first differences Mean Std. Dev. Maximum Minimum Mean Std. Dev. Maximum Minimum ggdppc? log(gdppc?) in_gdp? 4.4195 0.0326 0.1845 15.2140 0.1477 0.0972 106.5116 0.7252 0.7349 -54.3679 -0.7846 -0.0340 air? ex_gdp? 56.9172 0.2986 20.9985 0.1796 94.3000 0.8791 12.0000 0.0009 im_gdp? inf? log(ingdppc?) log(inpop?) 0.4024 54.3159 5.0705 14.9907 0.2324 835.3609 0.5763 1.1258 1.3089 23,773.1300 6.4606 17.7899 0.0630 -13.0566 4.0773 12.9456 d(ggdppc?) d(gdppcg?) d(in_gdp?) d(air?) d(ex_gdp?) d(im_gdp?) d(inf?) -0.2026 -0.0021 0.0006 -1.0142 0.0018 0.0016 0.1971 19.7533 0.1940 0.0494 0.4112 0.0536 0.0703 1,124.2550 119.4764 1.2704 0.2927 0.8000 0.2849 0.4486 21,786.2200 -115.0233 -1.2587 -0.3771 -3.7000 -0.2815 -0.3423 -23,231.2200 Tab. 2 – Summary statistics of the set of variables of interest, levels and first differences Mean Std. Dev. Maximum Minimum oda_gdp? ltd_gdp? fltd_gdp? 0.0893 0.5605 0.0418 0.0811 0.4415 0.0699 0.9474 2.6980 0.4326 0.000009 0.0086 -0.2755 fdi_gdp? tds_gdp? s_gdp? inted_gdp? log(inted_gdp?) 0.0114 0.0452 0.1035 1.6298 0.2486 0.0307 0.0466 0.1698 1.2042 0.6901 0.3065 0.7559 0.7898 5.3451 1.6762 -0.2578 0.0003 -0.8440 0.2968 -1.2148 d(oda_gdp?) d(ltd_gdp?) d(fltd_gdp?) d(fdi_gdp?) d(tds_gdp?) d(s_gdp?) Mean 0.0014 0.0253 -0.0019 0.0007 0.0011 -0.0005 Std. Dev. 0.0470 0.1329 0.0812 0.0267 0.0400 0.0601 Maximum 0.7655 0.9074 0.5057 0.2724 0.6437 0.4100 Minimum -0.4016 -0.7621 -0.3940 -0.2624 -0.6792 -0.4589 216 Tab. 3 – Mean of the set of variables of interest, cross-section and time specific values BEN BWA BFA BDI CMR CAF TCD ZAR COG CIV GAB GMB GHA KEN LSO MDG MWI MLI MRT MUS NER NGA RWA SEN SLE SDN SWZ TGO ZMB ZWE oda_gdp 0.090755 0.092545 0.128398 0.142325 0.040094 0.126304 0.112251 0.034406 0.068094 0.043156 0.028246 0.196903 0.058916 0.067079 0.184418 0.07848 0.168389 0.157651 0.211601 0.027716 0.12086 0.005781 0.16986 0.104417 0.085599 0.047204 0.06423 0.102696 0.120202 0.028178 tds_gdp fltd_gdp 0.018809 0.039966 0.019295 0.031889 0.013308 0.024902 0.023082 0.037822 0.039033 0.041796 0.017211 0.033562 0.00893 0.025275 0.02299 0.035647 0.099651 0.093862 0.105302 0.058497 0.062179 0.054859 0.054386 0.063757 0.043314 0.034808 0.071124 0.034498 0.027572 0.045393 0.043683 0.041088 0.056996 0.065075 0.022204 0.056529 0.09085 0.113727 0.054505 0.0316 0.040456 0.028995 0.054739 0.03293 0.007994 0.027722 0.051662 0.039256 0.039722 0.042363 0.015025 0.03855 0.0365 0.019034 0.050991 0.050178 0.113368 0.066804 0.045555 0.017534 fdi_gdp 0.004746 0.02136 0.002645 0.000919 0.006722 0.005556 0.011229 0.000432 0.024739 0.010256 0.019348 0.015614 0.007931 0.004353 0.048189 0.002886 0.009884 0.004565 0.00185 0.005361 0.006026 0.024283 0.006172 0.007027 0.001797 0.001626 0.045429 0.011969 0.013788 0.003826 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 217 oda_gdp? tds_gdp? 0.057443 0.019147 0.05899 0.020684 0.058979 0.024276 0.077099 0.017992 0.076629 0.02211 0.073945 0.022415 0.074577 0.025472 0.091703 0.027787 0.089149 0.031392 0.086825 0.041499 0.092146 0.049411 0.089168 0.050904 0.090874 0.055066 0.10457 0.064466 0.108649 0.072412 0.127017 0.07763 0.121054 0.064371 0.116457 0.062884 0.125014 0.05503 0.126542 0.056821 0.12581 0.051567 0.132824 0.045875 0.132195 0.045944 0.181778 0.065408 0.149472 0.073115 0.113073 0.050487 0.091717 0.045802 0.087386 0.047129 0.087878 0.050925 fltd_gdp? fdi_gdp? 0.027778 0.012283 0.028341 0.012499 0.046356 0.011132 0.047091 0.009072 0.036141 0.00025 0.03438 0.008626 0.063171 0.008377 0.06853 0.015131 0.065124 0.022263 0.064167 0.011684 0.027514 0.010395 0.046097 0.006684 0.02911 0.004542 0.005856 0.005327 0.096174 0.008946 0.123581 0.003365 0.146073 0.013433 0.00762 0.008139 0.013435 0.012953 0.071152 0.008675 0.029925 0.007031 0.006817 0.006682 0.021058 0.004725 0.073127 0.011289 0.050622 0.020187 -0.01046 0.023224 -0.02598 0.021781 0.046585 0.029771 -0.03042 0.02154 Tab. 4 - Matrix of selected cross-correlations of variables, mean values of four sub-periods Ai FDI_Gdp a FDI_Gdp b FDI_Gdp c FDI_Gdp d Ex 0,2 -0,22 0,77 -0,31 0,8 -0,13 0,29 Fd Fl Gg Gp Im 1 0,18 -0,05 0,27 1 0,1 0,51 0,32 0,71 1 -0,29 0,87 -0,79 0,92 1 -0,6 -0,5 0,27 0,02 0,18 FLTD_Gdp a 0,24 -0,42 0,34 0,1 FLTD_Gdp b 0,2 -0,15 -0,29 FLTD_Gdp c FLTD_Gdp d -0,36 -0,07 -0,6 1 1 1 1 0,37 0,32 -0,34 0,03 Ggdppc a Ggdppc b Ggdppc c Ggdppc d -0,23 -0,05 0,37 -0,35 0,3 0,51 0,32 -0,26 0,84 0,87 -0,34 0,24 0,17 -0,5 0,03 1 1 1 1 IN_Gdp a IN_Gdp b IN_Gdp c IN_Gdp d LTD_Gdp a LTD_Gdp b LTD_Gdp c LTD_Gdp d ODA_Gdp a ODA_Gdp b ODA_Gdp c ODA_Gdp d S_Gdp a S_Gdp b S_Gdp c S_Gdp d TDS_Gdp a TDS_Gdp b TDS_Gdp c TDS_Gdp d -0,34 0,28 -0,28 0,8 0,75 -0,25 0,68 0,73 -0,09 0,43 0,01 0,61 0,64 -0,61 0,23 0,01 0,1 0,25 0,2 -0,07 0,12 -0,14 -0,32 -0,07 -0,36 0,33 In Inf Ing Inp Int Lt Od 0,28 -0,06 0,21 0,2 0,3 0,1 0,75 -0,3 0,4 -0,49 0,33 -0,07 0,73 -0,37 0,5 -0,49 0,07 -0,32 0,01 0,78 -0,27 0,97 -0,02 0,33 0,45 0,61 -0,02 0,4 0,64 0,42 -0,27 -0,61 -0,36 0,22 0,23 -0,06 -0,72 0,76 -0,13 -0,04 -0,19 0,38 0,61 0,1 0,15 -0,41 0,24 0,01 0,21 -0,31 0,8 0,17 0,85 -0,04 0,29 -0,7 -0,1 -0,26 0,17 Sc St Td 0,44 0,1 0,31 -0,27 0,41 0,48 0,43 0,14 0,49 -0,23 -0,69 -0,28 0,63 -0,51 0,4 0,45 -0,09 0,35 0,22 0,64 -0,43 -0,18 -0,09 0,24 0,1 0,03 0,24 0,51 -0,51 0,3 -0,67 0,03 -0,36 -0,97 0,92 0,84 -0,36 0,77 -0,69 0,09 -0,5 -0,35 0,33 0,15 -0,79 0,41 -0,59 0,23 -0,54 0,34 0,03 -0,35 0,15 0,15 0,51 0,47 0,55 0,34 -0,15 0,3 -0,01 -0,07 -0,12 0,13 -0,44 0,41 0,04 0,81 -0,88 0,75 -0,81 0,6 1 1 1 1 -0,25 -0,77 -0,58 -0,26 0,07 0,35 -0,29 -0,26 0,46 0,52 0,89 0,66 0,6 0,62 0,71 0,32 0,08 0,16 0,09 0,37 0,61 0,1 0,47 0,01 -0,36 0,03 0,18 0,85 -0,5 0,53 -0,27 -0,26 -0,54 0,84 -0,59 0,42 0,14 -0,64 -0,89 0,06 -0,14 0,88 0,65 -0,1 0,38 0,41 0 0,24 0,05 -0,25 -0,52 0,04 -0,38 0,42 -0,19 0,72 0,64 0,04 -0,03 0,64 0,51 0,84 0,15 0,27 0 0,51 0,38 -0,68 0,49 0,72 -0,77 0,2 0,57 -0,08 -0,84 0,14 -0,42 -0,51 -0,65 -0,23 -0,05 0,97 0,41 0,82 0,84 -0,14 0,67 0,42 0,14 -0,64 -0,89 1 1 1 1 -0,08 -0,12 0,1 0,34 -0,08 0,07 -0,24 -0,63 -0,5 -0,11 -0,1 0,58 0,1 0,21 0,21 0,03 0,09 0,3 0,35 -0,5 -0,53 -0,37 0,53 0,41 0,74 -0,62 -0,09 -0,04 -0,35 0,34 -0,31 -0,29 -0,09 -0,81 0,19 -0,23 0,05 0,78 -0,92 -0,59 0,17 0,15 0,43 -0,76 -0,26 -0,41 -0,56 -0,52 0,07 0,04 0,16 0,47 -0,51 0,49 0,65 -0,76 0,74 0,4 -0,51 0,85 0,56 S -0,12 0,47 0,21 0,65 0,79 -0,09 0,4 0,4 -0,59 0,56 0,43 0,14 -0,23 -0,28 0,15 0,51 0,55 -0,15 1 1 1 1 -0,51 -0,19 0,25 -0,85 -0,44 -0,85 0,25 -0,42 -0,31 0 -0,33 -0,75 0,33 0,46 -0,08 0,51 0,14 0,41 0,38 -0,51 0,22 0,19 0,52 -0,3 0,53 -0,23 0,4 0 -0,19 -0,54 0,51 0,66 -0,24 0,88 -0,4 0,4 -0,25 -0,85 -0,62 0,74 0,62 0,19 0,6 0,52 -0,53 -0,38 -0,85 1 1 0,66 1 0,72 1 0,53 0,64 0,36 0,18 0,45 0,6 0,17 0,31 0,4 -0,07 0,35 0,08 -0,18 0,06 -0,13 0,67 0,72 -0,31 0,36 -0,34 0,21 -0,27 0,45 -0,12 -0,29 0,07 0,16 -0,27 0,02 -0,12 0,62 0,64 0 0,18 0,22 0,56 -0,86 0,42 0,41 -0,09 0,13 0,03 0,24 0,09 -0,26 0,29 -0,03 0,32 0,04 -0,33 0,45 0,68 0,64 -0,78 0,59 0,48 0,35 -0,44 -0,35 0,57 0,37 0,63 0,34 0,34 -0,21 -0,03 -0,75 0,6 a = 1971-78, b = 1979-85, c = 1986-92, d = 1993-2000 Ai = Air, Ex = Ex_Gdp, Fd = Fdi_Gdp, Fl = Fltd_Gdp, Gg = Ggdppc, Gp = Gpop, Im = Im_Gdp, In = In_Gdp, Inf, Ing = Ingdppc, Inp = Inpop, Int = Inted_Gdp, Lt =Ltd_gdp, Od = Oda_gdp, S = S_gdp, Sc = Scr_gdp, St = Ste_gdp, Td = Tds_gdp 218 1 1 1 1 Tab. 5 – corss-contry OLS regressions’ results (with White Heteroskedasticity-Consistent Standard Errors & Covariance correction) Period 1971-78 Dependent Variable: GGDPPC Variable Coefficient IN_GDP INGDPPC AIR C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 24.41083 -0.021204 -0.034912 14.46184 0.307599 0.221049 3.529349 298.9512 -72.88334 2.003127 Std. Error t-Statistic Prob. 10.10535 0.008385 0.044991 4.712778 2.415634 -2.528725 -0.775981 3.068644 0.0237 0.0184 0.4453 0.0053 Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion F-statistic Prob(F-statistic) 13.10807 3.998890 5.491667 5.681982 3.553999 0.029328 Dependent Variable: IN_GDP Variable Coefficient Std. Error t-Statistic Prob. S_GDP 0.444205 0.19059 2.330687 0.0281 FLTD_GDP 0.634448 0.583513 1.087291 0.2873 ODA_GDP 1.172296 0.295704 3.964423 0.0005 FDI_GDP 2.472084 1.090141 2.267673 0.0322 R-squared 0.374272 Mean dependent var 0.208989 Adjusted R-squared 0.299185 S.D. dependent var 0.093381 S.E. of regression 0.078174 Akaike info criterion -2.13232 Sum squared resid 0.152778 Schwarz criterion -1.94373 Log likelihood 34.91869 F-statistic 4.984498 Durbin-Watson stat 1.957165 Prob(F-statistic) 0.007583 Period 1979-85 Dependent Variable: GGDPPC Variable Coefficient IN_GDP INGDPPC EX_GDP AIR C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 18.62626 -0.001769 1.003702 -0.021127 -1.693483 0.126140 -0.040309 4.899024 504.0091 -75.43087 2.136494 Std. Error t-Statistic Prob. 17.02094 0.004702 10.21443 0.059908 5.886582 1.094314 -0.376183 0.098263 -0.352660 -0.287685 0.2862 0.7106 0.9227 0.7279 0.7764 Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion F-statistic Prob(F-statistic) 0.168962 4.803174 6.186990 6.428932 0.757828 0.564239 Dependent Variable: IN_GDP Variable Coefficient Std. Error t-Statistic Prob. C 0.14517 0.038262 3.794085 0.0009 S_GDP 0.016219 0.198453 0.081729 0.9356 ODA_GDP 0.053585 0.281328 0.19047 0.8506 FLTD_GDP 0.314678 0.640564 0.491251 0.6279 FDI_GDP 3.479832 1.285744 2.706473 0.0126 R-squared 0.373059 Mean dependent var 0.204103 Adjusted R-squared 0.264026 S.D. dependent var 0.083778 S.E. of regression 0.071872 Akaike info criterion -2.26743 Sum squared resid 0.118808 Schwarz criterion -2.02954 Log likelihood 36.74401 F-statistic 3.421516 Durbin-Watson stat 2.372427 Prob(F-statistic) 0.024636 219 Period 1986-92 Dependent Variable: GGDPPC Variable Coefficient INGDPPC IN_GDP AIR EX_GDP C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 0.004974 35.79654 0.082083 8.075365 -10.50056 0.545869 0.463300 3.824069 321.7170 -71.76212 1.668816 Std. Error t-Statistic Prob. 0.002697 10.28552 0.046015 4.952548 4.241000 1.844377 3.480286 1.783837 1.630548 -2.475962 0.0786 0.0021 0.0883 0.1172 0.0215 Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion F-statistic Prob(F-statistic) 5.142483 5.219870 5.686083 5.926053 6.611038 0.001187 Dependent Variable: IN_GDP Variable Coefficient Std. Error t-Statistic Prob. C 0.177871 0.065279 2.724768 0.015 S_GDP -0.38804 0.153333 -2.53069 0.0223 FLTD_GDP 0.357974 0.656912 0.544935 0.5933 ODA_GDP 0.099794 0.267722 0.372751 0.7142 FDI_GDP -0.64388 0.929168 -0.69296 0.4983 AIR -0.00139 0.000881 -1.58035 0.1336 EX_GDP 0.240653 0.115383 2.085694 0.0534 INF -6.25E-05 4.52E-05 -1.38216 0.1859 INGDPPC 7.68E-05 7.29E-05 1.053977 0.3076 R-squared 0.631678 Mean dependent var 0.184077 Adjusted R-squared 0.424498 S.D. dependent var 0.084337 S.E. of regression 0.06398 Akaike info criterion -2.37677 Sum squared resid 0.065495 Schwarz criterion -1.89289 Log likelihood 40.89806 F-statistic 3.048923 Durbin-Watson stat 1.439624 Prob(F-statistic) 0.024994 Period 1993-99 Dependent Variable: GGDPPC Variable Coefficient INGDPPC IN_GDP AIR EX_GDP C R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood Durbin-Watson stat 0.000914 13.36158 -0.034494 -4.186263 -1.125452 0.143488 -0.027815 3.553637 252.5667 -64.38345 1.344382 Std. Error t-Statistic Prob. 0.001218 16.16944 0.041661 5.574594 3.574488 0.750328 0.826348 -0.827960 -0.750954 -0.314857 0.4618 0.4184 0.4175 0.4614 0.7561 Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion F-statistic Prob(F-statistic) -1.098421 3.505223 5.550676 5.794452 0.837628 0.517384 Dependent Variable: IN_GDP Variable Coefficient Std. Error t-Statistic Prob. C 0.148438 0.05525 2.686653 0.0177 S_GDP 0.623182 0.109161 5.708837 0.0001 FLTD_GDP 0.846821 0.450553 1.879516 0.0811 ODA_GDP 0.305614 0.103847 2.942935 0.0107 FDI_GDP 0.936297 0.960364 0.97494 0.3461 EX_GDP 0.126636 0.091404 1.38545 0.1876 AIR 0.000864 0.000528 1.637166 0.1239 INGDPPC -2.36E-05 1.03E-05 -2.28867 0.0382 GPOP -0.04936 0.019565 -2.52268 0.0244 R-squared 0.842228 Mean dependent var 0.183563 Adjusted R-squared 0.729534 S.D. dependent var 0.061419 S.E. of regression 0.031942 Akaike info criterion -3.74961 Sum squared resid 0.014284 Schwarz criterion -3.2133 Log likelihood 57.8701 F-statistic 7.473567 Durbin-Watson stat 2.483439 Prob(F-statistic) 0.000444 220 Tab. 6 – Joint levels of aid and external debt conditioned on some iso-investment rates ODA_GDP vs. LTD_GDP (IN_GDP<10%) ODA_GDP vs. TDS_GDP (IN_GDP<10%) 0.3 0.2 0.2 ODA_GDP ODA_GDP 0.3 0.1 0.1 0.0 0.0 0.5 1.0 1.5 0.0 0.00 2.0 0.05 LTD_GDP 0.8 0.8 ODA_GDP ODA_GDP 1.0 0.6 0.4 0.2 0.6 0.4 0.2 0.5 1.0 1.5 0.0 0.00 2.0 0.05 LTD_GDP 0.10 0.15 0.20 TDS_GDP ODA_GDP vs. TDS_GDP (IN_GDP>15% and <20%) ODA_GDP vs. LTD_GDP (IN_GDP>15% and <20%) 0.6 0.6 0.4 0.4 ODA_GDP ODA_GDP 0.20 ODA_GDP vs. TDS_GDP (IN_GDP>10% and <15%) 1.0 0.2 0.0 0.0 0.15 TDS_GDP ODA_GDP vs. LTD_GDP (IN_GDP>10% and <15%) 0.0 0.0 0.10 0.2 0.5 1.0 1.5 2.0 0.0 0.0 2.5 LTD_GDP 0.2 0.4 TDS_GDP 221 0.6 0.8 ODA_GDP vs. TDS_GDP (IN_GDP>20% and <26%) 0.5 0.5 0.4 0.4 ODA_GDP ODA_GDP ODA_GDP vs. LTD_GDP (IN_GDP>20% and < 26%) 0.3 0.2 0.3 0.2 0.1 0.1 0.0 0.0 0.5 1.0 1.5 2.0 0.0 0.00 2.5 0.05 ODA_GDP vs. LTD_GDP (IN_GDP>26% and <35%) 0.20 0.25 ODA_GDP vs. TDS_GDP (IN_GDP>26% and <35%) 0.4 0.4 0.3 0.3 ODA_GDP ODA_GDP 0.15 LTD_GDP LTD_GDP 0.2 0.1 0.2 0.1 0.0 0.0 0.5 1.0 1.5 2.0 0.0 0.00 2.5 0.05 LTD_GDP 0.10 0.15 0.20 TDS_GDP 0.4 ODA_GDP vs. TDS_GDP (IN_GDP>35%) 0.4 0.3 0.3 ODA_GDP ODA_GDP vs. LTD_GDP (IN_GDP>35%) ODA_GDP 0.10 0.2 0.2 0.1 0.1 0.0 0.0 0.0 0 1 2 3 0.1 0.2 TDS_GDP LTD_GDP 222 0.3 0.4 Tab. 7 – pooled analysis: linear effects on growth. Dependent Variable: D(GDPPC?) Y = D(GDPPC?) OLS FE RE-GLS GLS SUR D(GDPPC?(-1)) 0.785 0,689 0,947 0,708 0,678 D(IN_GDP?) 1,884 88.1 130.4 123.9 91.5 D(AIR?) 0,489 -4.2 1.1 0.732 -4.4 D(EX_GDP?)+ D(IM_GDP?) 108.3 97.8 36.8 46.5 98.8 D(FLTD_GDP?) 208.9 172.6 312.8 89.6 145.5 D(ODA_GDP?) 34.4 33.4 59.8 34.0 43.4 D(FDI_GDP?) 399.4 409.4 305.5 203.0 367.4 D(TDS_GDP?) -47.2 -76.3 -66.4 -17.7 -76.6 Adjusted R-squared 0.619 0.623 0.544 0.529 0.622 Durbin-Watson stat 1.468 1.460 1.456 1.503 1.441 Note: RE-GLSis based on Variance Components, GLS on Cross Section Weights Underscored and italics values are non significant, bold values are significant at 1%, bold and italics at 5%, bold and underscored at 10%. 223 Appendix 2 – A case-study of foreign capital flows inter-links: Italian development aid and debt relief 1. Italian Official Development Assistance in 2000. Volume and instruments In terms of disbursement of Official development assistance (ODA), the latest detailed data available are from the year 2000, and compared to the previous three years, 1997 – 1999. Table. 1: Net payments by Italian ODA (in millions of US dollars). Period 1997-2000 1997 1,265.55 1998 2,355.55 (86,1%) 1999 1,805.72 (-23,3%) 2000 1,376.26 (-23,8%) 278.63 (22%) 453.73 (35.9%) 360.78 (28.5%) 531.69 (22.6%) 763.93 (32.4%) 332.43 (14.1%) 246.53 (13.7%) 450.72 (25%) 550.76 (30.5%) 261.11 (19%) 376.80 (27.4%) 524.81 (38.1%) 57.79 (4.6%) 40.42 (1.7%) 53.19 (2.9%) 27.16 (2.0%) developmental food aid (as % of total aid) 16.11 (1.3%) 39.43 (1.7%) 43.78 (2.4%) 31.95 (2.3%) emergency and distress relief (as % of total aid) Soft loans (as % of total aid) Multilateral ODA (% del totale) of which: UN system (as % of total aid) 50.25 (4.0%) 92.95 (7.3%) 811.83 (64.1%) 16.57 (0.7%) 431.50 (18.3%) 1,591.62 (67.6%) 102.97 (5.7%) - 100.04 1,355.00 (75%) 72.36 (5.3%) -148.01 999.46 (72.6%) 163.60 (12.9%) 172.31 (7.3%) 151.22 (8.4%) 202.36 (14.7%) European Commission (as % of total aid) 613.70 (48.5%) 713.35 (30.3%) 679.00 (37.6%) 637.62 (46.3%) The World Bank Group (as % of total aid) 17.43 (1.4%) 498.78 (21.2%) 302.83 (16.8%) 18.70 (1.4%) Regional development banks (as % of total aid) 1.35 (0.1%) 193.31 (8.2%) 146.65 (8.1%) 75.89 (5.5%) Total ODA (as % annual variation) of which: to Sub-Saharan African countries (as % of total aid) Bilateral ODA (as % of total aid) bilateral grants (as % of total aid) of which: technical co-operation (as % of total aid) Source: Elaborations on DAC-OECD, 2001 For the year 2000, as for the preceding years, one may note the Italian predilection for the use of the multilateral channel. 65-75% of total Italian ODA use the multilateral channel, which is the opposite of what happens in other donor countries. This policy to provide such a high proportion of ODA through the multilateral channel is not explicitly set out in the 1995 Inter-ministerial Committee for Economic Planning (Comitato Interministeriale per la Programmazione Economica, 224 CIPE) Guidelines governing Italian development co-operation, although Prospective 2000 1 states that 60% of Italian co-operation will be implemented in co-operation with international organisations. In 2000 Italian resources underwent a serious reduction ($M 1,376.26), and cooperation via the bilateral channel correspondingly felt the consequences of it ($M 376.8, or 27.4% of total Italian aid). Quantitative data for 2000 confirm many basic elements, which were already to be seen in the three years before. While the Finance Law for the year 2000 was passing through Italian Parliament, the smallness of resources became apparent. The Foreign Affairs Ministry had been assigned a mere 672 billion Lire to fund all the year’s activities. 8.2% of the total (55 billion Lire) were destined to running costs in the Directorate-General for Development Co-operation (DGCS, Direzione Generale per la Cooperazione allo Sviluppo), which oversees cooperation and development policies in the same Ministry, and 617 billion Lire was destined to on-site cooperation activities. A further 51 billion Lire had been assigned to the Foreign Affairs Ministry for making obligatory contributions to Italian and international organisations. As well as this, 400 billion Lire were transferred – pursuant to Law N° 266/1999 – from the rotating Fund for aid credits to the cooperation Fund for grants. In the grants Fund in the year 2000, therefore, 1,068 billion Lire (617 + 51 + 400) were available for development cooperation activities; in the rotating Fund, on the other hand, monies available for soft loans in the year 2000 stood at about 2,200 billion. Overall, totalling the sums available for grants and soft loans, Italian bilateral development cooperation policy in the year 2000 could count on more than 3,000 billion Lire. From the total sums available, the administrative machine would not therefore appear particularly able to manage these resources. Indeed, in terms of concrete disbursements, Italian co-operation has met with difficulties in running the finalisation of the initial phase, which has often meant the impossibility of paying out what has been laid aside for this purpose from year to year: if we take, for example, the case of soft loans, payments are well below budget capabilities. The important tool of credit repayments with long grace periods for financing joint ventures (pursuant to Art. 7 of Law N° 49/1987), so important in helping the private sector, and particularly the structure of small- and medium-sized firms, has been largely under-used. In 2000, approximately 90 billion Lire of funds were available at Mediocredito Centrale. In 1999, eight finance projects were approved, for a total of 22.8 billion Lire; in 1998, only seven projects were approved, for a total of 14.8 billion. Moving forward a major strengthening of the management structure and capacities for Italy’s development co-operation policy, supporting the work of DGCS, is considered the most urgent element in making Italian policy work better, as can be seen from the Peer Review of activities carried out between 1996 and 2000. The Review was made by the Development Assistance Committee (ODA) of the OECD, together with Canadian and Swedish reviewers between March and June of 2000. 2. External debt reduction as the main component of Italian ODA If on the one hand we can see from the data in Table 1 a large reduction in Italian ODA resources (especially in terms of the bilateral channel), we can also see that this figure has been resized by the particular nature of operations put in place to deal with funds for development aid. To sum up, bilateral aid has tended to reduce the importance of operative actions (development programmes and projects) in order to move over to the cancellation and rescheduling of developing countries’ external debt. Debt cancellation is marked down as a grant, whereas rescheduling is a soft loan. In Italy, there is no real additional nature to development aid – recommended, among other things, by 1 Prospective and programmatic report on development co-operation activities for the year 2000 (Prospective 2000) is the annual report on development co-operation policy. In this report, the Minister of Foreign Affairs established poverty alleviation as the main priority of the Italian aid programme. 225 the World Bank and International Monetary Fund’s Heavily Indebted Poor Countries (HIPC) Initiative and subscribed to as an objective by most of donors – of measures for reducing external debt. In fact, they become the single or main element of bilateral policies for Italian development co-operation policy. It does represent the main gap between rhetoric and performance. An analysis of the fifteen main beneficiaries of Italian aid during the year 2000 throws some clear light on the subject. First of all, concerning one of the main problems Italian authorities underlined in Prospective 2000, namely, of a tendency toward dispersion and lack of critical mass resulting from spreading aid over too many programmes and projects, the fifteen main beneficiaries of Italian cooperation – when considering both grants and loans – received 94.35% of the net total paid out in Italy (the first ten beneficiaries received 76.67% of total bilateral aid). Overall, 89 developing countries are seen to be overall net beneficiaries of the flow of Italian overseas development aid; eighteen show net overall losses (Croatia, Slovenia, and Turkey in Europe; Algeria, Egypt, and Tunisia in Northern Africa; Ghana and Kenya in Sub-Sahara Africa; Argentina, Columbia, Ecuador, Jamaica, Guatemala, Peru, the Dominican Republic, and Uruguay in America; China and India in Asia). Table 2: The fifteen main beneficiaries of Italian bilateral aid during 2000 (in millions of US$) Country Uganda Bosnia-Erzegovina Cameroun Ethiopia Zambia Malta Yugoslavia Fed. Rep. Honduras Benin Eritrea Albania Senegal Mozambique Palestinian Nat.Auth. Burkina Faso Sub-total Total net (a) cancelled (b) rescheduled aid debt debt 82.09 78.82 32.97 25.03 26.96 24.41 1.92 25.97 24.02 22.83 20.83 19.30 19.07 10.49 5.36 19.07 18.52 18.61 18.34 2.04 0.54 15.21 10.70 13.09 11.78 8.20 6.95 355.51 199.79 7.82 (a)+(b) as % of total aid 96.02 75.92 97.66 0.00 95.05 0.00 0.00 83.11 97.12 0.00 14.07 70.35 0.00 0.00 84.76 58.40 Source: Elaborations on DAC-OECD, 2001 During the year 2000, the only countries which benefited from Italian bilateral aid were those which obtained a cancellation of their overseas debt (Uganda, Cameroon, Zambia, and Benin), or where this was the main factor (Bosnia-Herzegovina, Honduras, Senegal, and Burkina Faso). Exceptions are to be found – if one does not take into consideration the special agreement made between Italy and Malta – those countries which show the lasting nature of Italian interests owing to their proximity or the complex emergency pay-out (Albania and the Federal Republic of Yugoslavia), the post-war emergency situation due to natural disasters in regions where Italy used to have interests such as the Horn of Africa (Eritrea and Ethiopia), and Mozambique. The Palestinian Authority is a now-consolidated reality where priority Italian action is required in development and technical assistance programmes. An aspect which directly affects the overseas development aid budget regards the emphasis which Italy, more so than other donor countries, has given to the cancellation of overseas debt in countries 226 which traditionally have high debt levels. This was confirmed in Italian Law N° 209, dated 29 July 2000, and its application, passed by Parliament as N° 185, dated 4 April 2001. Italian international role as a member of the Group of Seven Leading Industrialized Countries (G-7) - with the Presidency in 2001 -, the European Union (EU) and all of the major multilateral institutions as well as the particular interest of Italian public opinion on debt relief makes Italy an important actor in the international commitment toward external debt reduction of poor countries. Moreover, Italy experienced two important civil society’s campaigns on debt cancellation, one was Campagna Sdebitarsi, linked to the International Jubilee 2000 network, and the other was the Campagna della Conferenza Episcopale Italiana, linked to the Vatican Catholic State. Table 3: Share of debt relief 2 in Italy compared to DAC total, year 2000 (in millions of US$) Net ODA Italy Total DAC 1,376 53,737 Net ODA (as % of of which: (as % of for debt relief net ODA) bilateral debt relief) HIPC 239 17.3 217 90.8 204 2,236 4.2 1,988 88.9 1,180 (as % of debt relief) 85.4 52.8 Source: Elaborations on DAC-OECD, 2001 In the light of these data, the concrete problem to be found today, at an Italian and – more generally – international level is the ratio between debt reduction/cancellation and development co-operation. The international debt cancellation initiative is expressly based on the maintenance and strengthening of overseas development aid’s currently undertaken commitments, to which debt cancellation should be merely an adjunct. Unfortunately, this is not the case, and in view of the very long times required in debt cancellation procedures, there is the danger that donor countries budget for overseas development aid initiatives over many years. This does not necessarily mean that they actually pay money: they merely budget for it. For the poorest countries, differently from middle income countries who are still paying-back their debt service to bilateral creditors, no added resources are freed up as they have already suspended repayments to their bilateral creditors. The non-additional nature of resources for bilateral cancellation of foreign debt is confirmed in the given summary, which refers to 1999 and 2000, for items in ODA which can be traced back to this type of aid. A quick glance at the many items present in the various types of ODA dealing with debt cancellation is enough to show how much development cooperation is tied up – at least in budgets – with measures for reducing overseas debts in developing countries. 3 In Italy, the figures for the year 2000 merely underline what had already been understood in 1999, when the two main beneficiaries of ODA were Congo and Tanzania. They were not the countries where most aid was destined, nor did they benefit from development cooperation budgets, but they had their external debts restructured by Italy (78.16 million dollars for Congo, 19.15 million for Tanzania). 2 In this case, comparison is made between debt repayment (accounted for as part of grants) and rescheduling (part of soft loans) in the bilateral channel, and finance from the Trust Fund for HIPC activities in the multilateral channel. 3 The relationship between debt cancellation and overseas development aid is not simple, even when considered from a purely accounting point of view. Total resources at multilateral level for debt reduction initiatives in poor countries correspond to the current net value of the foreign debt which shall not be paid back. Accounting in donor countries, such as Italy, considers cancellation and rescheduling as part of overseas development aid, although they do not generate new resources. This is because they were registered as ODA flows at the moment of original payment. 227 Table 4: ODA items concerning foreign debt reduction for developing countries (1999-2000, in millions of US$) Amounts Amounts Net Amounts extended received 1999 2000 1999 2000 1999 2000 1,998.98 1,598.93 -193.26 -222.67 1,805.72 1,376.26 643.98 599.47 -193.26 -222.67 450.72 376.80 550.76 524.81 0.00 0.00 550.76 524.81 I. ODA I.A. Bilateral ODA 1. Grants of which: i) Debt forgiveness, total (incl. forgiven interest) (a + b + c) 101.92 201.47 101.92 201.47 101.92 0.00 201.47 0.00 93.22 74.66 3.41 3.41 15.04 15.04 89.81 59.62 a) ODA claims b) OOF claims c) Private claims Memo: Grants for debt service reduction ii) Other action on debt (a + b + c + d) a) Service payments to third partiesi b) Debt conversion c) Debt buybacks d) Other 2. Non-grant bilateral ODA / 101.92 201.47 / 0.00 / 101.92 201.47 / 0.00 / 101.92 201.47 / 0.00 0.00 / 0.00 / 0.00 / 0.00 / 0.00 -193.26 -222.67 -100.04 -148.01 of which: a) Rescheduling, total (a.i + a.ii) a.i. ODA claims (capitalised interest) a.ii. OOF claims b) Other lending c) Acquisition of equity (total) / d) Offsetting entry for debt forgiveness Memo: - Loans included in Associated Financing packages / / 1,355.00 999.46 1,355.00 999.46 - Interest received - Offsetting entry for forgiven interest I.B. Multilateral ODA 1. Grants and capital subscriptions, total 0.00 0.00 3.41 15.04 3.41 15.04 0.00 -193.26 -222.67 -103.45 -163.05 / 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1,355.00 999.46 0.00 0.00 1,355.00 999.46 of which: 297.06 5.77 146.65 i) IDA ii) IBRD,IFC,MIGA iii) Regional development banks 18.70 75.89 297.06 5.77 146.65 18.70 75.89 Memo (bilat. + multilat.): 23.80 - HIPC Initiative - IDA Debt Reduction Facility 189.63 189.63 189.63 0.00 II. Other official flows (OOF) (II.A+II.B) II.A. bilateral OOF 1. Rescheduling, total 1.1 Non-concessional rescheduling (a + b) a) OOF claims (capitalised interest) b) Private sector claims 189.63 / 0.00 / / / / 1.2 OOF component of debt service reduction 2. Offsetting entry for debt relief II.B. OOF on Multilateral channel Memo: - Interest received on OOF, total, (bil. + multil.) - Bilateral - Multilateral - Offsetting entry for forgiven interest Source: Elaborations on DAC-OECD, 2001 228 / 0.00 23.80 / 0.00 103.01 -170.77 -206.50 18.86 -103.49 103.01 -170.77 -206.50 18.86 -103.49 103.01 0.00 0.00 189.63 103.01 9.48 0.00 0.00 0.00 9.48 9.48 0.00 9.48 0.00 93.53 189.63 93.53 -170.77 -206.50 -170.77 -206.50 0.00 0.00 0.00 0.00 0.00 0.00 -120.77 -88.50 -120.77 -88.50 In the right perspective, overseas debt and poverty reduction policies via overseas development aid should balance one another out: when the debt crisis has been sorted out via cancellation policies, a new season of development financing should begin. In this way, developing countries would not restart on the downward spiral of debt, but would promote cooperation strategies for effectively reducing poverty. On the other hand, improper use of debt cancellation will mean giving up on playing an active role in development cooperation, which will then become a mechanism for reducing debt, without offering any active measures for international cooperation. It is also true that, for Italy, mention may be made in general of prevailing trends on the operating level of “reducing” commitments. In other words, whereas Italian cooperation tends to be multilateral, this depends on the freeing up in the bilateral channel of other commitments to multilateral cooperation. The latter thus become preponderant because the conventional commitment to bilateral aid lessens. And the same thing can also be seen within the bilateral channel: Italy’s commitment to reducing developing countries’ external debt becomes ever more important, not because this item increases in absolute terms, but simply because this money, once budgeted for, becomes an expenses commitment when other bilateral channel expenses are drastically reduced. A measure of this is given by the graph which – in terms of annual ODA commitments – compares trends in Italian bilateral ODA and the amount spent on reduction of developing countries’ external debt (which is, in turn, broken down into grants and soft loans). 3000 Bilateral ODA commitment, technical assistance excluded (US$ million) Debt Relief (grants) Debt Relief (soft loans) 2500 Debt Relief (total) total 2000 1500 1000 500 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1997 1998 1999 2000 -500 Source: Elaborations on DAC-OECD, 2001 Trends in the use of bilateral spending and cuts in developing countries’ international debt follow their own lines, and do not necessarily coincide, up to the middle of the 1990’s. Until then, indeed, bilateral spending first follow a progressive tend towards growth (early Eighties), and then enjoy a period of relative abundance of resources (late Eighties) followed by a crisis period (Nineties). On the other hand, in the case of bilateral resources for reducing developing countries’ external debt, the birth of this aid component coincides with the period of abundance of resources, and this is also the period when external debt rescheduling was implemented after agreements made at the Club de Paris. Thereafter, during the early Nineties, this commitment was maintained and strengthened, and the grant element was introduced to lighten debt burdens: this was used to compensate for trends in the soft loans channel. And so we finally reach the late Nineties: then, bilateral cooperation was 229 noticeably reduced, whereas cooperation to reduce developing countries’ external debts became the locomotive for other commitments. This can be seen in the mirroring of the two paths; developing countries’ external debt reduction measures became the main ingredient of bilateral expenses budgeting and, at the same time, took over from the policy of rescheduling (which may be expressed in terms of soft loans) as debt cancellation (which may be expressed in terms of grants), as was developed during international debates, and, especially, agreed upon at the G-8 summit in Cologne in 1999. In this sense, therefore, the idea of a policy for excluding or reducing commitments means that with the drying up of resources, new trends in Italian bilateral aid policies – starting with the commitment to reducing developing countries’ external debt – translate at best into the maintaining of commitments already undertaken when resources were relatively plentiful. 3. The structure and main components of Italian aid in 2000 The evidently different nature of the two main instruments used in development cooperation (grants and soft loans) require a closer examination in terms of the trends seen in these two channels of Italian overseas development aid during the year 2000. At a resources level, as was mentioned above, the grant channel was “subsidised” by the loans one over the last two years, and the graph below shows this extremely clearly. What this means is that a minimum amount of resources was available for making grants, at the cost of a corresponding reduction in credit resources – as can be seen in the mirroring between the two curves – which determines for both 1999 and 2000 an overall debt in credit expenditure. 2.500,00 ODA, Total Bilateral grants 2.000,00 Soft loans 1.500,00 1.000,00 500,00 0,00 1997 1998 1999 2000 -500,00 Source: Elaborations on DAC-OECD, 2001 A general consideration which may be made concerning the Italian situation is a reminder of development financing, at the centre of the United Nations conference in Monterrey in Mexico. ODA is not in itself a motor for economic growth, but can be a useful catalyst for mobilising internal and international resources. Foreign debt crises, although a real and deep danger, risk making people distrust the tool of credits (even those made on concessional terms) for developing countries, without finding, in the short term, any other solutions or mechanisms to replace them. In Italy, over the last two years, the credit system has been widely used as a resource to draw upon to 230 keep cooperation grants going, thanks to the success of the rotating Fund, which has seen the return of capital loans. Thanks are also due to red tape, whereby monies budgeted for were not spent. Another method for using the rotating Fund set out in Law N° 49/1987 has been the financing of venture capital operations for the internationalisation of small- and medium-sized Italian enterprises, especially in the Mezzogiorno. Finally, the law on Italian participation in the stabilising, reconstruction, and development of the Balkans (Law N° 84, dated 21 March 2001) states that part of the fund set up for helping companies to take part in the stabilising, reconstruction, and development process in the area should be paid out by the Ministry for Foreign Trade, in terms of micro-credits (up to 200 billion Lire) and an increase in rotating Fund credits. These monies are to be used for long-term investment in factories in countries outside the European Union, by providing free guarantees for finance to companies damaged by the lack of payments from Yugoslavia after the war in that area. The question which must now be asked is this: can this process be sustained in the long term? In other words, will the monies paid into the rotating Fund be used up quickly, requiring a new input of funds with future Finance Bills, and in what scenario? Will this tool be used less in financing, meaning that only the grant channel will be used? Or will it be used to guarantee access to credits for countries with medium-to-low earnings, as is already the case? It is easy to imagine an increasing use of the credit component in overseas development aid, in terms of the interest already seen in economic internationalisation of Italian industry. In this sense, signs can be seen both in central government (the idea of the “Italian system” involvement in the Balkans stability pact, and the strengthening of links via measures for promoting foreign trade) and in the growing interest shown by regional and local government. In any case, data on the geographical destination of concessions and credits identify a series of priorities which do not coincide. As far as grants are concerned (when dealing with those which are not primarily aimed at reducing external debt), a very worrying trend emerges in the year 2000 in terms of the resources actually paid out. Based on the general data, as shown in the table giving details for the period 1997-2000, in view of the small size of resources available, the increase in monies for grants in bilateral aid may seem interesting. Indeed, the amounts rose from 360.78 million dollars in 1997 to 332.43 in 1998, rising to 550.76 million in 1999 and then falling again slightly to 524.81 million dollars in 2000. However, when one removes the money destined for developing countries’ external debt reduction (which amounted to 38.39% of the bilateral grants total in 2000), one quickly realises that bilateral grants money has been decreasing over the last two years, and now stands at about 300-320 million dollars of resources effectively paid out for development aid. An amount, in real terms, much lower than the bilateral grant total (minus the amount for debt reduction) paid by any other country in G-7, and most members of ODA. Only Austria, Finland, Greece, Ireland, Luxembourg, New Zealand, and Portugal – which can in no way be compared with Italy in terms of riches produced each year in absolute terms – destined fewer resources. Table. 5: Net bilateral concessions paid, in millions of US dollars (and percentage of total bilateral concessions), year 2000 Italy Grants, Project and Technical Food aid Emergency Debt relief Support to Grants, total total programme assistance NGOs (excluded debt aid relief) 524.81 106.57 27.16 31.95 72.36 201.47 35.27 323.34 (20.3) (5.2) (6.1) (13.8) (38.4) (6.7) Source: Elaborations on DAC-OECD, 2001 If one considers the main beneficiaries during 2000 of bilateral grant cooperation, one sees they are countries which confirm what emerged during the general interpretation of bilateral cooperation. The “special” nature of absolute continuity over time is only seen in the protocol signed with 231 Malta 4 , and then geopolitical reasons enter into play (Albania, Federal Republic of Yugoslavia), interest in the Horn of Africa (Eritrea and Ethiopia), Mozambique and Angola in Sub-Saharan Africa, and Palestine in the Middle East. Bringing up the rear are other countries where the importance of the grant channel is more due to the follow-on effects of the credit channel (Egypt), other planning considerations currently underway (Venezuela, Sudan, and Djibouti), or to an Italian presence in a country with good economic prospects, where the “anti-Western” ideological connotations have not yet been removed, nor has full integration with the world economy begun (Cuba). 5 Table 6: The 15 main beneficiaries of bilateral Italian grant, 2000 (in millions of US$) Ethiopia Malta Yugoslavia Fed. Rep. Eritrea Mozambique Albania Palestinian Nat. Auth. Senegal Bosnia-Erzegovina Angola Egypt Venezuela Sudan Djibuti Cuba Total Grants, Total of which: Technical assistance 26.16 2.02 20.94 19.32 0.01 16.40 3.29 13.09 0.57 11.82 1.63 11.76 5.60 11.06 0.06 7.95 0.08 7.23 0.14 6.42 0.98 6.17 0.16 5.17 0.10 5.12 4.74 0.04 173.35 14.68 (8.47%) Food Emergency aid aid 1.43 1.59 0.19 19.05 4.28 6.27 5.24 2.56 0.95 1.68 0.95 0.01 0.08 2.38 0.10 1.19 4.81 2.34 0.95 0.01 2.38 1.57 16.10 43.91 (9.29%) (25.33%) Source: Elaborations on DAC-OECD, 2001 A particularly interesting item, in view of these considerations, is the weight given to emergency and food aid when debt reduction is not brought up. This is not the place for debating the relationship between humanitarian or emergency aid, food aid, and development aid. There are scholars who make a net distinction between them, and those who, on the other hand, say that there must be complete continuity between the various phases and problems which lead from emergency to development. In any case, as far as Italian aid is concerned, it cannot be denied that humanitarian and emergency aid – which make up more than a third of total grants to the main fifteen beneficiary countries – are the most important part of bilateral aid for poor countries. When one then moves on to analyse soft loans, even when the monies used for overseas debt cancellation, and capital and interest repayments on previous loans have been removed (i.e. when 4 One of the two countries, together with Slovenia, which OECD wants to exclude from the list of ODA beneficiary countries as of January 2003, because of the high level of economic development now reached. This means that in future, Italian ODA will feel the negative consequences of perhaps the only stable net paybacks. If calculations are made, for the year 2000, on total bilateral gifts (524.81 million dollars), not including the aid for reducing overseas debt (201.47 million, or 38.39% of the total) and gift made to Malta (20.94 million, or 10.30% of a total of 201.47 million), then the final net figure is barely 302.4 million dollars. 5 Cfr. J. L. Rhi-Sausi (2000), “La via italiana contro le sanzioni: il caso di Cuba”, in Limes, n. 2, 2000, pp. 147-156. 232 considering new credits), the picture that emerges shows important differences between the overall status of bilateral aid (marked by the reduction of external debt) and grants payments. Table 7: The 15 main beneficiaries of bilateral Italian soft loans, 2000 (in millions of US$) Argentina Tunisia China Egypt Syria Albania Honduras Eritrea Bolivia Kenya Lebanon Morocco Algeria Ecuador Jordan 2000 15.24 12.81 7.11 5.22 4.04 3.94 2.60 2.21 1.53 1.27 0.84 0.72 0.69 0.64 - 1999 10.69 6.84 1.72 0.41 9.71 17.42 1.41 2.30 4.68 0.26 0.62 5.04 22.43 5.76 total 25.93 19.65 7.11 6.94 4.45 13.65 20.02 3.62 3.83 5.95 1.1 1.34 5.73 23.07 5.76 Source: Elaborations on DAC-OECD, 2001 First of all, 2000 appears as a continuation of the preceding year, with confirmation of the priority given to the same countries, as per the long-term plans for credit aid. In addition to this, geographical priorities change when one comes to grant channels. Credits continue to be the main tool in foreign policy: China, Egypt, and Tunisia, as is the case with several South American countries, appear among the list of the fifteen beneficiaries of Italian aid, whereas Eritrea and Kenya are the only countries in Sub-Saharan Africa on the list. Facing up to the reduction in Italian credit aid which, therefore, was positive – it must be repeated – to the tune of 100.04 million US$ in 1000 and 148.01 million US$ in 2000, one notes that new credit payments have a different function from grants. This trend began in the late Eighties, when resources were available, and the whole world was interested in overseas development aid, and it continued throughout the Nineties. 4. Distribution by sectors of Italian development aid in 2000 As indicated above, Italian payments for development aid over the last years have noticeably decreased. This is especially the case with resources for bilateral aid, which owes its continuing financial “existence,” in good part, to the 400 billion Lire subsidy paid from the rotating Fund (pursuant to Law N° 266/1999) for granting credits to the cooperation Fund for grants. What is more, these reduced resources can be mostly attributed to forms of external debt rescheduling and cancellation for developing countries. This means that analysis of resource distribution must start at this point of departure. It also means much discretion must be used when using the resources available, a conventional characteristic of Italian overseas development aid, due in part to the fact that long-term budgeting is not binding, and can indeed change from year to year 6 . The scaling down of resources means more discretion is to be used, and underlying trends are harder to identify. Indeed, it is often enough for a country to change any sort of plan from one year to the next for 6 Italy’s new Right-wing government has not moved away from this tradition: the brief career (seven months) of Foreign minister Renato Ruggiero, and the interim taking on of the office by Premier Silvio Berlusconi, are signs of potential discontinuity in strategy in this field. 233 geographical and sector priorities to change. The time lag between the moment of financial planning and that of actual payment of monies implies a structural delay of years in terms of operative application of defined strategies. It may reasonably be said, as we have done here, that the basic choice made by Italy in 1999-2000 was to “marry” the cause of developing countries’ external debt cancellation. There was widespread agreement among Italian public opinion on this point, but it tended to mean a reduction in development aid. At this point, any analysis of real trends in Italian overseas development aid during the last few years becomes extremely relative, as is the case with any evaluation of its new planning direction, even though there are many of these. At a political level, the Nineties saw the fall-off in resources coincide with the re-appropriation by the Foreign Ministry of aid policies, in a fight-back against what Foreign Minister Colombo defined as “outside incrustations.” 7 Development co-ooperation in Italy has become a powerful tool in foreign politics, largely occupying the place once held by foreign economic politics. Indeed, what remains of Italian overseas development aid in South America and Asia (see paragraph above) confirms this view, as does the proximity to Italy of the main aid areas (Balkans and Mediterranean) in terms of security, stability, and the control of immigration flows. The decision to take on the ideal of overseas debt cancellation within the reduced resources of overseas development aid has meant that the ties with foreign policy have been strengthened. Via its commitment to cancelling external debt, Italy has played a major role on the international cooperation stage, both during Holy Year 2000, and when Italy was President of G-8 (2001). The growing use, over the last few years, of Trust funds held by international organisations (multilateral development banks, United Nations agencies) and the European Commission – especially when they are not tied up to funds from other donor countries – help the Italian Foreign Office in its policy of re-appropriation of development co-operation policies. At a sector level, fact checking must be considered together with the planned approach of Italian aid and Italy’s main commitments, last repeated by the then Foreign Minister Renato Ruggiero, when he published the Ministry forecast for ODA for the year 2002. The document laid out the following eight lines: 1. Poverty reduction, the main objective of development cooperation; 2. Help, via the adoption of the Genoa Plan for Africa, for the New African Initiative, which centres on democracy and good governance, conflict prevention, food safety, education and health, international trade, and the promoting of private investments; 3. Help for the private/public partnership initiative for the global AIDS, malaria, and tuberculosis Fund; 4. Education and, in particular, basic primary schooling; 5. The cancellation of developing countries’ external debt; 6. The setting up in developing countries of the right climate for foreign investment; 7. Help for developing countries so they can participate in international trade; 8. Promoting women, children, and the handicapped. These lines re-affirm the principles that Italy has subscribed to at international level over the last years, the core mission of which is the reduction of poverty8 . Re-affirming the centrality of Africa and the emphasis given to education and basic health care are an important corollary to it. In terms of numerical data, however, it is evident that – at least in the year 2000 – the smallness of overall resources and the absolute predominance of activities bound up with external debt reduction 7 Cfr. J. L. Rhi-Sausi, (editor) (1994), La crisi della cooperazione italiana: rapporto del CeSPI sull'aiuto pubblico allo sviluppo, Rome, Edizioni Associate, and J. L. Rhi-Sausi (editor), Ripensare la cooperazione. Rapporto Cespi sull'aiuto pubblico allo sviluppo, Rome, Memoranda. 8 In CIPE’s 1995 framework, poverty reduction was considered as one among many different goals. In 1999, The DGCS addressed these issues by developing specific guidelines on poverty reduction, based on the work undertaken by the DAC Informal Network on Poverty Reduction. 234 have slowed down the implementation of the understandings which Italy had been enunciating since the mid-Nineties. 12% 14% infrastructure and social services infrastructure and economic services productive sectors 1% 6% 4% multi-sector 4% 6% commodity aid and programme aid debt relief 8% emergency aid 12% support to NGOs general expenses not specified 33% Source: Elaborations on DAC-OECD, 2001 Data concerning the division into sectors of bilateral Italian overseas development aid shows that fully a third of all bilateral payments in 2000 went on operations linked to debt reduction. On the other hand, 14% was spent on operations that could not be directly linked to others. If, then, one excludes these two types of expenditure, and considers operations destined to well-defined sectors, debt-reduction expenditure amounts to 38.4%. This figure confirms and accentuates the trend first noticed in 1999, when debt reduction expenditure accounted for 18.5% of bilateral payments. Otherwise, emergency, infrastructure, and social service expenditure are the biggest part of the total, $M 599.48. This figure is also mostly up of grants (524.81 million dollars, as shown in table 5, which means that credits amount to a mere 74.66 million, which is even less – as said above – than what is paid back on credits granted beforehand). Infrastructure and social service payments could, in theory, show up in items which are considered proxies for commitments to poverty reduction, especially those on education and basic health care. However, if one goes on to read the breakdown of these figures, it is easy to see how marginal they are to the day-to-day reality of bilateral cooperation. In the education field, indeed, the trend has been from tertiary (1999) to secondary education (2000), leaving very little over for spending on primary schooling. Exactly the opposite of what a reading of the intentions would lead one to suppose. The same is true in the health field. The importance of the objective of reducing poverty, as the main focus of overseas development aid, nowadays unanimously recognised, and subscribed to by Italy, in fact continues to have a very ambiguous application in the field. 235 Table 8: Bilateral gross ODA payments, per destination sector (in millions of US dollars) 2000 70.18 25.87 11.90 0.03 11.03 2.91 22.59 17.50 5.09 5.71 3.51 3.66 8.84 8.60 5.17 33.12 27.86 3.35 22.54 48.89 31.95 201.47 72.36 35.27 22.23 84.82 599.48 Infrastructure and social services: Education - not specified - basic - secondary - tertiary Health - general - basic Population and reproductive health Water supply and sanitation Government and civil society Other Infrastructure and economic services: Transport and communications Productive sectors: - agriculture - industry Multi-sectors Commodity aid and programme aid: Developmental food aid Debt relief Emergency aid Support to NGOs DGCS general expenses Not specified Total 1999 114.46 31.88 19.70 0.78 1.34 10.06 39.95 27.34 12.61 1.37 9.78 25.38 6.10 24.73 18.26 22.44 13.60 2.94 43.78 78.27 43.78 101.92 102.97 25.69 22.64 13.85 550.76 Source: Elaborations on DAC-OECD, 2001 An intrinsic risk concerns the general nature and dimension of this definition, which can be used as a lifebelt to try to save everything and nothing (projects in the field of finance, economics, the social sphere, and politics). In this sense, what is actually done is particularly important for poverty reduction. Italian ODA, over the last few years, has been concerned with providing a wide spectrum of priority strategies for poverty reduction. The 1999 DGCS guidelines on poverty reduction outline the approach and contents of an Italian poverty reduction initiative to be implemented with an initial allocation of US$ 120 million. The initiative will consist of regional programmes in Central America, South America (Brazil and Andean countries), the Maghreb, the Middle East, the Horn of Africa, Sahel, Southern Africa and India. The initiative aims to provide support to the poverty reduction strategies in two/three partner countries in each of the above regions, within the framework of a consistent regional and international approach. Such programmes will build on the experience gained in previous initiatives, financed by the Italian Co-operation through trust funds to international organisations, mainly the UNDP, and executed by the United Nations Office for Project Services (UNOPS). In operational terms, it is a multi-bilateral initiative subject to direct monitoring by the DGCS. Thus, on a day-to-day basis, this effort should have been translated into a vast programme of multi- and 236 bilateral-type projects, covering seven regions, giving particular importance to the empowerment of women and childcare. What is new about the programme launched in 1999, specifically aimed at poverty reduction, is how it faced up to administrative difficulties, which have blocked it in its tracks until 2001. It is true, though, that in terms of cooperation strategies, the most important element in Italian cooperation will be the effective definition of what is actually meant by “poverty reduction” as the core mission of Italian development aid. This choice has been made, and it can be seen in the homogeneity of Italian aid programmes compared with international commitments and cooperation (the DAC role is particularly important as a structure for stimulating and coordinating the main donor countries, including Italy herself). When one examines Italy in detail, it can be observed that there are some traditional strong points in cooperation practice, and these can be directly linked to poverty reduction policies. In particular, social, health, and children’s programmes are a solid asset in the field of experience in the field. 9 5. The presence of tied aid in Italian development co-operation The mention made just now of Italian ODA reminds the writer of another element of development cooperation, which was hotly debated over the last few years. By this is meant tied aid, which is another way of saying credits or grants which imply the provision of goods or services by companies in the donor country. 10 In other words, with tied aid the tie is made at source, i.e. the beneficiary must use credits and grants in the donor country. At an international level, the practice of “tying” much bilateral aid has spread, as have criticisms of the same. Tied aid implicitly reduces the value of aid payments, because the tie to the donor country reduces the beneficiary country’s freedom in choosing where the money can be spent on the free market. It means an increase of 20-25%, and the principle of freedom of competition is violated (if the donor country has placed a tie on the aid given, perhaps the goods to be acquired are not very competitive). If, hypothetically, many countries speak out in favour of breaking the tie of this type of aid, in practice nobody actually wants to take the first step: this would mean that the donor countries could keep their own ties in place, and take advantage of other countries’ unbonding of aid. This is the reason why, within DAC, there have been such long debates about eliminating tied aid. On 14 May 2001, DAC published the agreement signed by donor countries on untying aid to the least developed countries (the 49 most backward ones, within the category of developing countries). The agreement deals with types of bilateral aid which currently amount to five billion dollars (of the eight billion paid out bilaterally to the least developed countries, or 17% of total bilateral aid). Italy played an important role in the preparatory stages of the negotiations in favour of untied aid, but this goes against the current tendency in Italian jurisprudence. Law N° 49/1987, which governs Italian development co-operation policy, forbids untying of credits. During the last Parliament, a Bill to reform Italian aid policy was not passed, and the implementation of the line adopted is not an easy task. Even the mechanism of Trust Funds with international organisations helps tying in the multi- and bilateral channels. Final data for bilateral cooperation during the year 2000 show that, in terms of actual monies spent, and excluding technical assistance and administrative costs (which are necessarily tied), 38.15% of aid is untied, whereas the remaining 61.85% is tied. To be precise, 41.73% of grants and 28.96% of credits are not tied, while 58.27% of grants and 71.04% of credits are. The percentage of tied aid is high, but lower than it used to be. A proper analysis shows the reasons why. 9 The Foreign Affairs Ministry has recently brought out a book called “Italy for Children’s Rights,” which publishes much information concerning Italy’s aid activities in favour of children. 10 Partially tied aid is the term used when it supply is limited not merely to the donor country, but to several others as well. This may also include beneficiary countries. With untied aid, there is no obligation towards suppliers in the donor country. 237 Table 9: Italian bilateral development aid commitments, exclusive of technical assistance and administrative costs (in millions of dollars) Direct financing of import - Untied aid - Partially tied aid - Tied aid (solo all’Italia) In kind aid (tied aid) Debt relief (untied aid) Grants 178.78 2.92 0 175.86 109.50 201.47 Soft loans 175.98 40.29 0 135.70 0 15.04 Total 354.76 43.21 0 311.56 109.50 216.51 Source: Elaborations on DAC-OECD, 2001 Once again, measures for the reduction of poor countries’ external debt are a significant part of the overall result. The high percentage of bilateral operations for the reduction of external debt – which cannot be classified as tied – account for almost all non-tied aid. Compared with the past, and especially before the late Eighties, when more than 90% of all bilateral aid was tied and Italy was near the top of the list of countries who offered tied aid, what has changed in the intervening period is that the amount of monies paid in aid has decreased, but measures for reducing external debt continue to increase. 2000 ODA commitments (bilateral tied aid percentage, US$ million) Direct financing of import (tied aid) Direct financing of import (untied aid) 1500 in kind aid (tied aid) Debt Relief (untied) 1000 other (untied) 500 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1997 1998 1999 2000 -500 Source: Elaborations on DAC-OECD, 2001 6. Historic evolution of Italian bilateral cooperation aid policy. A comparison with preceding decades In percentage terms, the Eighties started with 78.7% of Italian aid going to sub-Sahara Africa (compared with 2.9% for the Middle East and a deficit in negative terms for European countries). The same decade closed in 1989 with 54.5% going to sub-Sahara Africa (as against 1.4% for the Middle East and 0.5% for European countries). During that time, the importance of Latin America and the Caribbean had increased: from 10.5% in 1980, the figure more than doubled, reaching 22.7% in 1989. A decade, therefore, of greater resources for cooperating in development, and a greater Latin-Americanisation of the same. 238 Geographical distribution of Italian ODA (% of total aid) 90,0 Sub-saharan Africa North Africa 80,0 America Middle East 70,0 Asia 60,0 Europe 50,0 40,0 30,0 20,0 10,0 0,0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 -10,0 Source: Elaborations on DAC-OECD, 2001 During the Nineties, though, two areas of “proximity” grew and were consolidated. The importance of geo-politics and security was evident: the Middle East and the Balkans. More than 95% of Italian aid in 1999 and 2000 was concentrated in Africa, the Middle East, and Europe. Table 10: Relative weight of the three regions who benefited most from Italian Aid (in percent) Africa Middle East Europe Sub-total 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 58.5 32.5 33.9 31.3 22.7 52.0 43.2 66.1 68.5 57.9 64.6 1.9 1.9 2.2 2.8 2.3 3.9 9.2 8.5 4.1 7.3 6.4 4.1 19.8 10.6 11.8 5.1 9.2 21.8 12.4 8.0 33.1 24.6 64.5 54.1 46.8 45.9 30.0 65.1 74.2 87.0 80.5 98.3 95.7 Source: Elaborations on DAC-OECD, 2001 Over the last two years, the Balkans region has jumped up to a third and a quarter of total Italian overseas development aid. On the other hand, when one considers the final balance after repayments on aid credits, it can be seen that Asia (which has shown a stable negative balance since 1997) and Mediterranean Africa (which has also shown a negative balance in 1999 and 2000) have practically disappeared. North African countries are interesting for Italy. Since these countries have average earnings, the mostly receive credits, rather than grants. This means that Italian aid offers African nations around the Mediterranean an essential development tool – credit – which is decisive because it offers access to credit per se (otherwise hard to find in these countries), and access to particularly good repayment terms. For Italy too, credit aid to North African countries seems to work rather well, given that the final balance is negative: debtor countries can be trusted to pay back loans, presumably because such low interest rates have been applied to them. With this region, accounting for development aid would appear to be positive for the donor country as well: an important activity – giving access to credits which are not fully integrated into international finance markets – and at the same time financially sustainable by the Italian Exchequer, because these countries pay back their debts (granted on easy terms). It has also allowed Italy to sign agreements – in particular with Morocco and Egypt – for converting old debts from previous aid into programmes for social and 239 environmental programmes. Given that these countries pay back their debts, Italy uses the repayments for new programmes based on the objective of creating job opportunities, especially for small- and medium-sized firms. This mechanism is one of the most interesting innovations in the field of Italian bilateral aid for external debt when swaps are involved. Indeed, it was extended to Peru and will spread to Ecuador. Negotiations along the same lines are underway with Pakistan, the Philippines, Yemen, and Djibouti. An important tool for promoting the private sector is found in concessional credit for the financing of joint ventures (pursuant to Section 7, Law N° 49/1987), which has been as little applied over the last three years as it was during the whole decade. Overall, in the period between 1990 and June 2000, only 56 initiatives had been decreed under Section 7 of Law N° 49/1987, for a total of 170.9 billion Lire, 28 of which were for Asia, 11 for Latin America, 11 for the Mediterranean basin and Near East, 5 in Eastern Europe, and 1 in Africa. Concerning the focus on the least developed countries, they represented the bulk of Italian aid beneficiaries in the 1980s (around 60% of total aid). Then, in the 1990s, Italian interests and aid core mission changed – in correspondence with the prevalence of proximity, security and economic interests – with a bigger amount of financial resources devoted to aid, and the weight of the poorest countries decreased. In the last five years, Italy renewed its interest on the poorest countries, as a way to operationalize the objective of poverty reduction and, again thanks to the effect induced by debt relief (which is basically targeted to poor countries), recent data can confirm this trend. Tab. 11: Italian ODA to the least developed countries (as % of total aid) 1986 62.9 1987 59.6 1988 55.9 1989 41.6 1990 48.7 1991 26.9 1992 28.3 1993 29.6 Source: Elaborations on DAC-OECD, 2001 240 1994 18.9 1995 37.6 1996 28.3 1997 53.3 1998 69.9 1999 38.4 2000 64.4 7. The HIPC Initiative: its rationale and implementation 1. – Pre-HIPC Debt relief measures involving SSA countries Latin American crisis, exploded in 1982 with the Mexican debt crisis, was firstly regarded as an acute short-term liquidity problem, for which appropriate measures existed within the framework of conventional debt consolidation exercises1. Because of the possibility that widespread defaults could be a serious and immediate threat to the stability and efficient functioning of the international financial system, it was important to provide more time, required from creditors in order to protect themselves, through partial bail-outs, and save the credit system from collapse. In fact, it is always important to take in account that traditionally relief operations are intended to serve and protect the interests of the creditors, not less than those of debtors, in order to ensure that they get paid and that normal market conditions are restored. Thus, the mechanism of debt rescheduling was widely adopted. At the beginning of the 1980s international debt crisis was considered mainly due to liquidity problems. As a solution to the crisis industrialised countries were pumping more money to the developing countries as well as rescheduling the payments on the existing debts of the problem debtors. These exercises provided the countries with short-term relief as rescheduling of repayments and granting of grace periods slightly reduced the net present value of the debt but did not greatly affect the size of the debt burden. When country is suffering form balance of payments difficulties rather than liquidity crisis, it is in need of long-term relief. Then restructuring even on highly concessional terms might not be effective enough but debt-stock reduction could be required. By the end of the decade it was obvious that no sustainable solution would be achieved without further, more extensive relief. It was becoming more evident that in most of the developing countries debt servicing problems were not caused by liquidity problems but by substantial balance of payment disequilibrium. During the nineties the official creditors have provided large debt relief packages and the commercial banks have taken major write downs in developing country debts. On the other hand, the debt problem of SSA countries did not represent a threat to the stability of the world financial system. It was considered the effect of the long-term development debts of the SSA countries, that is an insolvency crisis, which could wait until African economies recovered and developed. Unfortunately development did not occur, and development loans retarded it, by targeting limited foreign exchange resources to debt service payments, whereas the level of total external debts outstanding has continued to rise. Differently from Latin American countries, SSA countries had servicing capacity severely constrained by low growth, poor investment opportunities and low credit-rating. Notwithstanding the existence of two main types of debtors among the developing countries (middle-income countries highly indebted in absolute terms vs. low-income countries highly indebted in relative terms) there was a common response of international system to manage debt crisis2. In contrast to middle-income countries, it became increasingly clear that the mounting debt burdens of low-income countries reflected deeper problems—they were solvency problems 1 For a discussion of the factors leading to high indebtedness in a sample of ten low-income countries, see D. Ross and K. Ross (1998), “External Debt History of ten Low-income Developing Countries: Lessons from Their Experience”, IMF Working Paper, WP/98/72. 2 G. C. Abbott (1993), Debt Relief and Sustainable Development in Sub-Saharan Africa, Edward Elgar, Cambridge. 241 that required not only a temporary reduction in debt service, but also a reduction in the level of debt. Paris Club creditors began to grant such debt reduction in the form of concessional flow reschedulings for low-income countries in late 1988 under the so-called “Toronto Terms” that involved a debt reduction of about one-third of the eligible amounts. The level of debt forgiveness was subsequently raised in steps: to London terms in late 1991 (50 percent debt reduction), and to Naples terms (two-thirds debt reduction) at the end of 1994. This resulted in increasingly longer and lower payment profiles on restructured debt. Bilateral creditors not participating in the Paris Club - mainly oil exporters in the Middle East, but also China, Taiwan and a number of other developing countries, including some heavily indebted poor countries - provided more limited debt restructuring than other creditors, but in turn often saw their claims increasingly falling into arrears3. The 1978 Retroactive terms adjustment scheme At UNCTAD IV, in May 1976, the creditors countries accepted the need of relief, within a multilateral framework, to countries suffering from debt service difficulties. At the end of a long discussion, in 1978, Resolution 165 (S-IX), which formalised the concept of the Retroactive terms adjustment (RTA) and called on donor countries to adopt measures to support developing countries4. This scheme was very flexible, enabling creditor countries to employ a variety of possible measures: ¾ cancellation of all outstanding bilateral ODA loans, ¾ conversion of past loans to grants, ¾ cancellation of interest payments for a given period of time on outstanding bilateral ODA loans, ¾ provision of new funds on highly concessional terms to refinance existing service payments, ¾ adoption of local currency payments and cost financing (the so called “equivalent measures”). Basically, Germany, Britain and the United States committed to this scheme. Out of the 58 beneficiary countries, 34 were SSA countries, receiving almost three-quarters of the total value of relief measures taken (the total nominal value of operations, up to 1986, amounted to $6.2 billion, mainly for ODA debt cancellation). Tanzania topped the list with RTA from 11 different creditors. However, the amount provided by this scheme was too small to reduce debt levels, lacked of uniformity in terms of creditor commitment and was not linked to ODA policies. Nevertheless, RTA was the first mechanism to introduce debt cancellation as part of the strategy of debt relief5. In fact, by 1987, all the G-7 member States, Denmark, the Netherlands, Sweden and Belgium implemented some bilateral debt relief initiatives to cancel the outstanding ODA debt of the low-income SSA countries. Gambia, Kenya, Madagascar, Senegal, Tanzania and Zaire were the main beneficiaries of these initiatives. 3 For a discussion of the motivations underlying traditional debt relief and the evolution of Paris Club terms see C. Daseking and R. Powell (2000), “From Toronto Terms to the HIPC Initiative: A Brief History of Debt Relief for Low-Income Countries”, International Economic PolicyReview, Vol. 2, IMF, pp. 39–58. 4 UNCTAD (1978), Resolution 165 (S-IX). Debt and Development Problems of Developing Countries on the trade and Development Board at its Ninth Special Session”, 494th Meeting, Geneva, 11 March. 5 The 1981 the Berg Report noted that Liberia, Sierra Leone, Sudan, Zaire, and Zambia had already experienced “severe debt-servicing difficulties” in the 1970s and “are likely to continue to do so in the 1980s”. 242 The 1987 Lawson initiative In 1987, the UK Nigel Lawson initiative, recognising that SSA countries could not repay their debts, were dependent on concessional aid, implied a moral and economic responsibility of creditors and needed macroeconomic conditionality to introduce economic reforms, proposed quick actions to reduce interest rates on debts. The 1987 G-7 Venice Summit recommended to apply lower interest rates, longer repayment and grace periods, in the Paris Club, to reduce the debt burden. And the Paris Club adopted these new conditions for SSA countries. The 1988 Toronto terms In 1988, the G-7 nations invited, at the Toronto Summit, the Paris Club to define a menu of options for softening rescheduling terms on non-concessional debt. The Paris Club identified three options: (A) partial write-off, to provide fewer resources during the grace period, but more relief later: cancel 33% of the debt service, and reschedule the remainder at market rates over 14 years with an 8-year grace period; (B) longer repayment terms, to provide relief over an extended period of time, being the least costly to creditors: reschedule debt service payments at market interest rates with a maturity of 25 years with a 14-year grace period; (C) concessional interest rates, to provide more resources up-front, but becoming more expensive later: reschedule debt service payments at reduced rates of interests – one half of market rates. All these options were lower than the conventional terms of Paris Club rescheduling. In order to become eligible to relief, a debtor country was required to conclude a structural adjustment agreement with the international financial institutions, thus receiving the Fund’s structural adjustment facility (SAF) and becoming eligible to the Bank’s Special programme of assistance (SPA). Creditors used all of these three options and some SSA countries had their debts rescheduled twice. The net present value of service payments was reduced by 20 percent, compared to traditional Paris Club terms6: the level of reduction was defined as 33.33%. Concerning ODA credits, they were rescheduled at an interest rate at least as favourable as the original concessional interest rate applying to these loans, with a 25-year repayment period including 14-year grace. This rescheduling usually resulted in a reduction of the net present value of the claims, as the original concessional rate was lower than the appropriate market rate. The reduction in the net present value varied on a case-by-case basis, depending on the original interest rate of the claims of each creditor on each debtor country. Thus, the Toronto terms introduced the idea of reducing the stock of official nonconcessional debts, but it was still inadequate: it covered only debts owed to Paris Club creditors – being 40 percent of non concessional official debts of SSA countries -, rescheduled one year’s maturities at a time, in a short-term approach, and the process of relief was slow7. 20 countries benefited from Toronto terms between 1988 and 1991, when these terms were replaced by London terms. The 1990 Houston and Trinidad terms In 1990, the Houston G-7 summit considered “more concessional rescheduling for the poorest debtor countries”. The UK proposed “Trinidad terms” that would increase the grant element of debt and debt service reduction to 67 percent, from 20 percent under the “Toronto 6 Classic terms are the standard terms applied to a debtor country coming to the Paris Club. 57 countries have benefited from the Classic terms. 7 In 1993, after the G-7 Tokyo summit, the Paris Club applied Enhanced Toronto Terms that were even more concessional. 243 terms”8. These new terms implied the increase of cancellation to two-thirds of the total stock of eligible debt, a longer repayment period lengthened to 25 years rather than 14 years, capitalisation of interest payment occurring in the first five years. The Washington Consensus’ conditionality, represented by the structural adjustment programmes, was confirmed. The 1991 London terms In 1991, Paris Club creditors agreed to implement a new treatment on the debt of the poorest countries. This new treatment, the “London terms”, raised the level of debt cancellation from the 33.33% defined in Toronto terms to 50%. Non-ODA credits were cancelled to a 50% level. Creditors chose to implement the terms through one of the four following options: - "debt reduction option" ("DR"): 50% of the claims treated were cancelled, the outstanding part being rescheduled with 23 years repayment period including 6-year grace and progressive payments. - "debt service reduction option" ("DSR"): the claims treated were rescheduled at a reduced interest rate with 23 years repayment period with progressive payments. - "moratorium interest capitalisation option" ("MIC"): the claims treated were rescheduled at a reduced interest rate with 23-year repayment period including 6-year grace and progressive payments. - "commercial option": the claims treated were restructured at the appropriate market rate over a longer period (25-year repayment period including 14-year grace). This was a nonconcessional option. Concerning ODA credits, they were rescheduled at an interest rate at least as favourable as the original concessional interest rate applying to these loans, with a 30-year repayment period including 12-year grace and progressive repayment. This rescheduling usually resulted in a reduction of the net present value of the claims, as the original concessional rate was lower than the appropriate market rate. The reduction in the net present value varied on a case-by-case basis, depending on the original interest rate of the claims of each creditor on each debtor country. 23 countries benefited from London terms between 1991 and 1994, when these terms were replaced by Naples terms. The 1994 Naples terms In 1994, after the G-7 Naples summit, the Paris Club creditors agreed to implement a new treatment on the debt of the poorest countries. These new terms, the “Naples terms”, granted two substantial enhancements with respect to London terms, that can be implemented on a case by case basis, on the level of reduction and the conditions of treatment of the debt: - for the poorest and most indebted countries, the level of cancellation is at least 50% and can be raised to 67% in net present value terms of total bilateral debt of eligible non-ODA credits9; 8 W. Easterly (1999), How did highly indebted poor countries become highly indebted? Reviewing two decades of debt relief, World Bank, Washington D.C., September. 9 Creditors agreed in September 1999 that all Naples terms treatments would carry a 67% debt reduction. 244 - stock treatments may be implemented, on a case-by-case basis, for countries having established a satisfactory track record with both the Paris Club and IMF and for which there is sufficient confidence in their ability to respect the debt agreement. Creditors were invited to chose to implement the 67% debt reduction of non-ODA credits by one of the two following options: - "debt reduction option" ("DR"): 67% of the claims treated are cancelled, the outstanding part being rescheduled at the rate of 23 years repayment period with a 6-year grace and progressive payments); - "debt service reduction option" ("DSR"): the claims treated are rescheduled at a reduced interest rate according to 33 years repayment period with progressive payments. Two other options were also designed, but have been very seldom used: - the "Capitalisation of moratorium interest" ("CMI") option, similar to the "DSR" option (with a reduction of 67% in net present value) but with slightly different repayment profiles; - the "commercial option", with longer repayment profiles but no reduction of the claims in net present value. It was agreed that creditors would refrain from using this option to very exceptional circumstances. Concerning ODA credits, they were rescheduled at an interest rate at least as favourable as the original concessional interest rate applying to these loans, according to 40 years with 16year grace and progressive repayment. This rescheduling results in a reduction of the net present value of the claims, as the original concessional rate is smaller than the appropriate market rate. The reduction in the net present value varied from one country to another, depending on the original interest rate of the claims. Naples terms also included the possibility for creditor countries to conduct, on a bilateral and voluntary basis, debt swaps with the debtor country. 31 countries have benefited from Naples terms. What however has been the common characteristics of all these mechanisms, presented as a needed and effective action to redress the SSA countries’ imbalances, is that neither the commercial banks have participated, for fear of setting a dangerous precedent for countries in which their exposure is greater, nor the multilateral banks in the cancellation schemes. 2. The Initiative for Heavily Indebted Poor Countries (HIPC) Due to rising evidence that the development prospects of many of the SSA countries suffered from unsustainable debt, the member governments of the International Monetary Fund (IMF) and the World Bank agreed in fall 1996 the initiative for “Heavily Indebted Poor Countries” (HIPC). At that time, SSA countries’ external debt amounted to US$ 1,003.2 billion, gross national product to US$ 168.4 billion and the debt-to-GNP ratio was 595.7%. The HIPC Initiative was launched and, for the first time, it was based on specific contributions of the multilateral institutions so that they were committed to reduce the level of their claims on the concerned countries. 245 Tab. 1 - SSA countries classified as IDA-eligible, HIPC, with low HDI, and Lldcs10 IDA-eligible SSA Angola Benin Burkina Faso Burundi Cameroon Cape Verde Central African Rep. Chad Comoros Congo, Republic of Congo, Democr. Rep. Cote D'Ivoire Eritrea Ethiopia Gambia Ghana Guinea Guinea-Bissau Kenya Lesotho Liberia Madagascar Malawi Mali Mauritania Mozambique Niger Nigeria * Rwanda Sao Tome & Príncipe Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Zambia Zimbabwe * HIPC in SSA11 Angola Benin Burkina Faso Burundi Cameroon SSA with low HDI Angola Benin Burkina Faso Burundi The Lldcs in SSA Angola Benin Burkina Faso Burundi Central African Rep. Chad Central African Rep. Chad Cape Verde Central African Rep. Chad Comoros Congo, Republic of Congo, Democr. Rep. Côte d'Ivoire Equatorial Guinea** Congo, Democr. Rep. Côte d'Ivoire Congo, Democr. Rep. Ethiopia Gambia*** Ghana Guinea Guinea-Bissau Kenya Eritrea Ethiopia Gambia Equatorial Guinea Eritrea Ethiopia Gambia Guinea Guinea Bissau Guinea Guinea Bissau Liberia Madagascar Malawi*** Liberia Madagascar Malawi Mali Mauritania Mozambique Niger Nigeria ** Rwanda São Tomé & Príncipe Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Zambia Mali Mauritania Mozambique Niger Nigeria Rwanda Sao Tome & Principe Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Zambia Lesotho Liberia Madagascar Malawi Maldives Mali Mauritania Mozambique Niger Rwanda Sao Tome & Principe Senegal Sierra Leone Somalia Sudan Tanzania Togo Uganda Zambia * = Blend Countries (eligible to both IDA and World Bank loans) ** = Included at the beginning in the HIPC list, later excluded *** = Excluded at the beginning in the HIPC list, later included After the G-7 Lyon summit, in November 1996, the Paris Club creditor countries accepted to raise up to 80 percent the level of bilateral debt reduction in net present value terms for the poorest countries with the highest indebtedness, going beyond Naples Terms. Lyon terms also confirmed the possibility for creditor countries to conduct, on a bilateral and voluntary basis, debt swaps with the debtor country. This initiative is the current one, enhanced by the decision, in November 1999, of the Paris Club creditor countries, after of the 10 Botswana, Gabon, Mauritius, Mayotte, Namibia, Seychelles, South Africa, and Swaziland are excluded by all these four categories. 11 Other HIPCs: Bolivia, Guyana, Honduras, Laos PDR, Myanmar, Nicaragua, Republic of Yemen, Vietnam. 246 G-7 Cologne Summit, to raise the level of cancellation for the poorest countries up to 90% or more if necessary in the framework of the HIPC initiative. As this is the most advanced debt relief initiative involving SSA countries and it has been under implementation for six years12, we will analyse it more deeply. The HIPC Initiative was endorsed by some 180 governments around the world as an effective and welcome approach to help poor, severely indebted countries and reduce debt as part of an overall poverty reduction strategy13. Declared objective of HIPC was to reduce the external debt of the world's poorest, most heavily indebted countries (33, that is three fourth of them, are in SSA) to “sustainable” levels. At the beginning, 41 countries were considered as eligible to the HIPC Initiative. This included, for analytical purposes, 32 countries with a 1993 gross national product per capita of $695 or less and 1993 present value of debt to exports higher than 220 percent or present value of debt to gross national product higher than 80 percent. Also included were nine countries that received, or were eligible for, concessional debt rescheduling from bilateral creditors. In 1998, Nigeria no longer met the criteria, and Malawi was added. In 1999, the number of countries was reduced to 40 because Equatorial Guinea no longer met the criteria for “low income” or “heavily indebted”, as – with the onset of oil production - GDP levels rose above those required for IDA-only assistance14. In the summer of 2000, the Gambia was added as well, and the number of HIPC countries returned to 41. The debt of the 40 HIPCs represents less than 10 percent of total developing country debt. While this amount is but a small fraction of the total debt of developing countries of more than $2.5 trillion, the debts of HIPCs are, on average, more than four times their annual export earnings, and well exceed their annual GNPs. These are about twice the levels considered to be sustainable. Thus, thirty-three African countries are recognised to have debt that is unsustainable. HIPCs have already been receiving substantial debt relief through traditional channels such as Paris Club debt relief operations, forgiveness of aid debts, and commercial bank debt buybacks – reviewed in chapter 4 - at discounts in the order of 85 percent. These operations have provided debt relief of about $25 billion during 1990-96. Even though the debt of the HIPCs is on more concessional terms than that of other developing countries, the debt burden of this group of countries is far more severe, in relation to their capacity to meet debt obligations. By 1996, the ratio of the net present value (NPV) of debt of the HIPCs to exports had fallen by 23 percent, from about 600 percent in 1991 to 450 percent. It largely reflected increasingly concessional debt relief provided by bilateral and commercial creditors, but remained more than twice as high as those of all developing 12 In the context of a Paris Club concessional agreement, some debts that were previously reduced may be further reduced under a concessional treatment with an increased level of cancellation. In this case, there is a topping-up from the previous concessional treatment to the new one. If the initial amount due by a debtor was $10 billion, and a creditor has cancelled in 1993 50% of the debt (London terms), then remaining debt is $5 billion. With a new 1998 treatment, involving the cancellation of part of the debt with a 80% debt reduction level (under Lyon terms), the remaining debt is $2 billion, that is the difference between the initial amount and 80% cancellation. For creditors, who have already cancelled 50% of the debt in 1993, it means that they have to cancel 60% of the non-ODA treated debt. Thus, in a strict sense, the net contribution of a given debt relief initiative, as the HIPC is, should be calculated considering the difference of final relief compared to the outcomes of the previous mechanisms. 13 "A Program for Action to Resolve the Debt Problems of the Heavily Indebted Poor Countries -- Report of the Managing Director of the IMF and President of the World Bank to the Interim and Development Committees,” September 20, 1996, page 2 (EBS/96/152, Revision 1 and SecM96-975/1). 14 But the Equatorial Guinea's amount of arrears equally suggests that this country is unable to carry its debt burden. 247 countries as a group. Within the HIPC group, individual country indebtedness ratios varied widely, ranging from below 200 percent to above 1,000 percent. The actual debt-service payments of the HIPCs, expressed in percent of exports, are approximately in line with those of all developing countries. However, in contrast to the situation in other developing countries, the debt service paid by the HIPCs is only around two-thirds of debt service due in 1995-97. Arrears, debt forgiveness and restructuring make up the difference. High levels of gross new inflows of official resources have also contributed to the ability of HIPCs to make debt-service payments and finance development programs. The ratio of gross inflows (from long-term debt and grants) to debt service paid averaged about two to one for the HIPCs as a group during the 1990’s and ranged upward of four to one in half of these countries. Annual net transfers to the HIPCs on medium- and long-term resource flows (including grants) have averaged about 10 percent of GNP over the 1990-96 period. As we mentioned, the HIPC Initiative is particularly important, as it was the first comprehensive effort by the international community to reduce the external debt of the world's poorest countries. It went beyond earlier debt-relief initiatives in that it included debt from multilateral creditors like the IMF and the World Bank. The implementation of the HIPC Initiative has been reviewed regularly by Executive Directors, and progress reports have been sent to the Interim and Development Committees on the occasion of their regular semi-annual meetings, in April and September, since 1997. When the Interim and Development Committees endorsed the Initiative, they agreed that it would remain open for two years to HIPCs that pursue or adopt programs of adjustment and reform supported by the IMF and IDA, after whom a decision would be made whether it should be continued. This two-year “sunset clause” on entry reflected the intention that the HIPC Initiative would not have been a permanent facility. The sunset clause gave countries an incentive to adopt IDA- and IMF-supported adjustment programs. It also limited the time available for build-up of new debt, and thus provided for relief on debt which mostly predated the Initiative. The Executive Directors of the IDA and Fund agreed to an extension of the initial deadline (sunset clause) for meeting the entry requirement until end-2000. They also agreed that a comprehensive review of the Initiative would be undertaken as early as 1999. The Initiative provides debt relief to countries that implement critical social and economic reforms as part of an integrated approach to lasting development. Debt sustainability targets A country may be considered to achieve external debt sustainability when it is able to meet its current and future external debt-service obligations in full, without recourse to debt relief, rescheduling, or the accumulation of arrears. Sustainable debt levels under the Initiative are defined on a case-by-case basis as: the net present value15 of public and publicly guaranteed external debt in percent of exports within the range of 200–250 percent; the ratio of external debt service to exports within the range of 20–25 percent. An NPV of debt-to-export target of 200 percent is assumed in the baseline scenario for all countries, excepting countries which: 15 The net present value of debt is a measure that takes into account the degree of concessionality, in contrast to the face value of the external debt stock, which is not a good measure of a country’s debt burden if a significant part of the debt is contracted on concessional terms; for example, with an interest rate below the prevailing market rate. The net present value is defined as the sum of all future debt-service obligations (interest and principal) on existing debt, discounted at the market interest rate. 248 (i) have already reached the decision point, (ii) have had preliminary HIPC discussions at the Boards; or (iii) would be projected to qualify under the fiscal/openness criteria. These criteria were agreed in April 1997, for countries where a large export base could exaggerate the capacity to service external debt, in relation to other measures, such as fiscal revenues. For such countries, being very open economies with a heavy fiscal debt burden, which are also making a strong effort to generate fiscal revenues, that is cases with a revenues-to-GDP ratio above 20 percent and an exports-to-GDP ratio above 40 percent, the sustainability threshold under the original framework was the objective of reducing the NPV of debt-to-fiscal revenue ratio to 280 percent at the completion point, other than an NPV debt-to-export target below 200 percent16. These target ranges were based on the research findings17. Countries eligibility The Initiative is open to the poorest countries, those that: (i) are eligible only for highly concessional assistance such as from the World Bank's International Development Association (IDA) and the IMF's Poverty Reduction and Growth Facility ( formerly called Enhanced Structural Adjustment Facility); (ii) face an unsustainable debt situation even after the full application of traditional debt relief mechanisms; and (iii) have a proven track record in implementing strategies focused on building the foundation for sustainable economic growth. The entry requirement The first milepost toward establishing the policy track record necessary to qualify for HIPC assistance is the entry requirement. The Program of Action states that “the Initiative would be open to all HIPCs that pursue or adopt programs of adjustment and reform supported by the IMF and IDA in the next two years, after which the Initiative would be reviewed and a decision made whether it should be continued”. Staffs have interpreted this to include any country which had an ESAF arrangement approved or midterm review completed starting one year prior to the inception of the Initiative, together with an ongoing Policy Framework Paper (PFP) and/or adjustment operations supported by IDA, as having met the entry requirement. The decision point and amendments Eligible countries reaching their decision points have generally completed three-year track records of adjustment and reform. Based on the assumption of satisfactory performance under IMF- and IDA-supported programs of adjustment and reform, a total of 26 countries (twothirds of all HIPCs) could have reached their decision points by the end of 2000. Post-conflict countries represent a special challenge. Their needs are great, opportunities for progress are substantial, but institutional and administrative capacity is often severely limited. 16 IMF and World Bank (1997), “HIPC Debt Initiative: Guidelines for Implementation”, IDA/R97-35 of 4/22/1997 and EBS/97/75 of 4/21/1997. 17 S. Claessens, E. Detragiache, R. Kanbur and Peter Wickham (1997), “Analytical Aspects of the Debt Problems of Heavily Indebted Poor Countries”, Z. Iqbal and R. Kanbur (eds.), External Finance for LowIncome Countries, Washington, D.C., 1997. 249 Many post-conflict countries have a substantial debt burden, and might ultimately be eligible for HIPC assistance. A joint Bank/Fund issues note on providing additional assistance to postconflict countries was prepared for consideration by the Interim and Development Committees. In recognition of the exceptional needs of these countries, it was proposed to provide an additional element of flexibility in the evaluation of the first three-year track record period leading up to the decision point for post-conflict countries. Specifically, satisfactory performance under economic recovery and emergency assistance programs supported by the Bank and the IMF could be counted, on a case-by-case basis, toward reaching the decision point. Transitory measures Several forms of HIPC assistance are available during the transitory period between the decision and completion points. Bilateral and commercial creditors are in general expected to provide flow rescheduling on eligible debt service involving an NPV reduction of up to 80 percent during the second stage. The Initiative also envisages that multilateral institutions could, at their discretion, provide some of their assistance during the interim period. In this context, IDA is providing grants instead of loans to eligible countries. In addition, under certain conditions, supplemental IDA allocations could be made during the interim period18. The IMF may provide some additional ESAF access in cases where this was justified by a strong program and a balance-of-payments need. The completion point Countries reach completion points after establishing a further track record of adjustment and reform, including appropriate social development policies, normally for three years following the decision point. If a country experiences delays in meeting the performance requirements during the interim period, this may lead to delays in reaching the completion point, as is envisaged in the HIPC framework. The IDA and IMF Boards would decide this on a case-bycase basis, and the completion point would not be reached for a given country unless both Boards agreed. The amount of assistance committed at the decision point would be presumed to remain the same as long as the projected NPV of debt-to-export ratio remained within the plus/minus 10-percentage point target range. The ratio at the new completion point would be based on the latest annual debt stock and export data available. Significant delays in performance could require a country to return to the beginning of the second stage and staffs would seek a reassessment from the Boards of the appropriate timing of the completion point. Delivery of HIPC assistance By commitments, 54 percent of the approved debt relief under the HIPC Initiative is covered by multilateral creditors, with bilateral creditors providing 46 percent (on top of the debt relief they have provided under earlier efforts). The World Bank and IMF account for the largest shares of total costs among multilateral creditors, at 25 and 9 percent, respectively. In providing the debt relief, both the World Bank and IMF are committed to pay for their full share of the cost under the Initiative. To this end, the World Bank transferred $850 million to the HIPC Trust Fund, the Bank’s principal vehicle to deliver the debt relief. The IMF has 18 The share of a country’s program support to come from IDA grants (rather than loans) is determined on a sliding scale according to the projected NPV of debt-to-export ratio at the completion point. Grants would account for up to one-third, one-half, or three-fourths of the IDA lending program, depending on whether the projected debt-to-export ratio at the completion point (on a present value basis) was between 250 and 300 percent, 300 and 350 percent, or over 350 percent. See “World Bank Participation in the Heavily Indebted Poor Countries Debt Initiative” (IDA/SecM96-926 of 8/26/1996). 250 provided for $520 million so far to finance its share of the debt relief. However, not all multilateral institutions will be able to finance out of their own resources the needed HIPC debt relief. In support of these institutions’ own efforts, the HIPC Trust Fund immediately obtained about $450 million in bilateral contributions and pledges from 19 countries. These contributions aim also at demonstrating the continued strong support this Initiative enjoys in the donor community. Both the World Bank and the IMF have allocated sufficient funds to cover the costs of the assistance packages agreed to date. Both have stated their commitment to meet their full share of the cost of the Initiative as eligible countries advance in the HIPC process. Ensuring full financing of the Initiative by all participants remains the main challenge. The Bank’s participation in the HIPC Initiative takes place primarily through the HIPC Trust Fund, which provides relief on debt to IDA, either through the purchase and subsequent cancellation of outstanding IDA credits or through servicing of a portion of the beneficiary country’s IDA debt. The Bank transferred US$750 million from IBRD net income and surplus to the HIPC Trust Fund, thereby front loading its contribution ahead of actual cash needs, which arise only when countries reach their completion points. In July 1998, the Executive Directors recommended to the IBRD Board of Governors another transfer of US$100 million from fiscal year 98 net income, which will be considered in early October, during the Annual Meetings. On a commitment basis, the Trust Fund has earmarked US$500 million for the six countries that have reached their decision point to date, leaving an uncommitted balance of about US$300 million (including accrued investment income). The remainder of the relief committed by the Bank to this group of countries (US$205 million in NPV terms) is being provided by replacing IDA credits with grants, which will amount to about US$660 million in nominal terms over the period of fiscal years 1998-2001. The Bank intends to meet all the costs of its participation in the HIPC Initiative from its own resources. IMF participation takes the form of special ESAF grants at the completion point that are deposited into an escrow account to cover debt-service payments to the IMF under an agreed schedule19. In addition, in March and April 1998, the IMF Board decided that no reimbursement would be made to the General Resources Account (GRA) from the ESAF Trust Reserve Account for the cost of administering the ESAF Trust during the fiscal years 1998 and 1999. It also decided that an equivalent amount will be transferred from the ESAF Trust Reserve Account to the ESAF-HIPC Trust. In May 1998, the IMF transferred SDR 41 million to the ESAF-HIPC Trust for fiscal year 1998 and expects to make a similar payment on a quarterly basis to the ESAF-HIPC Trust for fiscal year 1999. The IMF Board has also authorised the transfer of up to an additional SDR 250 million from the ESAF Trust Reserve Account to meet the IMF’s commitments under the Initiative20. IMF staff expected that projected resources are sufficient to meet the Fund’s commitments under the HIPC Initiative through 1999. Other Multilateral development banks (MDB) participate in the HIPC Initiative, too. Most MDBs have obtained the appropriate institutional approvals to enable them to participate in the HIPC Initiative, and have defined the modalities through which they intend to deliver their assistance. These vehicles include: 19 To finance these grants, several countries have contributed or made investments for the benefit of the ESAFHIPC Trust totaling SDR 35 million as of end-June 1998. 20 IMF commitments for the six countries that have already reached their decision points amount to US$270 million (SDR 200 million). 251 (i) channelling resources through the HIPC Trust Fund, either for debt-service reduction or debt buybacks; (ii) using similar, self-administered, trust funds; (iii) rescheduling current maturities or arrears on concessional terms tailored to provide the agreed NPV debt relief; and (iv) refinancing on grant terms. Several MDBs face constraints in covering the full cost of their participation in the HIPC Initiative from their own resources. Some have opted to use the Bank’s HIPC Trust Fund mechanism to help deliver their share of HIPC relief to individual countries, which enables them to receive additional contributions, beyond their own resources if necessary, from interested donors. Thus far, sixteen bilateral donors have made contributions or pledges to the HIPC Trust Fund to assist MDBs, amounting to about US$210 million. For those countries, which have already received commitments of assistance, the division of costs by creditor group reflects the amounts shown in each decision point document. For prospective cases, burden sharing is assumed to be fully proportional. The costing analysis of the HIPC Initiative must be based on the country-specific debt sustainability analyses presented to the Boards, supplemented in some cases by additional information prepared by Bank and IMF staff. A number of important caveats apply. The cost estimates rely on important assumptions, and on debt projections which have mostly not been fully reconciled between creditors and debtor governments. Therefore, the estimates need to be interpreted with caution and should be seen as subject to a substantial margin of uncertainty. In making these estimates, staffs have aimed to provide realistic but conservative estimates of costs; thus, in cases where a choice of targets or timing was required, the option implying a higher cost was used. HIPC costs are divided between multilateral and bilateral creditor groups based on decision point agreements reached and, for future decision points based on proportional burden sharing. On this basis, baseline costs for multilateral creditors would be US$4.2 billion in 1996 NPV terms, or 51 percent of the total US$8.2 billion in costs. Of this, the World Bank and IMF’s share would be, respectively, US$1.7 billion and US$0.7 billion. Bilateral and commercial creditors would meet just under half of total costs, or US$4.0 billion. In comparison with 1997 estimates, the cost to bilateral creditors has risen in 1998 estimates, while that of multilateral creditors is unchanged. This fluctuation in cost share is due to the different mix of country-specific costs21. Indeed, costs excluding countries, which have not yet entered the Initiative, are lower in total and for each creditor group. Cost estimates for the group of mostly post-conflict countries are likely to remain volatile. 21 Most of the increase for bilateral creditors reflects the increase for the Democratic Republic of Congo 252 THE HEAVILY INDEBTED POOR COUNTRIES (HIPC) INITIATIVE Summary First Stage Paris Club provides flow rescheduling as per current Naples terms, i.e. rescheduling of debt service on eligible debt falling due during the three-year consolidation period (up to 67 percent reduction on eligible maturities on a net present value basis). Other bilateral and commercial creditors provide at least comparable treatment. Multilateral institutions continue to provide adjustment support in the framework of a World Bank/IMF-supported adjustment program. Country establishes first three-year track record of good performance. Decision Point Exit · Either... Paris Club stock-ofDebt operation under Naples terms (up to 67 percent present value reduction of eligible debt) and comparable treatment by other bilateral and commercial creditors is adequate for the country to reach sustainability by the completion point-country not eligible for HIPC Debt Initiative. Borderline Eligible Or... Paris Club stock-of-debt operation (on Naples terms) not sufficient for the country's overall debt to become sustainable by the completion point-country requests additional support under the HIPC D e b t I n i ti a t i v e , a n d Wo r l d Ba nk/IMF Boar ds de te rmine eligibility. Second Stage Paris Club goes beyond Naples terms to provide more concessional debt reduction of up to 80 percent in present value terms. Other bilateral and commercial creditors provide at least comparable treatment. Donors and multilateral institutions provide enhanced support through interim measures. Or... for borderline cases, where there is doubt about whether sustainability would be achieved by the completion point under a Naples terms stock-of-debt operation, the country would receive further flow reschedulings under Naples terms. If the outcome at the completion point is better than or as projected, the country would receive a stockof-debt operation on Naples terms from Paris Club creditors and comparable treatment from other bilateral and commercial creditors. If the outcome at the completion point is worse than projected, the country could receive additional support under the HIPC Debt Initiative, so as to achieve exit from unsustainable debt. Country establishes a second track record of good performance under Bank/IMF-supported programs. Completion Point Paris Club provides deeper stock-of-debt reduction of up to 80 percent in present value terms on eligible debt so as to achieve an exit from unsustainable debt. Other bilateral and commercial creditors provide at least comparable treatment on stock-of-debt. Multilateral institutions take such additional measures, as may be needed, for the country's debt to be reduced to a sustainable level, each choosing from a menu of options, and ensuring broad and equitable participation by all 253 3. The implementation of the HIPC Initiative22 Under the original HIPC Initiative approved in 1996, seven countries (Bolivia, Burkina Faso, Côte d’Ivoire, Guyana, Mali, Mozambique, and Uganda) reached their decision points and five additional countries had their debt situations reviewed. Total assistance committed to these seven countries under the original framework was US$3.5 billion in net present value (NPV) terms. During the first two years of implementation, ten countries were reviewed for eligibility for assistance under the agreed framework of the HIPC Initiative. Eight of these were found to face unsustainable debt levels, after full use of traditional debt-relief mechanisms, and hence to require HIPC assistance. Debt relief was committed for six countries (Bolivia, Burkina Faso, Côte d’Ivoire, Guyana, Mozambique and Uganda) and a decision was pending for Mali, which was being proposed for assistance. The Boards also discussed a preliminary HIPC document for Guinea-Bissau; the decision point for that country was delayed by the outbreak of civil conflict. In two other countries (Benin and Senegal) traditional debt-relief mechanisms were found to be sufficient to enable them to achieve sustainable debt situations. The first country approved for HIPC debt relief (Uganda) reached its completion point in April 1998, and creditors started to deliver the promised assistance. Two other countries (Bolivia and Guyana) were close to their completion points. Some of the first ten countries reviewed for eligibility for assistance (Bolivia, Guyana, Mozambique and Uganda) had already demonstrated strong performance under Bank- and IMF-supported programs for long periods. A number of countries in the initial group (Benin, Bolivia, Burkina Faso, Guyana, Mali and Uganda) had also concluded stock-of-debt operations on Naples terms with Paris Club creditors in 1995-96. Côte d’Ivoire, GuineaBissau and Senegal were completing programs supported by three-year ESAF arrangements from the IMF and IDA-supported adjustment programs were broadly on track. All these countries were, therefore, ready to reach their decision points, subject to preliminary debt data reconciliation, completion of their tripartite debt sustainability analyses (DSAs), and agreement on appropriate policies with IDA and the IMF. On this basis, five countries were able to reach their decision points in 1997, and three more in the first half of 1998. In line with the need of more flexibility provided in the agreed framework of the HIPC Initiative, five of the six countries reaching their decision points which required assistance under the Initiative were granted a shortening of the second stage to the completion point. The shortening, in the case of countries with long-time and strong reform programs (Bolivia, Guyana, Mozambique and Uganda), was to around one year, subject to continued strong performance. Concerning debt sustainability targets, the HIPC framework provided for consideration of country-specific vulnerability factors, including the concentration and variability of exports and fiscal indicators of the burden of debt service. The targets for NPV of debt-to-export ratios were fixed at or near the bottom of the 200 to 250 percent debt sustainability range for Burkina Faso (205 percent), Mozambique (200 percent) and Uganda (202 percent). For Bolivia, a 225 percent ratio was agreed, since Bolivia is one of the least vulnerable of the HIPCs under consideration. In the cases of Côte d’Ivoire and Guyana, which qualified for HIPC assistance on the basis of the fiscal/openness criteria, the application of the guidelines resulted in NPV of debt-to-export targets of 141 and 107 percent, respectively. The maximum debt-service targets of 20-25 percent of exports considered in the framework of the Initiative 22 IMF and World Bank Development Committee (1998), “The Initiative For Heavily Indebted Poor Countries: Review and Outlook”, DC/98-15, Washington D.C. September 22. 254 have generally not been binding. In five of the six qualifying countries, the debt-service ratio was projected to be below 20 percent at the completion point even without HIPC relief. In the case of Bolivia, however, the debt service ratio (before HIPC debt relief) was expected to remain above the 20-25 percent sustainability range agreed under the framework for several years after the completion point. Creditors were therefore urged to front load their assistance under the Initiative— as is being done by the Bank and the IMF— and thus reduce debtservice obligations to as close to 20 percent of exports as possible. The HIPC assistance committed to the six qualifying countries that reached their decision points by July 1998 amounted to about US$3 billion in NPV terms, and US$5.6 billion in estimated nominal debt-service relief over time. In NPV terms, bilateral creditors as a group agreed to provide about 46 percent of the total and multilateral institutions 54 percent. Based on their respective shares of the debt, the World Bank provided about 45 percent of the total multilateral debt relief agreed so far, the IMF 17 percent, the IDB 13 percent and the AfDB 12 percent. Creditors were committed to deliver the agreed debt relief at the countries’ individual completion points. The Bank provides some relief earlier by substituting IDA grants for IDA credits during the interim period for eligible countries. The Paris Club has also provided interim debt relief for countries, which did not have existing stock-of-debt operations— namely Côte d’Ivoire and Mozambique— by increasing the NPV reduction under flow rescheduling from 67 to 80 percent. For the six qualifying HIPCs that reached their decision points so far, the reduction in NPV of debt from HIPC assistance at the completion point was projected to range from 6 percent in Côte d’Ivoire to 57 percent in Mozambique. The HIPC assistance package for Mozambique was by far the largest of those approved; to ensure debt sustainability at the completion point, Mozambique wais expected to receive assistance of over US$1.4 billion in NPV terms, or US$2.9 billion in nominal debtservice relief, equivalent to 70 percent of GDP. The debt burden of the eight countries that reached their decision points in the first two years of the HIPC Initiative (including the two HIPCs deemed sustainable) was expected to drop substantially over the medium term. The NPV of debt of these countries was projected to decline by a third between 1996 and 1998 and an additional 7 percent by the year 2000. This trend reflects both debt relief under traditional mechanisms and the projected impact of HIPC assistance. Co-ordination with Paris Club creditors in implementing the HIPC Initiative was been close. IMF staff took the lead in consultations with bilateral creditors, working with Paris Club creditors both in assisting the reconciliation of debt figures and in seeking creditors views’ on the DSA assessments and their commitment to providing the assistance envisaged under the HIPC Initiative. This was a particularly lengthy process in the case of Mozambique, as the bilateral effort required for fully proportional burden sharing exceeded Lyon terms (80 percent NPV relief on eligible medium- and long-term debt). After prolonged discussions, it was agreed that the difference between the two approaches would be covered by a combination of additional exceptional efforts from bilateral creditors and donors, and the Bank and the IMF, but that this should not be a precedent for future cases. Uganda became the first country to reach its completion point under the HIPC Initiative in April 1998. An updated DSA found the NPV of debt-to-export ratio at the completion point— at 196 percent after the committed debt relief— lower than the target of 202 percent, but well within the plus/minus 10 percentage point margin envisaged under the HIPC framework. Paris Club creditors of Uganda agreed in April 1998 on a stock-of-debt reduction of 80 percent under Lyon terms. As a result of HIPC assistance, Uganda’s debt service will be reduced by about 20 percent, compared to that after traditional debt-relief mechanisms, or US$30 million annually over the next ten years, and by about US$22 million, or 10 percent, for the following decade. Uganda is expected to use the savings to support the 255 implementation of a number of social programs, including, in particular, the Universal Primary Education Plan and the Poverty Eradication Action Plan. Another case cited by the IMF and the World Bank to demonstrate the success of impact of the HIPC initiative in action was, in particular, the case of Mozambique. Debt relief under the HIPC Initiative will amount to $4.3 billion. The Initiative will cut the debt by 72 percent (in present value terms). As a result, Mozambique's annual debt service payments will be reduced by up to two-thirds over the coming decade, to less than four percent of yearly export earnings and seven percent of government revenue beginning in 2002. 4. The Enhamced Initiative for Heavily Indebted Poor Countries (HIPC-2) Given the limited results after more than two years it has been in place and the consensus on the fact that more could and needed to be done, in late January 1999, the German Chancellor Schröder announced the Cologne Debt Initiative. At the G-7 Summit in Cologne, world leaders committed themselves to provide a major impetus to the HIPC Initiative. In 1999 the Boards of the World Bank and IMF have called for a comprehensive review of the HIPC Initiative, including updated cost estimates. 256 The enhanced HIPC Initiative adopted by the Boards of the Bank and the IMF in the fall of 1999 aimed at accelerating the delivery of HIPC Initiative assistance and linking debt relief more firmly and transparently to poverty reduction23. At the same time, the enhancements more than doubled the amount of relief projected to be provided under the Initiative. The enhancements provide broader, deeper and faster debt relief mainly through: (i) a lowering of the ratios considered to provide debt sustainability (together with a lowering of the minimum thresholds to qualify for the openness/fiscal criteria), (ii) replacing the principally fixed three-year period between decision and completion points by the concept of a floating completion point, and (iii) the provision of interim relief from some creditors between the decision point and the completion point. In order to accelerate the provision of debt relief, the provision of interim assistance beginning at the decision point, and the adoption of a “floating completion point” represented the two main changes to the original Initiative. The underlying objective of the enhanced Initiative is to channel the government resources freed up because of debt relief into poverty reduction programs. Under the programs being negotiated by the World Bank and the IMF with countries eligible for debt relief, government 23 B. G. Gunter (2001), “Does the HIPC Initiative Achieve its Goal of Debt Sustainability?”, in the WIDER/UNU Conference on Debt Relief Helsinki, Finland, 17-18 August. 257 spending on public services—such as preventive health care and primary education—that directly affect the poor must increase. Linking debt service relief to poverty reduction The initiative calls for countries to prepare a comprehensive, “country owned” poverty reduction strategy before completing the initiative. First, HIPC debt relief is primarily targeted at lowering external debt to sustainable levels, with benefits accruing over time and not only in the short run. The time profile of HIPC assistance may differ from that of desirable changes in social spending. Second, the fiscal space created by HIPC debt relief for direct increases in social expenditures is determined by the size of the actual cash flow savings generated by debt reduction. And may not be in some cases very large in the early years, when compared to debt service paid in the past, especially in the case of countries where not all debt service due was in fact being paid. Third, governments are often faced with absorptive capacity constraints to implement social programs expeditiously and efficiently. HIPC debt relief should be used in such a way as to maximise development effectiveness. Finally, when considering the resources available for social development, it should be recognised that most HIPCs are already receiving large positive net transfers from creditors and donors that enable them to pursue their development agenda. Debt relief should not be seen as a substitute for continued inflows of development finance. Poverty reduction strategy papers (PRSPs) will be introduced, country-by-country, according to a timetable that is as ambitious as possible while recognising these constraints; and is country-driven, reflecting the government’s stated commitment to poverty reduction. For HIPCs generally, the PRSPs will need to include clear, monitorable key actions that would allow, if endorsed by the two Boards, a country to reach its completion point under the HIPC Initiative. It is also essential in this context that all available resources are integrated in a transparent, accountable budgetary framework, which could include poverty/social funds, to ensure their effective use to combat poverty. For HIPCs that have already reached their decision or completion points under the original framework, additional debt relief will be assessed based on their progress in designing and implementing their poverty reduction strategy24. This approach underpins a proposed tri-partite (government/Bank/IMF) PRSP and provides the basis for ensuring that HIPC debt relief is an integral part of poverty reduction efforts. The use of PRSPs would be pioneered in the ESAF/IDA pilot cases for enhanced Bank-Fund collaboration and in HIPC countries where it would also provide a basis for ensuring a robust link between debt relief and poverty reduction25. Interim assistance and the interim PRSP In order to facilitate the eligibility of HIPCs, the concept of an Interim PRSP was introduced as a requested demonstration of good track records towards the implementation of a full PRSP. In September 1999, the Development and Interim Committees endorsed a new framework for Bank and Fund efforts to support our low-income members. Under this approach, building on the principles of the Comprehensive Development Framework (CDF), 24 IMF and the World Bank (1999), IMF and IDA Poverty Reduction Strategy Papers—Status and Next Steps, Washington D.C., November 19. 25 IMF and World Bank Development Committee (1999), Building Poverty Reduction Strategies in Developing Countries, Washington D.C., September 27, 1999. 258 nationally-owned, participatory poverty reduction strategies (PRSs)—embodied in PRSPs— form the basis for Bank and Fund concessional lending and for debt relief under the enhanced HIPC Initiative. Since the Boards’ discussions of the first interim PRSP (I-PRSP) in January 2000, momentum has been building steadily. By mid-September, 15 poverty reduction strategy papers had been considered by the Boards—13 I-PRSPs and two full PRSPs. Several other countries have either prepared I-PRSPs or are in advanced stages of drafting them. While we have a relatively small sample of country experiences to draw on, there have been a number of positive developments, and some emerging areas of concern. The interim PRSP can serve as a basis for taking a country to a decision point within the enhanced HIPC process. In an interim PRSP, a government should convey its commitment to poverty reduction, indicate its overall strategic goals and programs to address the issue, and define an action plan that would eventually lead to the articulation and adoption of a PRSP in a participatory process. The companion PRSP paper discusses in greater detail the early experience with this process. The provision of interim assistance from the decision point enables countries to benefit from debt relief while still preparing a full PRSP and putting in place other policy measures and actions that may be required for reaching the completion point. Interim assistance effectively reduces the burden of debt service payments on HIPCs while they prepare a full PRSP so that this can be done with due care and diligence. Under the enhanced Initiative all creditors are encouraged to provide some portion of their total assistance to each eligible country during the interim period. Interim assistance had been provided under the original Initiative by bilateral creditors in the Paris Club through flow rescheduling involving an NPV reduction of 80 percent on eligible debt. This interim assistance has now been topped-up by the Paris Club to 90 percent (or higher if needed) under Cologne terms. The predominant method of delivering relief on IDA debt will be through debt service relief. IDA delivers debt service relief, beginning at the decision point, with the aim of delivering its full share of debt relief to a country within 20 years after the decision point. Within this overall 20-year objective, IDA would provide annually, relief of not less than 50 percent of the IDA debt service due on the amounts disbursed and outstanding at the reference year when enhanced HIPC assistance is calculated. This level of debt service forgiveness would apply during the interim period, with maximum interim relief equal to one-third of the total NPV to be provided by IDA. The IMF, subject to receipt of satisfactory financing assurances from other creditors, is committed to deliver interim assistance in the form of grants. These grants are paid into a country’s account, administered by the IMF as Trustee, and used to help meet the country’s debt service payments to the IMF. A country may receive as much as 20 percent of total IMF assistance (i.e., 20 percent of the total IMF share of relief for the country) each year during the interim period, up to the total annual debt service due to the IMF by that country. Interim assistance is not to exceed 60 percent of the overall IMF assistance. The Central American Bank for Economic Integration, the European Union, the European Investment Bank, and OPEC Fund, are either providing or have indicated their willingness to provide interim assistance. The African Development Bank agreed on its modalities for participation in the enhanced framework, including provision of interim relief. Other multilateral development banks (MDBs) have been considering modalities to participate in a similar fashion. Floating completion point The second modification to the original framework designed to accelerate delivery of debt relief was the adoption of floating completion points. This eliminated the original three-year 259 interim period in favour of linking the completion point to the development and implementation of a PRSP, and the implementation of a set of key, pre-defined structural and social reforms. This places the timing of the completion point more clearly under the control of country authorities and enhances country-ownership of the reform program. Based on the assessment by authorities for each of the current decision point countries, the interim period is expected to average about 15 months. Due to their special nature, the interim period for retroactive cases is expected to average less than one year. Similar to the original initiative, the World Bank and IMF prepare a Debt Sustainability Analysis (DSA) and calculate whether a Paris Club stock-of-debt operation is sufficient to bring the debt to a sustainable level. The threshold levels to define debt sustainability, however, have changed somewhat If a Paris Club stock-of-debt operation is not sufficient to reach a sustainable debt level, the country can go on to the second stage. At this point, all creditors commit themselves to delivering a certain amount of debt relief at the Completion Point. The second stage has changed a lot compared to the original initiative: • The country must implement the PRSP. • World Bank and IMF provide ‘front-loaded’ interim debt service reduction, instead of providing IDA grants rather than IDA-credits, as proposed in the original initiative. • Other multilateral and bilateral creditors provide interim debt service reduction. Paris Club creditors offer flow rescheduling involving NPV reduction of up to 90% - instead of the original 80% - on eligible debt. When the country reaches the Completion Point, the creditors deliver the debt relief agreed at the Decision Point, minus the interim-assistance delivered in the second stage. Paris Club creditors have agreed to offer up to 90% stock of debt relief. Some creditors have announced their intention to go beyond this unilaterally and cancel up to 100% of export credits, including post cut-off date debts. Countries must seek equivalent treatment from non-Paris Club bilateral and commercial creditors. Multilateral creditors deliver additional debt relief so that countries reach debt sustainability targets. Deeper Debt Relief Under the enhanced framework, the target sustainability ratios were lowered to provide a greater cushion against unanticipated economic shocks to export earnings. This is intended to increase the probability of a permanent exit from unsustainable debt, and to significantly lower debt-service payments, freeing up resources to support poverty reduction efforts. The amended sustainability thresholds under the enhanced HIPC framework are 150 percent NPV of debt-to-exports and a debt service ratio of 15-20 percent. The fiscal window threshold was lowered to 250 percent NPV of debt-to-revenues, with qualifying criteria reduced to 15 percent for the revenues-to-GDP ratio and 30 percent for the exports-to-GDP ratio. Tab. 2 - Debt Sustainability Threshold Levels NPV debt-to-exports NPV debt-to-Government Revenues Qualifying criteria: - Exports/GDP - Government Revenue/GDP - Debt service-to-exports HIPC I 200-250% 280% HIPC II 150% 250% ≥40% ≥20% 20-25% ≥30% ≥15% 10-15% 260 The introduction of Poverty reduction and growth facilities (PRGF) A central objective of the enhanced framework was that countries would prepare credible, effective and monitorable poverty reduction strategies as the basis for access and use of extraordinary debt relief and other sources of concessional assistance provided by the Bank, the Fund, and other development partners. With regard to debt relief, institutional arrangements and procedures need to be developed to ensure that the savings from debt service are channelled towards programs linked to poverty reduction. Bank and Fund teams are collaborating to assess country capabilities and provide guidance to governments on mechanisms to track and report on poverty-related public spending in HIPCs. A preliminary assessment of early cases suggests that funds “saved” through debt relief should be seen as part of the overall budget and monitored through the country’s own public financial management system. While countries themselves should bear the primary responsibility for monitoring and reporting, Bank and Fund staffs’ work in this area should emphasise providing assistance to HIPCs to strengthen their own public expenditure management systems. In order to demonstrate the new emphasis on poverty reduction, IMF committed to move from ESAF to Poverty reduction and growth facilities (PRGF). The core aim of the PRGF is to arrive at policies that are more clearly focused on economic growth and poverty reduction and, as a result of better national ownership, more consistently implemented. To achieve this, PRGF-supported programs are being derived from the PRSPs, and are readily contrasted with ESAF-supported programs: • The objective is different: The PRGF explicitly makes poverty reduction a central goal, whereas under the ESAF poverty reduction was an implicit by-product. Hence, policies now have to be weighed in terms of their contribution to poverty reduction. • The relationship with the country's strategy is different: The specific measures supported by a PRGF loan arrangement have to derive from the poverty reduction strategy described in the country's PRSP, which is also the basis for all other official creditor support. The PRGF is therefore part of a more coherent and country-led approach to poverty-reduction policies, with the macroeconomic and povertyreduction elements of the economic program better integrated than in the past. • The way programs are formulated is different: The Policy Framework Papers (PFP), the basis of ESAF loan arrangements, were prepared jointly by country officials and IMF and World Bank staff without broader consultation. PRSPs, and thus PRGFsupported policy programs, are country-led and incorporate contributions to policy design from across society. Because PRGF-related documents are published more extensively than under the ESAF, programs are more transparent, enabling other donors to use PRSPs as the basis of support as well. • The nature of conditionality is different: The PRGF's emphasis on country leadership and enhanced collaboration with the Bank mean that IMF conditionality is less extensive and more focused on the Fund's core areas of responsibility than before. • The link with the World Bank is different: the Bank and Fund jointly assess the PRSP, which then serves as the basis for concessional lending by both institutions. That way, the two institutions can tailor assistance to fit their respective areas of responsibility in supporting the PRSP strategy. Thus, there is both more collaboration and a clearer division of labor. 261 These changes should imply: greater transparency; a pro-poor shift in public expenditures; more focus on governance and accountability for public resources; streamlined conditionality; and willingness by some donors (such as the United Kingdom and the European Union), to use the PRSP as a basis for their aid. Over the longer term, this entails working to strengthening their entire financial management architecture. In the short-run, Bank and Fund staffs will assist in selected areas of public expenditure management, including working with those tools currently in place in each country for monitoring specific types of poverty-oriented spending. Some countries, such as Rwanda and Malawi, have a basic capacity to program, track and report on poverty-related recurrent public expenditures and the potential to relate these to social indicators. Other countries, such as Cameroon, are responding to significant shortcomings in their overall public expenditure systems by establishing special HIPC-arrangements and accounts to identify and track spending on poverty related programs. In virtually all cases, technical assistance and extensive efforts at institution building will continue to be needed. Long-term debt sustainability The weakening of commodity prices, notably the double digit declines since 1997 of the prices of cotton, coffee, cocoa, maize, sugar, and copper, combined with a sharp increase in oil prices over the past year, has adversely affected a large number of HIPCs that rely heavily on commodity exports and petroleum imports. As a result, about half of the 41 HIPCs are estimated to have suffered major terms of trade losses in 1999–2000 compared to 1997. These developments have created a difficult external environment for many HIPCs as they seek to maintain macroeconomic stability and establish track records for qualification under the enhanced HIPC Initiative. They also highlighted the difficulties and challenges many HIPCs are facing in achieving long-term debt sustainability, a key objective of the HIPC Initiative. Debt relief delivered under the enhanced Initiative allows HIPCs to return to external debt sustainability. However, maintaining external debt at sustainable levels ultimately requires that HIPCs continue to pursue sound economic policies, in particular timely adjustment to any economic shocks that may weaken their debt service capacity, prudent debt management, and adequate financing on appropriate concessional terms. In the context of PRGF- and IDAsupported programs, many HIPCs have undertaken to accelerate structural and institutional reforms to improve the efficiency of domestic resource mobilisation and utilisation, thereby increasing their resilience to adverse external developments. In view of the continuing large financing needs of many HIPCs, adherence to a policy on new borrowing that is consistent with long-term debt sustainability will play a particularly important role in preventing their external debt from rising again to an unsustainable level. Depending on each country’s specific circumstances, such a policy may require a strict limit on, or even exclusion in some cases of any, new borrowing by the public sector on nonconcessional terms. It may also entail a need for restraint on borrowing on concessional terms, and renewed efforts to shift concessional external financing from loans to grants. Clearly, such a policy also entails action by donor agencies to provide financing for poverty expenditures and development projects. Again, the sustainability must be implicitly referred to commitments and actions from the lenders’ side of debt game. 262 THE HEAVILY INDEBTED POOR COUNTRIES ENHANCED (HIPC-2) INITIATIVE Summary First Stage Country established three-year track record of good performance and develops together with civil society a Poverty Reduction Strategy Paper (PRSP); in early cases, an interim PRSP may be sufficient to reach the decision point. Paris Club provides flow rescheduling as per current Naples terms, i.e. Rescheduling of debt service on eligible debt falling due during the three-year consolidation period (up to 67 percent reduction on eligible maturities on a net present value basis). Other bilateral and commercial creditors provide at least comparable treatment. Multilateral institutions continue to provide support within the framework of a comprehensive poverty reduction strategy designed by governments, with broad participation of civil society and donor community. Exit · Either... Paris Club stock-ofDebt operation under Naples terms (up to 67 percent present value reduction of eligible debt) and comparable treatment by other bilateral and commercial creditors is adequate for the country to reach sustainability by the completion point-country not eligible for HIPC Debt Initiative. Eligible Or... Paris Club stock-of-debt operation (on Naples terms) not sufficient for the country's overall debt to become sustainable by the completion pointcountry requests additional support under the HIPC Debt Initiative All creditors (multilateral, bilateral, and commercial) commit debt relief to be delivered at the floating completion point. The amount of assistance depends on the need to bring the debt to a sustainable level at the decision point. This is calculated based on latest available data at the decision point. Second Stage Country establishes a second track record by implementing the policies determined at the decision point (which are triggers to reaching the floating completion point) and linked to the (interim) PRSP. - World Bank and IMF provide interim assistance. • Oher multilateral and bilateral creditors and donors provide interim debt relief at their discretion. - All creditors continue to provide support within the framework of a comprehensive poverty reduction strategy designed by governments, with broad participation of civil society and donor community. “Floating” Completion Point Timing of completion point is tied to the implemention of policies determined at the decision point.All creditors provide the assistance determined at the decision point; interim debt relief provided between decision and completion points counts towards this assistance: - Paris Club goes beyond Naples terms to provide more concessional debt reduction of up to 90 percent in NPVterms (and if needed even higher) on eligible debt so as to achieve an exit from unsustainable debt. - Other bilateral and commercial creditors provide at least comparable treatment on stock of debt. - Multilateral institutions take additional measures, as may be needed, for the country's debt to be reduced to a sustainable level, each choosing from a menu of options, and ensuring broad and equitable participation by all creditors involved. 263 Financing the enhanced HIPC Initiative More debt relief means higher cost to creditors. Positive developments to date and the initiatives being taken to accelerate implementation risk being undermined if further rapid progress is not made on financing. Total costs for the enhanced HIPC Initiative have been estimated at US$28.5 billion in end-1999 NPV terms—a slight increase compared to the original HIPC costing. The breakdown between bilateral and multilateral creditors has remained roughly even. Paris Club members account for most of bilateral creditor costs (US$11 billion in NPV terms), and IDA, the IMF, AfDB, and IaDB account for the bulk of multilateral creditor costs (in NPV, US$6.2, $2.2, $2.4, and $1.2 billion, respectively). Commitments from bilateral creditors have mostly come from Paris Club creditors (many of which have announced bilateral debt relief beyond their assistance under the HIPC Initiative, according to G8 Köln meeting). The staffs have been working on a case-by-case basis with non-Paris Club official creditors to discuss their participation, recognising the difficulties some of them face - especially those that themselves qualify for assistance under the Initiative, such as Tanzania. Meanwhile, there has been steady progress in securing confirmation of participation from multilateral creditors. The African Development Fund’s Deputies agreed on the financing for its participation in the HIPC Initiative for the near term countries. A financing framework has been agreed for the Inter-American Development Bank, but donor pledges remain to be secured, holding up the release of IMF interim assistance to Honduras. More generally, shortfalls in resources could emerge in the next future. Tab. 3 - Status of Paris Club Rescheduling SSA Countries as of End-July 2001 Countries that rescheduling graduated Equatorial Guinea (1) Uganda (4) from Countries with agreements in effect rescheduling Countries with previous rescheduling agreements, but without current rescheduling agreements, which have not graduated from rescheduling 2/96 Benin (4) 9/00 Burkina Faso (4) Cameroon (4) Chad (4) Ethiopia (4) Guinea (4) Guinea-Bissau (4) Madagascar (4) Mali (4) Malawi (4) Mauritania (4) Mozambique (3) Niger (4) Rwanda (4) Sao Tome & Principe (2) Senegal (4) Tanzania (4) Zambia (4) Sources: Paris Club Secretariat; and Fund staff estimates. 12/01 12/01 12/03 3/03 3/04 4/04 12/03 2/04 12/01 12/03 6/02 12/01 12/03 12/01 4/03 12/01 3/03 3/02 Angola Central Afric. Rep. (2) The Côte d'Ivoire (3) Congo, Republic of Congo, Dem. Rep. of Gambia, The Liberia (4) Sierra Leone (2) Somalia Sudan Togo (2) 9/90 6/01 3/01 6/99 6/90 9/87 6/85 12/97 12/88 12/84 6/98 (1) denotes rescheduling on London terms, (2) denotes rescheduling on Naples terms, (3) denotes rescheduling on Lyon terms, and (4) denotes rescheduling on Cologne terms. Depending on progress with individual country cases, and consistent with the objectives of the overall funding package for the interim PRGF and the Fund’s contribution to the HIPC Initiative, the IMF would need to consider carefully whether to continue to provide debt relief 264 under the HIPC Initiative beyond late 2001 in the absence of the release of the remainder of the investment income on the profits from gold transactions. Additional funding is also required for MDBs other than IDA, a funding need that will become particularly acute as the decision and completion points for countries with large exposure to the IaDB and AfDB are reached. IDA itself will require donor-funding beginning around 2005. The Executive Boards of IDA and the IMF discussed the costing and financing issue at their meetings, and called on all creditors to meet their obligations under the Initiative as expeditiously as possible, and to provide the additional financing needed to ensure full implementation of the HIPC Initiative. 5. The implementation of the HIPC-2 As of June 2001, 23 countries have reached their decision point under the enhanced Heavily Indebted Poor Countries (HIPC) Initiative framework and are now receiving debt service relief under the Initiative which will amount to about US$34 billion over time, or a reduction of US$20 billion in the net present value (NPV) of their outstanding stock of debt. This is approximately 70 percent of the total relief projected to be delivered under the Initiative. 265 Within these 23 countries, three countries (Bolivia, Mozambique and Uganda) reached its completion point under the original HIPC Initiative. The IMF and the World Bank expect that about 35 countries could ultimately qualify for assistance under the HIPC Initiative. However, roughly a dozen of the countries which have yet to qualify for HIPC debt relief are either currently engaged in, or have recently ended, internal or cross-border armed conflict, or are struggling with severe governance problems which have made it impossible to move forward with HIPC assistance. For the 23 countries that have reached their decision point so far, the enhanced HIPC Initiative aims at slashing debt stocks, lowering debt service (compared with actual annual payments made in 1998-99) and boosting social spending. Within the group of 23 countries, four countries - Guyana, Honduras, Mauritania, and Senegal – qualified under the fiscal criterion, 18 countries, which qualified under the export criteria. Referring to a specific case of implementation, on September 2001, Mozambique became the third country to reach this point (after Bolivia and Uganda). Debt service relief under the enhanced HIPC Initiative from all of Mozambique's creditors will amount to approximately US$600 million. US$306 million in net present value (NPV) terms - this includes an additional $53 million to be provided in light of revisions that have been made to the debt data available at the April 2000 decision point. Including assistance provided under the earlier HIPC Initiative, this brings total estimated debt service relief to about $ 4.3 billion. As a result of HIPC assistance and bilateral debt relief already committed, Mozambique's total external debt is reduced by some 73 percent, and possible additional bilateral relief could raise this figure. The NPV of debt-to-export ratio is expected to remain well below the target ceiling of 150 percent throughout the period 2000-2020 (averaging 77 percent from 2000-2010; and about 44 percent from 2011-2020). Debt service payments are cut almost in half (from over $100 million in 1998 to an average of US$56 million from 2002 to 2010), creating room for additional public expenditures on poverty reduction. Debt service as a percentage of government revenue is reduced from 23 percent to an average of just under 10 percent over 2000-2010 and 7 percent over 2011 – 2020. Resources made available by debt relief provided under the HIPC Initiative will be allocated to key anti-poverty programs, which are outlined in Mozambique's Poverty Reduction Strategy Paper (PRSP). Mozambique's PRSP (which is officially designated as the Action Plan for the Reduction of Absolute Poverty) was approved by Mozambique's Council of Ministers in April 2001. The PRSP makes explicit the key role of accelerated and broad-based economic growth as a significant element in Mozambique’s poverty reduction effort, and sets out six priority areas for continued focus (i) education; (ii) health; (iii) agriculture and rural development; (iv) basic infrastructure; (v) good governance; and (vi) macroeconomic and financial management. In a joint assessment, Bank and IMF staff indicated that Mozambique's PRSP contains a sound strategy for poverty reduction, including its incorporation of inputs from civil society and key stakeholders. Despite the impact of the floods of early 2000, GDP growth for 2000 measured 2.1 percent in real terms, reflecting aid-financed reconstruction, new foreign direct investment, and the start of production at a new aluminium smelter. These same forces have continued to drive the economy; a return to near double-digit growth is expected in the near future. Progress was also recorded in a smaller-than-expected primary deficit, stabilisation of consumer prices after the flood, and a significant narrowing of the current account deficit as a result of higher net exports from large aluminium and energy projects. By end-2000, all quantitative criteria were met for the PRGF-supported program. 266 Looking at the broad figure, the roughly US$20.7 billion in NPV terms of debt relief committed to the first 23 enhanced decision point countries amounts to nearly half of their total stock of external debt in NPV terms. In combination with other co-ordinated debt reduction mechanisms, the external indebtedness of these countries will be reduced by about two thirds. Looking at debt as a percentage of GDP, this measurement of indebtedness is slashed from about 60 percent in 1999 to 28 percent after HIPC relief, some ten percentage points lower than the average for developing countries overall. Reflecting the reduction in the HIPCs’ stock of debt, there is a significant reduction in long-term debt service obligations beginning in 2001, whether measured against debt service paid in the past, due in the future, or in relation to other economic indicators. For the 23 HIPCs that have reached their enhanced decision points, overall debt service requirements are cut by one-third, or about US$1.1 billion annually during the 2001-2003 period, compared with actual annual payments made in 1998-99. This translates into annual savings per country of close to US$50 million. Moreover, substantial annual debt service savings will be sustained over the coming two decades, releasing to governments over the medium to long term predictable amounts of additional resources for expenditures on poverty reduction programs. And while the absolute dollar amounts vary across countries, the actual debt service savings for these 23 countries over the initial years represents an average of 1.2 percent of GDP. The savings, of course, are even greater when post-HIPC relief debt service obligations are compared to scheduled debt service due in the absence of debt relief under traditional mechanisms and the HIPC Initiative. On that basis of comparison, the average annual savings over the coming three years amount to about US$2.4 billion (of which more than US$1.6 billion is HIPC relief), or 2.7 percent of GDP. 267 6. Recent changes of perspectives on the Current HIPC Framework The official documentation and data provided by the IMF and World Bank on HIPC Initiative implementation give us a detailed description of good performance and results. As we will describe in chapter 8, the picture is quite different from those analyses of other “actors”, who 268 are interested in debt game: international civil society. An important result can be directly derived from these critical external comments to HIPC Initiative, in terms of less enthusiasm of IMF and World Bank for the initiative ‘s outcomes. In fact, on April 2001, the Staffs of the World Bank and the International Monetary Fund presented a report, "The Challenge of Maintaining Long-Term External Debt Sustainability". They admit, for the first time, that HIPC Initiative will not necessarily represent the final exit solution to escape from unsustainable debt, for those poor countries having reached their completion point. This report assesses the extent to which the enhanced HIPC Initiative provides a solid basis for sustainability for the 22 countries, which have reached their decision points under the Initiative. The paper is based on the detailed Debt Sustainability Analyses for these countries and investigates what additional actions will be needed to ensure that debt sustainability is maintained over the longer term. This document stresses that: • To assess long-term debt sustainability, the focus of attention must shift away from this single debt indicator to a more complex and comprehensive view of the development process in which policies, institutions, exogenous factors and debt management play an integral role over time. • While HIPC Initiative assistance will substantially reduce the debt service due on existing debt, maintaining external debt at sustainable levels will depend critically on future policies and growth performance of the HIPCs and on support from the creditor/donor community. • This restricted objective of HIPC Initiative requires a focus on overall debt sustainability as an exit strategy from the rescheduling process, a focus on a country’s track record to ensure that HIPC debt relief will be put to good use, and a focus on the concessionality of new external finance, including grants, for countries benefiting from HIPC debt relief. • long-term debt sustainability will only be achieved if the fundamental causes that triggered the debt build-up in the first place have been redressed. Those fundamental causes include some external factors such as worsening terms of trade and protectionist policies that restrict access to export markets. HIPCs, typically the poorest members of the international community, have a narrow production and export base, heavily dependent upon a few primary commodities, which make them particularly vulnerable to external shocks. Future borrowing on market terms will exacerbate the debt burden of many of these countries. • Fiscal policies are very important. To the extent to which external imbalances are the result of fiscal imbalances (and they often are in SSA case), fiscal consolidation, including tax reform to strengthen the fiscal payments capacity, is a key factor in achieving debt sustainability. Prudent budgeting and reorienting of expenditures from non-productive (such as military expenditures) to growth enhancing activities within a medium-term framework would also help achieve a sustainable fiscal position. • Longer-term growth prospects can be undermined by natural disasters, war, or health threats such as the AIDS epidemic affecting many of the HIPCs, particularly several decision point cases such as Malawi, Rwanda, and Zambia. In such cases, in the absence of adequate grant financing, external indebtedness (and the NPV-of-debt to exports ratio) may need to rise to accommodate the financing of reconstruction and rehabilitation. • Raising the past average annual growth rate by 2-3 percentage points (as projected in the scenarios underlying the DSAs) over a decade should be not easy. A permanent 5 to 6 percent growth path is not feasible, and growth performance similar to the past decades will not ensure long-term debt sustainability. The impact of lower growth targets would be very negative. 269 Source: World Bank (2001), World Development Indicators • Trade policy is an area where trade partners should focus their efforts. The share of HIPCs in international trade has been eroding since the 1970s, down from 2.2 percent of world exports to only 0.7 percent in 1997 (see next Table 1. HIPC Initiative). Even as a proportion of developing countries’ exports, HIPCs’ share has fallen from 8 percent to 2 percent over the same period. • Rates of return on investment in the HIPCs have typically been low. It has been estimated that African HIPCs would need to increase their savings and investment levels substantially in order to achieve the international development goals over the next 10-15 years: at least, at 30 percent of GDP26. • African HIPCs need to be integrated in the world economy, even though there has been an almost complete withdrawal of commercial creditors from lending to the HIPCs. The share of commercial debt in total new disbursements down from one-third in 1980 to below 10 percent in 1990, and less than 1 percent by 1999. • Even though a very good growth performance is assumed, nevertheless the NPV of debtto-exports ratio of Malawi and Niger is projected to remain above 150 percent for 10 years or more. Export trends are important explanatory factors: Malawi had high tobacco exports before the decision point, while Niger is expected to face a continued decline in its main export, uranium. 26 World Bank (2000), “Can Africa Reclaim the 21st Century”, Washington D.C. 270 • In addition to lower export growth, the assumed level of new finance can have a significant impact on the debt stock ratios, and the terms of new borrowing can have a significant impact on the debt service ratios. The assumptions contained in the DSAs, including the volume of grants, and the terms and concessionality of new borrowing, have important implications for long-term sustainability in the HIPCs. While the average grant element of existing debt as of end-1999 for the 22 HIPCs was 30 percent, the DSAs assume that the average grant element in new borrowing for 2000 to 2010 is twice as high at 58 percent (80 percent for Niger). What seems to be particularly important in this document is the fact that, for the first time in a clear way, the IMF and the World Bank admit that the NPV target of 150 percent of exports is simply a guideline for determining HIPC debt relief. It is not an absolute level for sustainability beyond which countries would automatically face debt-servicing problems. Both the original and enhanced HIPC frameworks begin to acknowledge that debt sustainability involves other factors such as debt service ratios both in terms of exports and relative to fiscal capacity. These factors – and real sustainability – are heavily dependent on lenders’ policies. Some of the critical comments expressed by experts, campaigns and NGOs seem to have effectively influenced the passage from HIPC to HIPC-2 and from the very optimistic projections of the past years to the much more prudent conclusions in recent documents. 271 Bibliography G. C. Abbott (1993), Debt Relief and Sustainable Development in Sub-Saharan Africa, Edward Elgar, Cambridge. C. Daseking and R. Powell (2000), “From Toronto Terms to the HIPC Initiative: A Brief History of Debt Relief for Low-Income Countries”, International Economic PolicyReview, Vol. 2, IMF. W. Easterly (1999), How did highly indebted poor countries become highly indebted? Reviewing two decades of debt relief, World Bank, Washington D.C., September. International Development Association (2000), “Zambia. Enhanced HIPC Debt Initiative: President’s Memorandum and Recommendation and Decision Point Document”, International Bank for Reconstruction and Development / World Bank, Washington D.C., November. International Development Association and International Monetary Fund (2001), “Heavily Indebted Poor Countries (HIPC) Initiative: Status of Implementation”, International Bank for Reconstruction and Development / World Bank & International Monetary Fund, Washington D.C. International Development Association(2000), “Zambia. Preliminary Document on the Initiative for Heavily Indebted Poor Countries (HIPC)”, International Bank for Reconstruction and Development / World Bank, Washington D.C. July. International Monetary Fund (2000), “Then Impact of Debt Reduction under the HIPC Initiative on External Debt Service and Social Expenditures”, International Monetary Fund, Washington, D.C., November. International Monetary Fund (2001), “Debt Relief under the Heavily Indebted Poor Countries (HIPC) Initiative: A Factsheet”, International Monetary Fund, Washington, D.C. International Monetary Fund (IMF) (1998), “Report of the Group of Independent Persons Appointed to Conduct an Evaluation of Certain Aspects of the Enhanced Structural Adjustment Facility”, Washington, DC: IMF, June. International Monetary Fund (IMF) (2000), IMF Survey, Supplement, Vol. 29, September 2000. International Monetary Fund (IMF) and World Bank (2001) “The Challenge of Maintaining Long-Term External Debt Sustainability,” Progress Report, Washington, DC: IMF and World Bank. International Monetary Fund (IMF) and World Bank (July 2001) “100 Percent Debt Cancellation? A Response from the IMF and the World Bank,” Washington, DC: IMF and World Bank. International Monetary Fund and International Development Association (1999), “Modifications to the Heavily Indebted Poor Countries (HIPC) Initiative”, International Bank for Reconstruction and Development / World Bank & International Monetary Fund, Washington D.C. D. Ross and K. Ross (1998), “External Debt History of ten Low-income Developing Countries: Lessons from Their Experience”, IMF Working Paper, WP/98/72. UNCTAD (1978), Resolution 165 (S-IX). Debt and Development Problems of Developing Countries on the trade and Development Board at its Ninth Special Session”, 494th Meeting, Geneva, 11 March. 272 World Bank (2001) Aid and Reform in Africa: Lessons from Ten Case Studies, Washington, DC: World Bank. World Bank (2001), “Financial Impact of the HIPC Initiative: First 23 Country Case”, Washington D.C., June. World Bank (2001), Global Development Finance, Washington, DC: World Bank. 273 8. External critiques to the HIPC Initiative 1. Premise As George Bernard Shaw once noted, debts are like any other trap: getting in is easy, but getting out is pretty hard. Thus, the HIPC Initiative has been welcomed by a broad spectrum of observers as an important innovation in addressing the external debt crisis. For a long time, the multilateral debt problem was not even taken seriously. Even though the main part of the poorest SSA countries' debt repayments went to multilateral creditors, the World Bank and IMF for many years denied that there was anything like a multilateral debt problem. That’s why the HIPC Initiative is important, not only because for the first time, creditors acknowledged the problem of multilateral debt, but also because it is the first comprehensive attempt to deal with the debt crisis. The initiative involves commercial, bilateral, and multilateral creditors. Following the Group of Seven (G-7) summits in Toronto (1988), London (1991), and Naples (1994), debt relief provided on eligible debt were 33%, 50%, and 67%, respectively, whereby eligible debt was defined as pre-cut-off-date 1 , non-ODA bilateral debt. ODA debt was excluded from debt reduction, but was usually rescheduled. Easterly’s (1999) characterisation that substantial debt relief has been provided to HIPCs before the HIPC initiative is misleading as traditional debt relief provided first of all only temporary relief on largely unpayable debt, and second debt relief was only provided to eligible debt. Indeed, of the 37 lowincome countries that have rescheduled their debt with the Paris Club during the last two decades, fewer than one-fourth have graduated from the rescheduling process 2 . Sachs (1989) and Krugman (1988) have shown that debt relief can be more efficient than the provision of new loans in cases where debt has accumulated beyond some critical level. During the early 1990s, empirical evidence grew that many of the poorest countries had acquired unsustainably high external debt stocks that had a significantly negative impact on investment and growth3 . At the same time, the framework has been subject to a number of critical comments, directly involving the participation of a great number of civil society organisations, under the common umbrella of the Jubilee 2000 Campaign 4 . From a practical point of view, we can try to regroup different critical comments into some main areas. 2. Definition and targets of debt sustainability From a general point of view, the rationale for adopting an average target for the debt indicators remains weak, and the adoption of country-specific targets is a way to tailor debt relief more accurately to country needs. Insisting on allocating debt relief on the basis of single values of debt 1 The cut-off-date refers to the date a debtor country approached the Paris Club for the first time for a debt rescheduling. For many HIPCs this was in the 1980s. 2 IMF Survey, Supplement, Vol. 29, September 2000, p. 17. And, for a historical description of debt relief for lowincome countries from Toronto terms to the HIPC initiative, see Daseking and Powell (1999). 3 The empirical studies showed that unsustainable debt burdens crippled development, especially in Sub-Saharan Africa: Borensztein (1990), Cardoso (1993), Cohen (1993), Faini and de Melo (1990), Greene and Villanueva (1991), Schmidt-Hebbel and Müller (1992), and Serven and Solimano (1993). For more recent evidence see Elbadawi, Ndulu and Ndung’u (1997) and Gunter (2001). 4 The Jubilee 2000 campaign commenced, with funding from three Christian aid organisations: CAFOD (catholic), Christian Aid (ecumenical) and Tearfund (evangelical). In 1987, the campaign was launched in the US and elsewhere. In October of that year, Jubilee 2000 (UK) became a formal coalition of aid agencies. Compared to HIPC, Jubilee 2000 campaign shared aim with HIPC, though it wanted more countries covered and more money spent. It earmarked 52 countries (compared with 41 under HIPC) with debts totalling $350 billion ($205 billion under HIPC) for relief of $200 to $300 billion ($100 billion under HIPC). 274 indicators, to countries, whose debt-related problems are surely more diverse and complex than could possibly be revealed by a unique indicator, increases the risk of ending up allocating available funding inconsistent with individual country needs. It has been also argued that the target range for the NPV of debt-to-export ratio is too high either to attain the stated objective of debt sustainability or to ensure the expansion of public expenditures on basic social services. Some groups have recommended targets below 200 percent (e.g. 150 percent) and debt-service ratios below 20 percent, while others have argued for maximum relief on humanitarian rather than debt sustainability grounds, with recommendations ranging from partial to total cancellation of external debt of poor countries. It is unlikely that the enhanced HIPC Initiative will achieve even its limited aim of offering an exit from the rescheduling process. More importantly, it is unlikely to free up sufficient resources to tackle poverty in LDCs. First, several LDCs with significant debt burdens are not included in the initiative. For instance, Angola is considered to have a sustainable debt burden, even though all its 1998 debt indicators are above the threshold level and Angola’s arrears almost quadrupled from US$700 million (or 8% of total debt) in 1990 to US$2,704 (22% of total debt) in 1998. As Unctad puts it: “There has been a persistent tendency to underestimate what has been needed, which has in itself contributed to the build up of the debt”5 . Attempts to significantly reduce the burden of bilateral debt through Paris Club debt negotiations failed, mainly because of the exclusion of large part of the debt, as the case of Uganda demonstrates. In 1995 Uganda received a 67% stock reduction, but mainly because previously rescheduled debt was excluded, its US$3.2 billion debt stock only decreased by 3,2%. The problem is, first of all, that the threshold levels to measure debt sustainability are arbitrary and still too high. When the concept of debt sustainability is approached from a human and social development perspective many more LDCs have an unsustainable debt level6 . Thus, another important problem is that sustainability is defined in economic terms and not in terms of human and social development. This in spite of the fact that many in the international community think that debt needs to be seen in a broader context and that a human development perspective must be incorporated into the HIPC Initiative. The human and social development perspective should measure how much debt servicing is “affordable” as a percentage of government revenues, explicitly acknowledging that the scarce resources available to LDC governments should first be used to meet the basic human needs of their population. The HIPC definition of debt sustainability assumes that a country can meet its current and future external debt servicing obligations in full, without recourse to further debt relief, rescheduling or accumulation of arrears, and without compromising growth. Sustainability implies a permanent “exit” from the rescheduling process. But this definition is too narrow from an overall development perspective. Based on earlier work of Cafod and Jubilee 2000, Eurodad supports a new bottom-up approach to define debt sustainability, the “affordable debt-service approach”. Starting point of this approach is that resources available to LDCs’ governments must be first used for essential expenditures that are necessary to fight poverty: clean water, primary health care, education and basic infrastructure. Government revenues that are left after these essential expenditures can then be spent on other important but less essential items, such as capital expenditure, civil infrastructure, police, security, domestic debt servicing – and external debt repayments. In short, the approach defines how much a country can spend on external debt servicing, after it has made sufficient investments in social and economic development and after domestic debt has been serviced. Of course, this will differ for each individual country. This approach requires projecting what realistic potential fiscal revenues might be. 5 Unctad, Least Developed Countries 2000 Report, p139. 6 Sachs et al. (1999). 275 Tab. 1 - Simple method to calculate Affordable Debt Service 1. Domestic budget revenue 2. Essential spending on poverty reduction 16. Domestic debt repayments x 20% = ------------------------------------------------------4. Net Feasible Revenue x 20% = ------------------------------------------------------5. Affordable debt service Eurostep proposes a methodology to project fiscal revenues. A tax threshold is set at the normal international poverty line of $1 per day (conventionally taken as $1 /day at purchasing power of 1985 US dollars). Anyone living below this poverty line is held to be unlikely to be contributing any tax at all. For the better off, tax is paid as normal, but with a tax break of $1 per day. The rest of GDP is considered taxable, with a maximum tax rate of 25% (higher rates could be distortionary). This gives a figure for potential government revenues. Another, simpler method would be to look at recent fiscal receipts. The Cafod paper sets out the basis for $16 per capita spending on primary health care and $12 per capita spending on education and basic infrastructure. The drawback of this simple approach is that it does not reflect the variety of poverty reduction needs in HIPCs and that is probably too low. For instance, the 1997 UNDP Human Development Report estimates that an extra $80 billion must be spent annually on the 1.3 billion people living in poverty. This is an additional $62 spending for each person living under the $1 per day. The HIPC initiative has been designed around the concept of what debt reduction is needed, according to inappropriate debt sustainability indicators instead of what debt reduction is needed for sustainable development. It does not deal with issues of domestic debt, which are important for fiscal sustainability, nor does it measure the adequacy of public resources to address priority development programs after debt service has been paid. The basic indicators 7 don’t reflect the burden of debt. Their emphasis is on debt stock rather than debt service and, from a resources perspective, it is debt servicing that counts. In addition, the focus is on exports. Exports are an important source of foreign exchange, which is needed for debt repayments. But the debt-to-export ratio can be misleading as rapid growth in exports does not always translate into more budgetary resources for the government, and the volatility of currency and commodity markets also make the debt-to-export ratio an unreliable benchmark to predict debt sustainability in the medium term8 . And the threshold to 150% is high, compared to the fact that research suggests that the historical present value of debt-to-export threshold level for HIPCs is 140%. 3. The adoption of fiscal criteria In the mid-1990s debt analysts and policymakers began looking beyond the traditional balance-ofpayments perspective to debt problems, by adopting a fiscal approach. This shift in emphasis rested on the empirical observation that the bulk of poor country debt was (and is) a public sector liability. 7 a net present value of debt-to-exports ratio of 150%, a net present value of debt-to-government revenue of 250% - if the exports/GDP ratio is 30% or more and the government revenue/GDP ratio is 15% or more -, a debt-service-toexports ratio of 15%. 8 As pointed out in the Unctad Least Developed Countries 2000 Report, LDCs exhibit a high degree of vulnerability to external shocks and they are highly dependent on external financial resources. The government has limited control over its own finances: its income depends to a large degree on tax revenues related to raw material exports and spending is conditioned by aid flows and debt service requirements. The unpredictability of debt service obligations discourages private investors. 276 At the heart of the debt capacity of the public sector is what has become known as the “internal transfer problem”. If the government uses most of the borrowed funds for investments in such areas as infrastructure, education, health services, etc., the sustainable level of debt that the government can take on will depend on different factors. Not only on the relationship between the marginal social return on these investments and the marginal cost of borrowing, but also on the governments ability to appropriate sufficient domestic resources (through more tax revenue) for debt service. The fiscal source of debt service problems is thus evident if taxation is not expanded commensurately with maturing public debt service obligations 9 . The inclusion of a fiscal target in the HIPC debt initiative is thus a reflection of a legitimate concern for the fiscal sustainability of poor country debt. A number of commentators think that the framework gives insufficient attention to the fiscal burden of public debt. They argue that debt relief should not be based only on the foreign exchange burden of public debt, but also directly on the government’s debt-servicing capacity, as this would establish a more direct link to social expenditures, and increase debt relief. These commentators recognise that this would likely complicate the framework, as it would also require dealing with domestic debt and tax and expenditure issues. However, they feel that complementing the current framework with a fiscal criterion could likely result in lower debt-sustainability targets than under the current approach. The fiscal criterion is a more appropriate indicator, but it should be referred to a flow perspective rather than a stock one, and the threshold should be lower. For example, Sachs (2000) has expressed the view that the HIPC sustainability criteria have nothing to do with debt sustainability in any real sense. Indeed, the HIPC initiative neglects standard debt-to-GNP ratios and is far too restrictive in the application of the fiscal indicator10 . Evidence is growing that the only useful debt sustainability criteria for HIPCs have to be related to fiscal indicators. The NPV debt-togovernment revenue indicator could – together with poverty levels and other country specific vulnerability factors – be used for the assessment of the long-term debt sustainability11 . A debt service-to-government revenue indicator could – together with needed investments for anti-poverty programs – be used to determine the maximum annual debt service due. In any case, there should be no minimum thresholds to apply these fiscal indicators. Furthermore, countries should not be discouraged to improve their revenue collection and should at the same time not be punished (in form of lower HIPC assistance) for improvements in government revenue collection. Thus, the reference revenue components should be based on long-term backward looking averages ending with the year before the modification gets adopted as a framework, not the country-specific decision point. Referring to the SSA countries, more than ten years ago, Dittus (1989) had analysed the budgetary dimension of the debt crisis in low-income Sub-Saharan Africa and suggested to assign the debt service-to-revenue ratio a central role. But, as regards the fiscal dimension of the HIPC Initiative, the procedure for opening the fiscal window appears to be unsatisfactory. In fact, most of the countries that have reached decision point have used the export window. This emphasis on export-led strategies rather than on fiscal component of development (through public expenditures and revenues) is much more coherent with the market-oriented approach of neo-liberal 9 See: P. Hjertholm (2001), Debt Relief And The Rule of Thumb: Analytical History of Hipc Debt Sustainability Targets, Paper for a WIDER Conference on Debt Relief, Helsinki, August. With respect to the indebted low-income countries in sub-Saharan Africa, a study by Hjertholm (1997) found that fiscal debt burden indicators played a significant role in explaining the poor debt service performance of a large number of sub-Saharan African countries. This study for 20 sub-Saharan African countries, while not generating strong evidence for the narrow debt overhang hypothesis, showed clear evidence of the broad hypothesis in that public debt burdens had several (indirect) effects that were transmitted through macroeconomic variables, such as the inflation rate, exchange rates and exceptional financing. 10 These budgetary aspects of the transfer problem have been analysed by Dani Rodrik and especially by Helmut Reisen, see Hjertholm (1999) for further references. 11 A further sustainability criterion of having a debt service to government revenue ratio of no greater than 12%, is suggested by Martin (2001). 277 marketisation. Unsatisfying is the attached twin conditions that only very open economies with a strong government tax effort are allowed to benefit from the fiscal window. It is still clear that many HIPC countries are not going to benefit from the HIPC II fiscal window, since their tax efforts are often well below the threshold stipulated of 15 percent as a precondition for additional debt relief 12 . Indeed that is one of the reasons why they have a fiscal problem in the first place 13 . Moreover, As argued by Esquivel et al. (1998), the application of the joint tax-openness criteria tends to produce an undesired biased treatment of HIPC countries. It does so by penalizing those countries that are undertaking a substantial tax effort since the implied target value of the debt-toexports ratio (in present value terms) tends to increase as the tax rate increases, and by rewarding countries that are highly export dependent since the implied target value of the debt-to-export ratio tends to fall as the export-to-GDP ratio rises. So only highly open countries with a moderate tax effort are going to benefit from this amendment to the HIPC scheme, although it was intended (or so one would like to think) to deal with the fiscal aspect of the debt of all HIPC countries. 4. Criteria of country eligibility As a direct consequence of incorrect adoption of criteria of sustainability, some groups consider the eligibility criteria to be overly restrictive. They criticise the restriction to poor countries below the IDA operational cut-off as excluding some highly indebted countries that could benefit from relief, though most commentators agree that the poorest countries should have the highest priority in concessional debt relief. Some have suggested that human development criteria be developed as substitute for the NPV of debt-to-export and debt-service-to-export ratios, as measures of sustainability. Others have questioned why it is necessary to exhaust traditional mechanisms in a few cases where multilateral debt is dominant while the benefits of bilateral debt relief may be offset by reduced flows of new aid. For example, Nigeria is highly indebted: its NPV debt-to-export ratio is estimated to be 188% (38% higher than what is considered to be sustainable under the enhanced framework). It is also a poor country: GDP per capita is below US$300 a year and more than 70% of its 120 million population live in absolute poverty (below one dollar-a-day). The official reason for Nigeria’s exclusion from the group of HIPCs is that Nigeria does - due to its large oil reserves - not rely on IDA assistance. In other words, it is expected that Nigeria extracts and sells its oil to generate the foreign exchange and revenues necessary to repay its external debt, most of which was contracted by corrupt previous governments. In reality, to be very restrictive in terms of country eligibility to debt relief means to avoid to involve “big” debtors – as non- IDA middle income countries -, which could be much more expensive for official and commercial lenders. 5. Timing of debt relief The performance period to receive HIPC debt relief has frequently been perceived by outside commentators as too long and as unnecessarily delaying and reducing the potential benefits of the 12 . As noted by Tanzi and Blejer (1988), experience shows that it is very difficult to substantially raise the level of taxation in the short or medium term. Tanzi and Blejer reports from that experience that, unlike in industrial countries, no developing country had been able to raise the tax ratio by ten or twenty percentage points in a matter of one or two decades, or, for that matter, by just several percentage points in a few years. But this is precisely what is asked for in the HIPC context. Moreover, this scenario may then contradict the advice usually given to developing countries about the criticality of expanding the tax base as the primary source of higher government revenue. This advice is based on the assumption that higher tax rates, even if their initial values are relatively low, often act to discourage investment activity. 13 See: P. Hjertholm (2001), ibid. 278 HIPC Initiative. In particular, in the early cases (e.g. Uganda and Bolivia), where countries had already established track records of more than the required six years of good policy performance, there was criticism that the shortening of the interim period should have been more pronounced or even that the decision point should have coincided with the completion point. Some observers have argued that more weight should be given to past performance and the depth of reform, rather than only the length of the track record. They also stressed that delays in the receipt of HIPC debt relief have a high cost in terms of forgone social services. Some case studies demonstrate that the HIPC relief comes too late to address adverse impacts and too little to meet poverty targets. This calls on debtor countries to concentrate on complementary measures and the HIPC initiative to undergo refinements like addressing the time lag between decision and completion points, and incorporating ratio of recurrent expenditure as a threshold level14 . Again, in reality it is clear that to make the HIPC Initiative implementation faster means to oblige donor countries to finance faster the HIPC Trust Fund, which can be undesirable. 6. Amount of debt relief Another main problem is that the debt reduction on HIPC offer is too small. • Debt service is still squeezing out much needed expenditure in health and education and, in HIV/AIDS afflicted countries, preventing countries deal with, or exacerbating the HIV/AIDS epidemic (Oxfam,2000; Eurodad 2000; Drop the Debt 2001; World Development Movement, 2001). • Countries still face increasing debt service payment (Oxfam, 2000; Jubilee 2000, 2000; DFI, 2001). • Projections of future export levels and government revenues are overoptimistic and do not take adequate account of HIPC countries’ vulnerability to shocks (DFI, 2001). Zambia will actually pay more debt service in the years 2001-2005 than they did in 1998-99, implying that no fund is freed at all. For five out of sixteen HIPC-LDCs for which data are available, the NPV of debt to export level in 2005 is still above the IMF and World Bank threshold ratio of 150%. In addition, some countries will see much volatility in debt servicing, severely disrupting government budget processes. Furthermore, after 2005, debt service levels start rising again. For eleven out of the thirteen HIPC-LDCs, for which post-2005 debt service data are available, debt service starts to increase after 2005 and for nine HIPC-LDCs, future levels are far above present debt service levels. The fact that all HIPC-LDCs are projected to have sustainable debt levels in 2018 cannot be attributed to declining debt levels, but to predictions of improved economic performance. As has been the case in the past, however, these projections are very optimistic, and there is a real risk that in the long run, most HIPC-LDCs will be back to where they are now. Apart from rhetoric, the debt reduction offered is likely to be too small and there is no straightforward relation between decreased debt service levels and increased spending on social development. Increased social spending cannot be related directly to decreased debt service. HIPCs have always borrowed money to pay part of their debt service. Consequently, it is impossible to increase spending on social and human development with the same amount as the debt service is lowered without new borrowing or grants. The stated HIPC debt service savings—comparing actual payments before and after the HIPC initiative—are distorted by the fact that some HIPCs made unusual high debt service payments shortly before reaching the enhanced decision point in order to avoid any arrears that could have delayed the reaching of the enhanced decision point. This was the case for Guinea, Malawi, Mauritania, Nicaragua, and Zambia, where actual debt service payments 14 A.V.Y. Mbelle (2001), The Hipc Relief: Too Late, Too Little? Perspectives From A New Qualifier, Tanzania, Unu/Wider Conference On Debt Relief: Helsinki, Finland 17-18 August. 279 in the decision point year increased by 30%, 26%, 11%, 17%, and 24%, respectively, compared to actual debt service payments in the year before the decision point. Excluding these increases, the actual annual debt service savings of the enhanced HIPC initiative are at least US$100 million. Zambia is expected to pay US$22 million more in 2001 (US$158 million) than in 1999 (US$136 million) for debt service, and reached its enhanced decision point in December 2000. 7. The problem of burden-sharing and financing of the initiative Burden sharing is unrelated to economic power. There are largely developing country creditors and sub-regional MDBs that have—due to financial constraints—not fully participated in the provision of HIPC debt relief. Under the original framework, the costs among bilateral creditors were calculated based on the NPV of pre-cut-off date and non- ODA debt. Under the new framework, the burden sharing among bilateral creditors is based on total NPV debt outstanding at the decision point. Comparisons between the two methodologies show that the bilateral costs have been shifted towards non-Paris Club creditors as they have relatively more post cut-off-date debt than the Paris Club creditors. Given that less than 4% of HIPCs’ total external debt was private non-guaranteed (PNG) debt, the architects of the HIPC framework decided to exclude private non-guaranteed debt 15 . For bilateral creditors, cancelling additional debt poses few financial problems. A large part of LDC loans are non-performing loans, as indicated by the piling up of arrears. Writing off a non-performing loan does not cost any new money, it simply involves accepting that the loaned money will not return. In technical terms, it simply needs to be recognised that the revenues from the asset will not be forthcoming and that total revenues will be lower than expected. Provisions can then be made that reduce the size of the asset in the creditor’s account. In this way annual provisions gradually reduce the paper value of the asset to recognise the decline in the asset’s real value. When the paper value reaches zero, the asset has been simply ‘written off’ the financial accounts. For multilateral creditors, cancelling loans is a bit more complicated. Multilateral creditors administer revolving funds of money donated by the shareholders, and have 'preferred creditor status', meaning that their loans are serviced more diligently than those of other creditors. The evidence submitted to Drop the Debt by two independent experts commissioned to look at the question of how resources can be used to fund deeper debt cancellation by the World Bank and IMF suggests indeed that a 100% cancellation of HIPC debt by the Bank is not unrealistic. With the prudent use of a small proportion of the IBRD’s reserves and an ongoing commitment from its net income, and the future use of IDA’s greatly increased reflows 16 more than enough funds can be realised to cancel 100% of the outstanding debts owed by these poorest countries to the World Bank without affecting its financial position and IDA lending by any significance. For the IMF, there is a consensus that it can easily write off its debts to the HIPC countries by using the earning capacity of its general reserves, together with a repeat of limited revaluation of its undervalued gold reserves. Other multilateral creditors (which do not generate nearly as much net income from non- concessional lending in the way that the IBRD does), in contrast, could go bankrupt if they were to offer 100% debt cancellation. Multilateral development banks (MDBs), particularly the African Development Bank, are in fact already technically bankrupt and rely on transfers from donors or on trust funds administered by the World Bank 17 . 15 Only a few SSA LDCs, such as Mozambique and Angola, owe a significant amount of debt to commercial banks and other private creditors. 16 Its income from loans made earlier. 17 Drop the Debt (2001), Reality Check. The Need for Deeper Debt Cancellation and the Fight Against HIV/AIDS. April 10. 280 The problem of burden sharing implies a problem of financing of the initiative, too. Commentators favouring an expansion of the HIPC Initiative have recommended that net additional resources need to be made available by bilateral and multilateral creditors and donors for debt relief. They recommend more progress in increasing Official Development Assistance (ODA) levels towards attaining the UN target of 0.7 percent of GNP. They also call for ODA to be more effective and better targeted to benefit poor people. Other groups have pressured their governments and legislatures to increase funding for the Initiative. It is clear that multilateral assistance to the HIPC initiative can only be additional to traditional multilateral development assistance if overall budget allocations to MDBs are increased in real terms. Excluding increases earmarked to fight the HIV/AIDS epidemic, there is currently little indication for this to happen. It is thus realistic to assume that HIPC debt relief will reduce traditional multilateral assistance. Previous analyses found little evidence of additionality with regards to the resource transfers to HIPCs. For example, the Concluding Report of a November 2000 Bretton Woods Committee Roundtable Discussion on Reassessing Debt Relief stated that roundtable participants, including Bank and Fund officials, were discouraged that at an early stage, little evidence of additionality could be found. Excluding Bolivia, disbursements of commercial banks to the group of HIPCs decreased from US$1.11 billion during 1994-96, to US$0.85 billion during 1997-99, even though disbursements of commercial banks to Sub-Saharan Africa increased during the same time from US$1.98 billion to US$3.04 billion. In real terms (using the SDR interest rate as discount rate), the 1997-99 average disbursements of official creditors excluding the IMF to the group of HIPCs have been 18 percent lower than the 1994-96 average disbursements. What makes these overall negative trends worse is that even concessional disbursements, especially bilateral concessional disbursements, have decreased. As long as bilateral creditors believe that the HIPC initiative provides debt sustainability, they have no reason to continue shifting disbursements from decision point HIPCs to other countries that are not covered under the HIPC initiative. However, if bilateral creditors believed that the original HIPC initiative did not provide an exit from future debt rescheduling, they would continue to shift disbursements from HIPC decision point countries to non- HIPCs, even after reaching the decision point. Though there may be other explanations, data seem to generally support the hypothesis that bilateral creditors did not believe that the original HIPC initiative would provide a lasting exit from future debt rescheduling. Combining the various results, there is some indication that (i) the adoption of the HIPC Initiative led to a reduction in disbursements to HIPCs and (ii) HIPC debt relief has been deducted from traditional development assistance. Given that overall development budgets are generally decreasing in real terms, it is to expect that this tendency might continue. In other words, HIPCs could end up with no additional resources for poverty reduction. Italian case This ambiguity can be adequately demonstrated in the Italian case, taking in account the fact that Italy is the only G7 member country to have approved by the half of 2000 a national law related to the HIPC Initiative. In fact, the purpose of Italian law18 is to make operative the understanding reached by creditor countries at multilateral level regarding the treatment of the foreign debt reduction of the developing countries with the lowest income and highest debt levels. The developing countries subject to debt cancellation on the part of the Italian State are those qualified for soft loans from the International Development Association (IDA). Therefore the Italian programme for debt reduction potentially encompasses a larger group of countries (78, or 66 if we consider the IDA-only countries) than that considered in the multilateral negotiations which were restricted to the Heavily Indebted Poor Countries (41). Furthermore, in relation to the latter group, 18 Law N. 209/2000, “Measures for the Foreign Debt Reduction of Countries with the Lowest Levels of Income and Highest Levels of Debt”. 281 the Italian government reserves itself the capacity to bilaterally negotiate debt reduction and concede debt cancellation under terms, timing and conditions different from those agreed between the creditor Countries at multilateral level. The Parliament proposes to resort to the International Court of Justice for an opinion on the legal aspects of foreign debt reduction in order to proceed in accordance with the general principles of law and within the framework of the rights of man and of peoples. To qualify for debt cancellation, potential candidates must commit themselves to respect for human rights and fundamental liberties, to the repudiation of war as a means to solve international conflicts, and to the pursuit of full human and personal development, particularly with regard to the reduction of poverty. In the event of natural catastrophes and serious humanitarian crisis, credits relative to financial aid on the part of Italy may be totally or partially cancelled. The following sums have been established as object of total or partial debt cancellation: a) an amount not inferior to 3 thousand billion lire relative to aid credits (around US$ 1.5 billions); b) insured credits held by the Italian Export Credits Guarantee Department (SACE) in the amount not inferior to 5 thousand billion lire (around US$ 2.5 billions); and c) a total amount not inferior to 8 thousand billion lire (around US$ 4 billions). The debt may also be reduced through different kinds of “social swaps”, as conversions into investments in development and poverty reduction programmes. Finally, a term of Three years from the effective date of the present law has been established within which the above mentioned sum of at least 8 thousand billion lire should be cancelled. Criteria and formalities for the execution thereof have been fixed by decree of the Treasury, in 2001. The conditions, types and terms of the cancellations, including eventual conversion operations, will be defined in apposite bilateral intergovernmental agreements with each of the interested countries. The Treasury will submit to the Parliament a yearly report on the execution of the law, containing detailed information on each of the beneficiary countries, on the value and conditions of the original debt, the complete list of the projects and executors relative to cancelled aid credits, the complete list of insured credits cancelled, together with the list of respective guarantees supplied by debtor countries and the beneficiaries of compensations previously made by SACE relative to cancelled credits. If the spirit of this Law were respected by decree of the Treasury, it could effectively have represented an innovative instrument, to be considered as an immediate and una tantum initiative to cope with the foreign debt emergency and to prepare a new international order for financing development, on which this Law does not provide any specific detail as it could be part of the debate on the reform of Italian development cooperation19 . But, strictly linked to the current Law on debt cancellation and reduction, as the mechanism to translate its principles into operative terms, is the Treasury’s decree on criteria and formalities for the execution of Law. The rules for the enforcement of Law n. 209/2000 were approved the 4 April 2001, n. 185. This decree of the Treasury Ministry have drastically reduced the space of any national initiatives, the rules for the enforcement of Law, n. 185, by the Treasury and Ministry for Foreign Affairs, and it has carried back the application within the context of the Paris Club agreements. As a consequence, the embedded innovation of Law has been dramatically retrenched. It must be noted that this decree was rejected by the Parliament; nevertheless it was approved. 19 The new law embodying Italian bilateral ODA policy reforms has not passed. The legislative process was long, taking all the year, and it was considered to be the condition for expanding the limited current stock of direct bilateral ODA and for overcoming he constraints and deficiencies of the current aid management system. 282 What seems evident is that debt relief represents a great opportunity for bilateral donors to take officially their “responsibilities” in terms of ODA commitment, without real disbursement (NPV is lower than nominal value) and simplifying the mechanism of disbursements (compared to the procedures of project or programmes implementation). Italian recent figures confirm the real facts, as we demonstrated in chapter 6, Appendix 2 (in particular, table 4: the following table is an abstract from that table). Tab. 2 - The importance of debt relief for ODA statistics (1999-2000, US$ million) I. Overseas Development Assistance (ODA) I.A. Bilateral ODA 1. Grants. i) Debt forgiveness, total (incl. forgiven interest) (a+b+c) 2. Non-grant bilateral ODA i) Rescheduling ODA claims (capitalised interest) ii) Other lending I.B. Multilateral ODA 1. Grants and capital subscriptions, total i) IDA ii) IBRD,IFC,MIGA iii) Regional development banks - HIPC Inititative II. Other Official Flows (OOF) (II.A+II.B) II.A. Other Official Bilateral Flows 1. Rescheduling, total i) OOF claims (capitalised interest) ii) OOF component of debt service reduction 2. Offsetting entry for debt relief II.B. Transactions with Multil. Agencies at Market Terms - Offsetting entry for forgiven interest Amounts 1999 1998.98 643.98 550.76 101.92 93.22 3.41 89.81 1355.00 1355.00 297.06 5.77 146.65 189.63 189.63 189.63 189.63 / 0.00 / Extended 2000 1598.93 599.47 524.81 201.47 74.66 15.04 59.62 999.46 999.46 18.70 75.89 23.80 103.01 103.01 103.01 9.48 93.53 Amounts Received 1999 2000 -193.26 -222.67 -193.26 -222.67 0.00 0.00 / -193.26 -222.67 -193.26 0.00 0.00 -222.67 0.00 0.00 / -170.77 -170.77 0.00 -206.50 -206.50 0.00 -170.77 0.00 -120.77 -206.50 -88.50 Net Amounts 1999 2000 1805.72 1376.26 450.72 376.80 550.76 524.81 101.92 201.47 -100.04 -148.01 3.41 15.04 -103.45 -163.05 1355.00 999.46 1355.00 999.46 297.06 5.77 18.70 146.65 75.89 0.00 23.80 18.86 -103.49 18.86 -103.49 189.63 103.01 0.00 9.48 189.63 93.53 -170.77 -206.50 0.00 -120.77 -88.50 Source: Italian Ministry of Foreign Affairs data set. 8. Misuse of discount rates and time horizons For LDCs, and in particular for HIPC LDCs, the difference between the nominal and the market value of their debt is likely to be particularly significant. Building on previous econometric estimates of the secondary market prices of middle income countries prior to the Brady Initiative, a recent academic paper shows that the HIPC initiative is about ten times less generous than face value accounting would suggest20 . The economic value of LDC debts can also be calculated directly by using country specific discount rates. These rates can be derived either from data from secondary loan swap markets or data from debt buy-back operations. As interest rates are changing over time, any change in the DSA reference date implies changes in the discount rates and thus represents changes in NPV debt and HIPC assistance. 20 OECD-DC (2000), The HIPC Initiative: True and False Promises, Daniel Cohen, Ecole normale superieure and OECD Development Centre, June. 283 As Uganda’s reassessment under the enhanced HIPC initiative showed, high discount rates and too optimistic export projections can make a country’s debt look sustainable, even though it is not 21 . Together with volatile exports (or volatile government revenues), this methodology results in highly unpredictable numbers of HIPC assistance and HIPC costs. The impact of different discount rates on a country’s HIPC assistance is relevant. Modifications are needed with regards to the use of currency-specific discount rates in order to make the initiative fairer, more predictable, and even simpler. All of the problems related to currency specific discount rates could be avoided by adopting a single (a ten-year backward-looking average) fixed discount rate for all currencies, all creditors, and all debtors. Moreover, there is some doubt if the NPV calculations used in the HIPC framework are appropriate. Among many problems related to discounting, the key argument is that discounting unpayable debt at market discount rates gives the wrong picture about a HIPC’s debt burden. As a World Bank (2001, Table 3) HIPC study shows, at least four countries (Guinea, Mauritania, Niger, and Zambia) will continue to pay between 20-23 percent of their government revenues as external debt service on public or publicly guaranteed debt after enhanced HIPC debt relief. Another important aspect linked to time, is time horizon. In fact, the time horizons of governments, communities and the private sector lead to bad policy (such as real exchange rate overvaluation) and affect debt crisis. The decision to borrow, and the choice of what to do with the loan, is intimately linked to the rate of time preference of the borrower (or, put another way, the rate at which the future is discounted relative to the present). This in turn drives resource exchanges across time, and thus the current account position and other macroeconomic identities (Buiter 1981, Obstfeld and Rogoff, 1996). This is the approach taken by Easterly22 in seeking to explain why debt problems occur, and why they persist: "A country that has gotten an 'excessive' external debt may be one with a high discount rate against the future ... After receiving debt relief, the high discount rate country would like to accumulate the same amount of external debt again". But what causes such a high discount rate? Easterly offers a number of possibilities ranging from high rates of pure time preference (rooted in the shorter expected lifetimes of the populations of poor countries) to politicians with very short time horizons, leading to excessive (debt-financed) current spending without concern to future debtservice. He finds that average policies in HIPCs were generally worse than those of other LDCs, controlling for income, over the period 1980-97. Easterly concludes that governments impose their high discount rates on the rest of society through bad macro-economic policy, leading to debtservice problems. This result supports the IMF and World Bank focus on a country's track record of policy reform, and the use of 'good policy' as a key criterion for eligibility for debt relief. This manifests itself in a shortening of the government's time horizon (the bad policy effect in Easterly). This is a provocative thesis on the cause of the debt crisis and overhang, which deserves a detailed examination from both theoretical and empirical perspectives. 21 Based on Uganda’s original completion point DSA, Uganda’s NPV debt-to-export ratio at end-June 1999 was supposed to be 207%; however, as Uganda’s enhanced decision point DSA showed, the actual NPV debt-to-export ratio was 240%. 22 Easterly, W. (2001), “How Did Highly Indebted Countries become Highly Indebted? Reviewing Two Decades of Debt Relief”, Working Paper 2225, Washington DC, World Bank. 284 As its analytical basis, Easterly’s thesis rests on the inter-temporal borrowing/lending model. He argues that a country with an excessive debt is one with a high discount rate against future and /or a low inter-temporal elasticity of substitution. This views an `excessive debt’ of HIPCs as a reflection of their peculiar order of inter-temporal preference (in particular, that of the public sector), exhibiting a tendency to run down country assets. While interpreting the two key parameters of the model in this very specific perspective, his analysis completely ignores a number of other main structural characteristics of low-income developing economies. As Obstfeld and Rogoff (1996) show, this model visually links the current account concept and the domestic investment-saving gap, and illustrates the role of international borrowing and lending to fill the gap. Thus, accessing the international capital market, i.e. borrowing, allows a country to undertake the extra investment as well as to enjoy the extra first-period of consumption. With the constant property of intertemporal preference, the process of debt relief and a progressive substitution of concessional debt for a non- concessional one is seen as keeping the country perpetually heavily indebted. Thus, the granting of debt relief without ensuring a full adherence to policy conditionality set out by the Structural Adjustment Programmes (SAPs) is predicted to lead to negative saving and declining investment. This effect of debt relief is supposed to be in addition to other purported negative incentive effects, such as the delay of policy reforms in anticipation of “selling” reforms at a higher “price”, or the creation of a moral hazard for borrowing in the expectation of debt forgiveness. Easterly presents a number of disparate empirical evidences to support his thesis of “high discount behaviour” as the cause of HIPC’s misfortune against the alternative hypothesis suggesting that HIPCs became highly indebted due to external shocks. He concludes that debt relief is futile with unchanged long-run preferences 23 . 9. Performance criteria: the growing number of conditionalities Easterly’s thesis clearly supports the presence of conditionalities. Many observers accepted the need for conditionality to lessen the moral hazard problem and ensure that HIPC debt relief is put to good use. Nevertheless, they felt that conditions are too stringent. In particular, the tight link to IMF/ESAF-supported programs has been objected to, particularly by those who have concerns about the design of these programs generally. Conditionality under IDA and IMF adjustment operations has been criticised for not giving sufficient weight to poverty reduction objectives. The social development criteria have sometimes been seen as adding conditionality; and some observers have advocated instead providing a positive incentive for stronger programs to support human development by permitting lower debt sustainability targets in countries with particularly strong programs in this area. On the other hand, some observers have been concerned about the absence of conditionality after the completion point, and the risk of misuse of funds or of the re-appearance of debt problems. The introduction of the PRSP as a condition for debt relief not only involves a change of emphasis, but also an extension of policy conditions. In addition to traditional macroeconomic and structural reforms, the country must also implement a number of agreed social development policies. This 23 M. Nissanke and B. Ferrarini (2001), Debt Dynamics and Contingency Financing: Theoretical Reappraisal of the HIPC Initiative, Paper for a WIDER Conference on Debt Relief, Helsinki, August. 285 involves long, complex and comprehensive development planning, which may delay the road to the Completion Point, the point where countries receive unconditional and irrevocable relief. As is more detailed described in EURODAD (2000), GAO (2000) and Northover (2000), there is an unacceptable tension between the urgent need for debt relief and the time required to build a genuinely participatory PRSP process. Following a suggestion of Oxfam International (2000), the enhanced decision point could be reached once a government produces an Interim Poverty Reduction Strategy Paper that provides some details on the workings of a poverty fund. Moreover, the linkage between the HIPC Initiative and the PRS also erodes the quality of the PRS, as countries are in a rush to enter the initiative. At the same time, the PRSPs do not seem to succeed in aligning macro-economic issues and poverty issues more closely than in the past, and macro-economic policy frameworks have not changed significantly as a result of the PRSP approach. Strictly linked to this critique is that it was a fundamental mistake to let the creditors, led by the IMF and World Bank, work out the original framework of the HIPC initiative 24 . Eurodad’s ‘PRS-Watch’ program, which closely follows the formulation of PRSPs, has found that even though there are many differences between countries' PRS processes, the PRSP concept cannot be characterised as an overall success or a failure. There are some concerns that need to be addressed. First of all, the link between PRSP and the HIPC Initiative is delaying debt relief and lowering the quality of PRSs. PRSPs are another layer of conditionality in an already complicated qualification process for the HIPC Initiative. The interim- PRSP doesn’t solve this problem, simply postpones it. Secondly, the PRSPs do not succeed in aligning macro-economic issues and poverty issues more closely than in the past and macro-economic frameworks haven’t changed significantly as a result of PRSPs. Thirdly, as a result, growth is consistently prioritised as the primary motor of poverty reduction. There has been little attention to quality aspects of growth, such as equity and distribution. This partly results from a lack of a definition of what pro-poor growth is. In fact, poverty concerns are not being placed at the heart of policy making. Fourth, to ensure that policies proposed will maximise benefits for the poor, an open, transparent ex-ante impact assessment is needed. However, there is little evidence of this occurring so far. Fifth, the World Bank and the IMF remain the final arbiters of PRSPs. Given the general rules (and bargaining powers) of debt game, the PRSP therefore hardly involves a change in the relationship between countries and the World Bank and IMF. Instead of the World Bank and IMF approving or rejecting a PRSP on an all or nothing basis, it would be desirable to see a national poverty reduction strategy being presented to all donors equally, for example a UN round Table or a World Bank Consultative Group meeting. 24 Nevertheless, there are some interesting experiences of funds such as the Poverty Action Fund created in Uganda, which is considered a credible solution. Such structures have three main features: - They are integrated in the national budget so that the expenditures being funded are included in the overall development/poverty reduction strategy of the country. - Dedicated disbursement and reporting procedures ensure that the funds allocated to poverty reduction are truly additional to current expenditures and guarantee the transparency of the process. - The poor are involved in the monitoring of the structure. In Uganda, this framework has enabled the country to make an efficient use of the funds freed by debt reductions despite the fact that it was engaged in a war, highly corrupted and that no PRSP had yet been approved by the IFIs. Furthermore, the expenditures programmed through the Poverty Action Fund have had a very significant impact on social indicators (education doubled in a few years) and the participation of the different stakeholders (government, civil society organisations, donors) has been by all means significant and fruitful. 286 Sixth, although participation is better organised and being taken more seriously by governments than in the past, proper participation will take significantly longer than in the past – up to five years. Moreover, there are concerns about the lack of distinction between mere ‘consultation’ – where the views and ideas of civil society are merely solicited – and full participation, where civil society organisations share in decision making. The danger is that consultation could serve to rubber-stamp and legitimise a strategy which civil society has not really had influence over. Seventh, there is little evidence that the PRSP process – particularly because of the rush to get debt relief – builds on existing processes. On the contrary, existing home-grown processes are being shouldered aside by the arrival of PRSP. Although some of these existing processes were slow moving and inadequate, there are still significant implications for the quality of ownership when these are not built upon25 . It has been also argued that measures to provide assistance during the interim period should be strengthened. Observers have argued that if the interim period is to be as long as three years, interim measures should provide greater cash-flow relief, and more multilateral creditors should provide interim relief to allow for needed expansion of development expenditures. Observers have commented that there should be a closer and more visible link between HIPC debtrelief and poverty reduction. Specifically, some have called for debt-relief efforts to be explicitly integrated into a broader strategy to combat poverty as reflected in the human development targets set in “Shaping the 21 st Century: the Contribution of Development Co-operation”, issued by the OECD’s Development Assistance Committee (DAC). 10. Real impact of debt relief Some scholars considered how much poverty could be reduced through the HIPC Initiative. Using a simple distribution function and measures of inequality, US one dollar a day, poverty was estimated for the twenty-three countries. It is showed that the HIPC countries account for relatively little of developing country poverty. Further, full debt cancellation would have a small impact on reducing poverty in most of the HIPC countries themselves. The paper reaches the conclusion that neither a distribution-neutral debt cancellation, nor transferring all debt payments to investment for faster growth would achieve the International Poverty Targets. Therefore, debt relief must be combined with redistribution measures to achieve those targets 26 . Moreover, the HIPC is not able to prevent LDCs from falling again into the same debt trap: LDCs remain dependent on external support and current borrowing and lending practises, all of them being the conditions that generate the vicious circle of indebtedness. If the initiative is to fulfil even its limited aim of an exit from the rescheduling process, much deeper debt reduction is needed. The fact that debt reduction that is currently on offer may in the long run simply be not enough is also demonstrated by the fact that the eligible countries remain dependent on new borrowing, as is reflected in the building up of the debt and increasing debt service levels. As pointed out by the US General Accounting Office, debt reduction only frees up resources if HIPCs continue to borrow at the same rate as in the past. In effect, increased spending on poverty reduction now is being financed by new debt that will come due in the future. As the IMF and the World Bank admit, in their Debt Sustainability Analyses, a relatively neutral external environment' is assumed. However, the economies of LDCs and HIPCs are very vulnerable, because of the volatility of global commodity prices, and several other factors, including: 25 See: Eurodad (2000), Poverty Reduction Strategies: What have we learned so far? September. 26 H. Dagdeviren and J. Weeks (2001), How much poverty could HIPC reduce?, Paper for a WIDER Conference on Debt Relief, Helsinki, August. 287 • Fluctuations in import prices, such as the oil price; • Exchange rate devaluation and its impact on import prices; • The level of donor aid flows (which is assumed to increase significantly); • The occurrence of non-economic shocks, such as climatic shocks, wars, or social conflicts, but also slow-moving shocks, such as the AIDS/HIV pandemic. The vulnerability of HIPCs to shocks and hence to fluctuations in GDP growth and export earnings have been analysed by Martin (2001). He identifies six main sources of shocks, which clearly contradict the IMF and World Bank’s assumptions: 1. High aid dependence and volatility 2. High prevalence of climatic shocks 3. High export concentration and volatility 4. High import dependence and volatility 5. Low reserves 6. Domestic revenue (excluding grants) volatility Killick and Stevens (1996) present a comprehensive assessment of the traditional debt relief mechanisms applied to low-income countries against a set of efficiency criteria in terms of adequacy, productivity, transaction cost and transparency. In almost each criterion listed the preHIPC mechanisms were assessed inefficient. They are found to be applying short-leash mechanisms for dealing with mounting debt problems, thus burdened with the inadequacy of the relief provided and the need for repeated rounds of negotiations. Despite major efforts to alleviate the debt burden, the main debt indicators deteriorated with a series of convulsions. A question has been raised repeatedly as to why the debt burdens of poor countries remain so onerous. One of the answers to this lies in the reluctance of the donor community to grapple effectively with commodity price shocks or terms of trade shocks - one of the critical factors shaping debt dynamics. Killick and Stevens (1996) have made many recommendations for raising efficiency of debt relief measures by noting, among other things: (i) provision for possible supervention of external shocks by distinguishing between temporary vs. long-lasting shocks; and (ii) more flexible mechanisms for larger and severer shocks 27 . Unfortunately, the HIPC Initiative does not imply any innovation as referred to these aspects. The external structure of SSA countries affects the consistency of the sustainability ondicators themselves. In fact, as is well-known, the debt-to-export ratio has been used for mostly middleincome Latin American countries in the aftermath of the 1982 debt crisis, whereby a substantial part was private debt and exchange rate adjustments ensured substantial trade surpluses. However, most HIPCs import not only more than they export (Cameroon and Cote d’Ivoire are two exceptions). Indeed, as was shown for example by López and Thomas (1990), Sub-Saharan African economies depend highly on imports. Trade unbalances are likely to remain for HIPCs at least over the next 10 to 20 years. 11. The role of structural and exogenous factors in the vicious circle of indebtedness What seems to be a very critical point of view on the HIPC Initiative is the fact that some comments consider it as a partial and temporary solution to a structural problem of indebtedness, which should require a systemic approach. 27 Killick and Stevens (1996), p. 147. 288 Until recently, the new growth literature was a collage of theoretical and empirical studies, many of them stressing the importance of one or a few sources of growth. Furthermore, there were some influential studies suggesting that most of the high growth experiences were due to a rapid accumulation of labour and capital (either physical capital or human capital, or both) 28 . However, over the last few years, evidence has been mounting that the accumulation of labour and capital does not explain the huge differences in growth experiences across countries. Instead, attention has been shifting towards the residual representing total factor productivity, and institutions do matter 29 . A systemic international political economic approach means to involve all the different actors who play a role and have different interests in the debt game, from both the debtor and the creditor side. Tab. 3 - The Multiple Players of "Debt Relief" Game Lenders (good and bad) Multilateral II.OO. (=I.F.I.) the World Bank, IMF and RDBs Borrowers (good and bad) HIPC (small debtors, official exposure) Regional official lenders EU The Least developed countries Bilateral official lenders G7 Other members of the Paris Club Other countries Low income countries Private lenders Commercial Banks (London Club) Private Enterprises Export Credit Agencies Middle income countries (big debtors, private exposure) Civil society (indirect:stakeholders) Savers, consumers, NGOs, poor... Civil society (indirect:stakeholders) Savers, consumers, NGOs, poor... We have a map of multiple interests and preferences (political, strategic, economic, environmental, etc.), each of one having different time horizons (long-term or short-term horizons). The relationship between all the involved interests is thus of crucial importance. In concrete terms, it is clear the responsibility of LDCs’ government, who should represent the national interest. Corruption is one of the reasons why loans don’t generate the necessary resources to repay the debts. Many civil society organisations, particularly organisations from the South, point out that the debts that piled up under the rule of corruptive dictatorial and undemocratic governments should be considered ‘illegitimate debts’. The people living in these countries now bear the burden of these debts. For example, Nigeria’s former military regime, which is to a large extent responsible for the country’s enormous debt burden, has transferred substantial sums of money to foreign bank accounts. It has thus far not been possible to transfer this ‘stolen capital’ back to the country. 28 See Krugman (1994) and Young (1995). 29 For example, see Easterly and Levine (2000) and Senhadji (2000). Rama (1993) provides a good review of empirical investment function specifications for developing countries. The specification here broadly follows that of Gunter (1998) and Oshikoya (1994). 289 Capital flight is another crucial phenomenon to be addressed. Most analysts have also attributed sluggish growth and persistent balance of payments deficits in most developing countries including Nigeria 30 , despite private transfers and long term capital inflows, to capital31 . Capital flight is a product of natural and economically rational behaviour of wealthy residents of these debtor countries to diversify their portfolios in order to protect themselves against riskiness of any one particular investment. The immediate consequence of capital flight is to reduce foreign exchange reserves, which may, in turn, require increase external borrowing to finance development expenditure. Capital flight also has some income distributional consequences: it would not be easy to tax flight capital since such capital is normally not reported to the tax authorities32 . The basic problem of two/three gaps to be closed seems to persist, particularly when we are referring to SSA countries. a) Saving-investment (S-I), fiscal and foreign exchange gaps were all persistently large in the period 1980-1998 and in part had been widening over time. b) Net capital transfers and grants filling these gaps were generally declining, highly volatile and grossly insufficient for initiating a self-sustainable investment-growth saving cycle. c) External shocks, particularly in the form of persistently declining terms of trade of HIPCs depending on the export proceeds from a small number of primary commodities, make a sustainable accumulation process very difficult. d) As a result, external debt stocks of HIPCs had been rapidly rising over time, with large shares of new disbursements leaving the debtor countries under the guise of debt service on accumulated external debt. Accumulation of arrears, debt rescheduling and debt forgiveness had so far been inadequate for reducing accumulated debt stocks or making debtors’ position sustainable. The S-I, fiscal and foreign exchange gaps were filled by the recourse to external finance, showing the evolution of current account deficits including current income transfers and grants, S-I gaps and net resource transfers including grants. At least four common features regarding external resource flows apply to all SSA ountries. These are as follows. i) Current accounts, as defined here, were persistently negative, and there is any evidence that neither foreign exchange gap nor S/I-gap are narrowing over time. Current net resource transfers, almost exclusively from official creditors, and grants have been covering the ensuing capital shortages. ii) Official grant flows, the main source of external finance to the HIPCs, were declining significantly during the nineties after a temporary surge in the early years of that decade. iii) As compared to official grants, net transfers payments played a minor role, reflecting the fact that large shares of new disbursements were made to HIPCs for enabling repayment of existing debt, rather than for capital accumulation or as a cushion against external shocks. This fact is further underlined by the more detailed debt-profiles analysis given below. 30 E. A. Onwioduokit (2001), Capital Flight From Nigeria: An Empirical Re-Examination, West African Monetary Institute, Accra, Ghana. 31 See: Chang, P. H. and R. E. Cumby (1991), "Capital flight in sub-Saharan African countries", in J. Underwood (ed) Africa’s External Finance in the 1990s, The World Bank, Washington; Cuddington, J. T. (1986), "Capital Flight: Estimates Issues and Explanations", Princeton Studies on International Finance No. 58 Princeton, New Jersey; Khan, M and N. Ul Hague (1987), "Capital flight from developing countries", Finance and Development Vol. 24 (1). 32 Lessard, D. R and J. Williamson (1987), Capital Flight: The problem and policy responses,Institute for International Economics, Washington, D.C. 290 iv) The high degree of volatility of net transfers and grants, suggesting a high level of unpredictability of countries’ development. In summary, on all three fronts, gaps were persistent and mostly widening over time, while capital inflows were channelled towards servicing of debt rather than capital accumulation. Debt relief, without taking in consideration the implementation problems, is not able to solve these main constraints. The IMF(1999a) attributes the high external debt burden of HIPC countries to “a combination of factors including imprudent external debt management polices, lack of perseverance in structural and economic reform, deterioration in their terms of trade and poor governance” 33 . However, it is acknowledged that there may be culpability on the creditor side too. The IMF (1999b) implies that the behaviour of export credit agencies of developed market economies may be another factor as their lending activities were highly risky and there was a real possibility that eventually much of the debt would not be paid. Lending behaviour by creditor governments is also implied to be a factor as it was frequently driven by domestic economic factors – the need to protect or create employment and boost exports - and the desire to strengthen diplomatic relations with borrower countries (GAO, 2000) though this only holds true where aid is tied to purchases of the creditor country’s goods and services. A number of exogenous factors also contributed to the debt problems of HIPCs. Adverse terms of trade shocks and political factors such as war and political instability were also significant factors in a number of HIPCs. Furthermore, in contrast to the IMF/World Bank’s stress on the inadequacy of adjustment effort by developing country governments, some argue that aid itself has compounded the difficulties developing countries have had in adjusting to the global economic conditions of the 1980s and 1990s and attaining growth34 . At Eurodad’s 1998 Annual Conference in Rome, participants discussed a set of conditions for the monitoring of borrowing and lending. The conditions are rather simple, and needs to be refined, but it is a good starting point. Key assumptions are: (1) governments should not borrow from any source without the authority from parliament, and (2) loans should have a “productivity conditionality” attached to them. Loans should be used for productive activities, which generate sufficient resources to pay off the loan. 12. The reliability of official data Looking at the data on the distribution of enhanced HIPC debt relief in among the decision point countries prompts two comments. First, there was a concentration in the distribution of debt relief in absolute value. Four countries, Nicaragua, Zambia, Tanzania and Mozambique, account for approximately fifty per cent of the total relief in net present value terms. Second, in relative terms the shares of Guinea-Bissau, São Tomé & Príncipe, Zambia and Guyana were larger when measured on the basis of total savings accruing to each country as a percentage of their national output. Moreover, there are three main criticisms of the results of the HIPC Initiative, three complaints at the Decision Point Documents, which are the basis of the statistical evidence of all the analyses: 1. Tables often lack internal consistency. 2. Summary statistics presented are not transparent (e.g. it is often not clear how ratios and growth rates are calculated). 33 IMF(1999a), page 2 34 El Badawi (1999) provides an overview of these arguments. 291 3. The Decision Point Documents do not follow the same format – contents and the order in which they are presented vary format, and so become somewhat inaccessible (e.g. debt sustainability analysis, use of IMF included and IMF excluded debt service figures). And presentation of statistics on debt relief is often a great deal of rhetoric. As regards the cancellation of foreign debt, the G-7 (responsible for financial matters) in Genoa did no more than look over the progress made in existing initiatives, without taking any further action. The final document underlined that the HIPC is a valid contribution to the fight against poverty. In terms of statistics, the final Genoa report stress that, on the basis of the 1999 Cologne accords, 23 countries have qualified for the initiative’s benefits: with an initial debt of $74 billion, they are eligible for a reduction of over $53 billion, that is 71.6 percent. Negotiations with five more countries are underway, while nine others would be eligible but are at war. Yet, on the basis of the latest World Bank and IMF reports on the HIPC, we can proceed in a more detailed data analysis, which provides different figures. The $53 billion figure is the nominal value of all existing mechanisms in place – traditional initiatives, HIPC, bilateral initiatives, recent agreements with the Paris Club – not just the HIPC. And that if only the HIPC is calculated, the reduction only comes to $34 billion (equals to 45.9 percent). If the actual market value of the debt is used instead of the nominal value, that figure further reduces to $21 billion or 39.6 percent – 32 points less than indicated in the G-8 final document. Another matter to be cleared up is that the HIPC reduction of foreign debt stock is only granted to countries that manage to reach the “completion point”, before it the reduction is only promised. By the end of June 2001, only Uganda and Bolivia had managed to go through the long process to the completion point from the time the initiative was re-launched at the Cologne summit in 1999. Thus the total real reduction for the moment has been less than $13 billion, that is 17.6 percent of the nominal stock, equal to 15.3 percent in concrete terms. Tab. 4 - Different methods of calculation of debt reduction linked to HIPC Initiative (US$ billion) All the initiatives (nominal value) only the HIPC (nominal value) All the initiatives (net present value) only the HIPC (net present value) only the HIPC, real data (nominal value) only the HIPC, real data (net present value) Debt reduction 53 34 33 21 13 8,1 Total stock 74 74 53 53 74 53 Reduction (%) 71.6 45.9 62.3 39.6 17.6 15.3 Thus, compared to the “official” figure, presented in the final communiqué of the Genoa G8 summit (July, 2001), the HIPC reduction is much lower. Moreover, if we consider the outstanding debt of all the HIPC countries (more than US$200 billion), given that only Uganda and Bolivia have really reached their completion point (23 countries reached the decision point), it implies that no more than 6.2 percent of total debt of all the HIPC countries has been cancelled thanks to HIPC Initiative. Thus, we have provided a very different picture of the same current data and the estimates for the coming years are even less encouraging. The 23 countries admitted at the end of 2000 will continue to pay $2 billion annually to their creditors (on the average this is more than they spend on health – $1.35 annually). Zambia, after qualifying for HIPC increased its debt servicing by approximately 23 percent, Niger by 32 percent: partial and slow mechanisms for reducing stock of debt can not interrupt the vicious circle of debt service, which tends to increase as time passes. Sub-Saharan Africa spends $14.2 billion to repay its debt – considerably more than the $7-10 billion the UN considers necessary to combat AIDS. 292 13. Concluding remarks The HIPC Initiative was criticised by a wide range of organisations including UNDP, Oxfam, Christian Aid, Drop the Debt, Jubilee 2000, Eurodad and Cafod. Among them a consensus has emerged on three major aspects of HIPC. First, the original HIPC initiative offered too little to too few and too late. The most important factors for this were the criteria for debt sustainability and the conditionality accompanying debt relief. The thresholds for debt sustainability were especially problematic. Second, the HIPC lacked an explicit linkage with human development and poverty reduction. Third, resources did not adequately support the initiative to fulfil its promises. Generation of sufficient resources for HIPC initiative has been closely linked to the differing degrees of political support from donor countries. Modification to the original HIPC initiative came in 1999 as a result of increasing pressures from debt-campaign groups. Even though the passage from the original HIPC to the enhanced initiative has been well appreciated from outside, as an encouraging step to answer to some of the main critical comments, there remain many problems with the enhanced HIPC Initiative itself. First of all, evidence is once again mounting that even the enhanced HIPC framework does not provide long-term debt sustainability for many of the poorest countries, mainly because: (i) its growth assumptions are considered too optimistic 35 , (ii) its debt sustainability analysis is still inappropriate, and (iii) its country selection is too narrow. More radical critical comments come from the Jubilee 2000 Campaign, asking for a 100 percent cancellation of debt. The World Bank and the IMF typically argue that a write-off would undermine the ability to lend to the poorest countries. The capital base of the institutions would be reduced, and it is unlikely that governments of developed countries would increase their contributions to replace the debt cancellation. With the IBRD’s equity leveraged at ‘about 5:1’, ‘its capacity to lend would be reduced by $5 for every $1 distributed to debt relief’ (World Bank & IMF 2001, p. 5). In other words, a debt write-off for poor countries would be bad for poor countries. This is not a convincing argument. If it were, it would argue against any debt relief, or for as little as possible. The paper makes it clear that 100 percent debt relief would not be in the self-interest of the IMF: ‘Debt cancellation would…impair the Fund’s financial integrity’. A similar interest is stated for the World Bank: ‘it is likely that the write-off would result in a weaker equity capital position for the Bank…’ (p. 5). Apart from this specific radical comments 36 , there are different kinds of criticism, which comes from various sides: from within the organizations (the World Bank and the IMF), from the “insider stakeholders” (the major members states) and from outside (the “external actors” represented by the international civil society and NGOs). For example, in Spring 2000, the United States General Accounting Office concluded that the HIPC Initiative might not provide a lasting exit from debt problems, unless strong and sustained economic 35 As population grows at 3% per year, it must be assumed a target growth rate of the economy of 6%, in order to have an increase of 3% in the GDP per capita. It requires that SSA countries accumulate capital and invest. The estimates of the World Bank take an Incremental Capital Output Ratio, which implies that these countries should invest every year 24% of their GDP, which is substantially higher than the 16% which is the average I/Y for these countries in 1998. 36 Other critical comments say that, once the decision point has been reached, another way to accelerate HIPC debt relief would be to front-load debt relief, which is not receveing any priority in the enhanced HIPC Initiative. 293 growth is achieved. The report cautions that the growth assumptions used by IMF and World Bank staff for the country-specific debt sustainability analyses may be overly optimistic 37 . The United States General Accounting Office report concluded (p.13) that “unless strong, sustained economic growth is achieved, the initiative is not likely to provide a lasting exit from debt problems” 38 . Like others, it cautions that the growth assumptions used in the country-specific debt sustainability analyses (DSAs) may be overly optimistic. For example, IMF and World Bank assume that export earnings will grow in excess of 9% every year for 20 years in four of the seven HIPCs the GAO analyzed (GAO 2000, p. 9). Too optimistic growth rates affect the debt sustainability in two ways: first, they imply too optimistic growth rates of a country’s exports, and second, they underestimate a country’s future financing needs. Given that the HIPC framework defines debt sustainability largely by a debt-to-export ratio, overestimations of exports (which is in the denominator of the ratio) and underestimations of future financing needs/new debt (which is in the nominator of the ratio) result in highly unrealistic low future debt-to-export ratios, which then indicate unrealistic long-term debt sustainability. A detailed analysis of capital flows, structural transformation, and investment and savings rates of HIPCs show that there is little macroeconomic foundation for the high growth projections of HIPC DSAs. On April 2001, the IMF and World Bank have recently issued a paper on the challenge of maintaining long-term debt sustainability. The paper emphasises the importance of establishing an environment conducive to growth and poverty reduction, particularly in the areas of macroeconomic policies, structural reforms, public sector management, governance and social inclusion. It also notes that HIPCs are typically dependent upon a narrow export base, which makes them vulnerable to externally induced shocks. It examines the sensitivity of long-term debt sustainability to possible shortfalls in export revenues and less concessional financing than assumed in the debt sustainability analyses, yet, the paper does not assess the likelihood of these and other factors influencing growth prospects 39 . 37 United States General Accounting Office (GAO), Developing Countries: Debt Relief Initiative for Poor Countries Facing Challenges, Washington, DC: United States General Accounting Office. 38 Even the World Bank’s Global Development Finance (2001, p. 102) has cautioned that the projected growth rates may not be realistic. 39 IMF and World Bank (2001), The Challenge of Maintaining Long-Term External Debt Sustainability, Progress Report presented to the Development Committee and the International Monetary and Financial Committee (IMFC), Washington DC, April. 294 However, looking at the long-term trends (1990-99) as well as recent trends (1997-99) of the investment and savings ratios of Guinea-Bissau, Madagascar, and Rwanda, indicate that their HIPC DSA growth assumptions of more than 6 percent are highly optimistic. Guinea-Bissau’s investment and savings ratios were lower in 1999 than in 1997, and their 1997 levels were either the same or lower than in 1990. The great vulnerability of SSA countries’ dependence on external interests for their export earning performance is another main issue. For example, multinational enterprises own close to 90% of Guinea’s exports and use most of the foreign exchange earnings for imports of equipment, salaries of expatriate workers, and transfers of profits. In any case, considering world history, any long-term real GDP growth rate of more than 6 percent is highly exceptional. It seems unlikely that Mauritania, Guinea-Bissau, Madagascar, and Rwanda will repeat what has been known as the East Asian miracle. The average growth rates for 2000-10, assumed for the first 22 enhanced decisions point countries, are 5.5 percent for real GDP and 8.6 percent for exports (expressed in nominal US dollars, ranging from 4.4 percent for Guyana to 13.7 percent for Rwanda). There is an implicit comparison being made with Latin America in the 1980s, where debt appeared to be the major obstacle to resumed growth – and indeed, following the Brady Plan, private sector inflows re-emerged and growth did resume. But the situation in SSA low-income countries is a very different one. In the first place, most of their debt and resource inflows are from the public sector, multilateral or bilateral – approximately 37 per cent of total external debt of poor countries is multilateral, 48 per cent bilateral, and only 15 per cent private. In the Brady Plan, external public 295 funds provided the private sector with relief, revived the confidence of that sector, and led to resume private inflows. With HIPC, however, it is public debt, not private, that is being written off. Over the last twenty-five years, SSA export earnings have increased only at the pace of population growth (2.9% p.a.), greatly handicapping growth possibilities. This failure is partly due to slow growth in the volume of exports, but deterioration in the terms of trade has played a major role. Cumulative terms of trade losses for Sub-Saharan African non-oil exporters (excluding South Africa) amounted to nearly 120% of average GDP, 1970-1997. Over the same period, aid transfers amounted to 178% of GDP, but the increase after 1970-73 was little more than the terms of trade loss. Behind the poor performance lies a failure to diversify exports out of primary products, a failure, which has not been addressed at all effectively by successive adjustment programmes. Average net transfers as a proportion of GDP for Africa as a whole have been about 12 percent, representing half of all government revenue and most of all public investment. Despite high levels of lending and grant programs, average GDP per capita at constant prices is lower in 2000 than it was in 1960, and the number and proportion of poor people have increased. In sub-Saharan Africa, 40 percent of its population of 600 million today live on less than $1 a day. Civil war has been another cause and consequence of many African countries’ poor development achievements. Between 1960 and 1995, almost half the countries in the region were in conflict, and an estimated 1.5% of the 1995 population is estimated to have died directly or as a result of warrelated famine. Thus, the SSA problem is not debt by itself, nor insufficient flows of externally unearned resources, but lies elsewhere, in the failure of countries to earn foreign exchange, and in deep structural economic and political problems that are not addressed by HIPC Initiative 40 . From this point of view, the main risk of the HIPC Initiative is not the inadequate implementation, but the illusion that has been created that it will provide a major solution to poverty in poor indebted African countries. And, regardless of all slogans, catchwords and ideological disputes, today, also thanks to the mass mobilisation of international civil society it has been widely proven that globalisation is not a zerosum game: there are winners and losers, between classes and social groups within individual countries, within the same productive world, between countries and continents, between debtors and lenders. 40 G. Ranis and F. Stewart (2001), HIPC: Good news for the poor?, Paper for a WIDER Conference on Debt Relief, Helsinki, August. 296 Bibliography A. Geda (2001), Debt Issues in Africa: Thinking Beyond the HIPC Initiative to Solving Structural Problems, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Ajaya S. Ibi and Mohsin S. Khan (2000), External Debt and Capital Flight in Sub-Saharan Africa, Washington, DC: International Monetary Fund. L. Abrego and D. C. Ross (2001), Debt Relief Under the HIPC Initiative: Context and Outlook for Debt Sustainability and Resource Flows, International Monetary Fund, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Alami R. (2001), ‘Evaluating the HIPC Initiative: Draft 3’, Mimeo. Andrews.D, Boote.A.R, Rizavi.S.S and Singh.S (1999), “Debt Relief for Low-Income Countries: The Enhanced HIPC Initiative.”, IMF Pamphlet Series No.51, International Monetary Fund, Washington D.C. B. G. Gunter (2001), What’s Wrong with the HIPC Initiative and What’s Next?, mimeo. Berthélemy, Jean-Claude and Ludvig Söderling (2001), “The Role of Capital Accumulation, Adjustment and Structural Change for Economic Take-Off: Empirical Evidence from African Growth Episodes, World Development, Vol. 29, No. 2, February, pp. 323-343. Bleaney. M. and Greenaway. D. (2001), “The impact of terms trade and real exchange rate volatility on investment and growth in sub-Saharan Africa”, Journal of Development Economics, Vol. 64, Elsevier Science Ltd. Amsterdam, pp. 491-500. Block. S. (2001), “Does Africa grow differently?”,. Journal of Development Economics, Vol. 64, Elsevier Science Ltd. Amsterdam, pp. 443-467. Booker S. (2000), “The Myth of HIPC debt relief’. Daily Mail & Guardian. The Mail&Guardian. London, 12th December. Boote A.R. , Thugge K. (1997), “Debt Relief for Low-Income Countries and the HIPC Initiative”, IMF Working Paper, International Monetary Fund, Washington D.C. Borensztein, Eduardo (1990) "Debt Overhang, Credit Rationing, and Investment", Journal of Development Economics, Vol. 32, No. 2, pp. 315-35. Bread for the World (1999), “The World Bank / IMF Debt Relief Initiative: Does it Lower the Debt Service Payments of Poor Countries?”, Bread for the World, Debt and Development Dossier, Issue 1. Bread for the World (2001), “What Good Can Debt Relief and PRSP Do? The Case of Zambia”, Bread for the World, Debt and Development Dossier, Issue 5. Bretton Woods Committee (2000), Concluding Report of a Roundtable Discussion on Reassessing Debt Relief, November. CAFOD (2000), “PRS – Poverty Reduction or Public Relations Strategies”, September. Chang, P. H. and R. E. Cumby (1991), "Capital flight in sub-Saharan African countries" in J. Underwood (eds), Africa’s External Finance in the 1990s, The World Bank, Washington. Cheru, Fantu (2001) “The Highly Indebted Poor Countries (HIPC) Initiative: A Human Rights Assessment of the Poverty Reduction Strategy Papers (PRSP)”, New York: United Nations, Economic and Social Council, Commission on Human Rights. Claessens S., Detragiache E., Kanbur R. and Wickham P. (1997), “HIPCs’ Debt Review of the Issues”, Journal of African Economies, Vol 6, No. 2, Oxford University Press, Oxford, pp. 231-254. 297 Cohen, Daniel (1993) ”Low Investment and Large LDC Debt in the 1980s”, American Economic Review, Vol. 83, No. 3, pp. 437-49. Cohen, Daniel (1996), “The Sustainability of African Debt,” World Bank, Policy Research Working Paper No. 1621, July. Cuddington, J. T. (1986), "Capital Flight: Estimates Issues and Explanations", Princeton Studies on International Finance No. 58 Princeton, New Jersey. Culpeper, Roy (2001) “Capital Volatility and Long-Term Financing for the Poorest Countries.” London: International Development Committee (IDC). H. Dagdeviren and J. Weeks (2001), How much poverty could HIPC reduce?, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Daseking C. and Powell R. (1999), “From Toronto Terms to the HIPC Initiative: A Brief History of Debt Relief for Low Income Countries”, IMF Working Paper 99/142, International Monetary Fund, Washington D.C. Daseking, Christina and Robert Powell (1999) “From Toronto Terms to the HIPC Initiative: a Brief History of Debt Relief for Low-income Countries,” Washington, DC: IMF Working Paper, WP/99/142. DFI (2001), “Debt Relief and Debt Humps”, Report to DFID by Development Finance International, London. Dijkstra A.G. and Van Donge J.K. (2001), “What Does the ‘Show Case’ Show? Evidence of and Lessons from Adjustment in Uganda”, World Development, Vol 29, No.5, Pergamon-Elsevier Science Ltd., Oxford, pp. 841-863. Dittus, Peter (1989) “The Budgetary Dimension of the Debt Crisis in Low-Income Sub-Saharan Countries,” Journal of Institutional and Theoretical Economics, Vol. 145, No. 2 pp. 358-366. Dittus, Peter (1989), “The Budgetary Dimension of the Debt Crisis in Low-Income Sub-Saharan Countries,” Journal of Institutional and Theoretical Economics, Vol. 145, No. 2 (June 1989), pp. 358-366. Drop the Debt (2001), “Debt relief problems require urgent solutions, say indebted country finance ministers meeting in London, 5 June 2001”, Drop the Debt. Drop the Debt (2001/b), “Finance Ministers Tread On Dangerous Ground In Palermo As They Fail To Address ‘Unfinished Business’ Of Deeper Debt Cancellation. Drop the Debt response”, London, mimeo. Drop the Debt (2001/b), “No Guarantee Against Future Debt Problems”, Debt Briefing Note, No. 23, April. Drop the Debt (2001/c), “HIPC Initiative Offers “No Guarantee Against Future Debt Problems”, World Bank & IMF Admit: A Drop the Debt briefing note’, London, mimeo. E. A. Onwioduokit (2001), Capital Flight From Nigeria: An Empirical Re-Examination, West African Monetary Institute, Accra, Ghana. Easterly, W. (2001), “How Did Highly Indebted Countries become Highly Indebted? Reviewing Two Decades of Debt Relief”, Working Paper 2225, Washington DC, World Bank. Easterly, William (1999) “How Did Highly Indebted Poor Countries Become Highly Indebted” Reviewing Two Decades of Debt Relief,” World Bank Policy Research Working Paper 2346, Washington, DC. 298 Easterly, W. and R. Levine (1997), “Africa’s Growth Tragedy: Policies and Ethnic Divisions”, Quarterly Journal of Economics, CXII (4), 1203-1250. Elbadawi A.I., Ndulu B.J. and Ndung’u N. (1997), “Debt Overhang and Economic Growth in SubSaharan Africa”, Chapter 5 in Iqbl Z. and Kanbur R. (Eds), External finance for low-income countries, IMF Institute, Washington D.C. Elbadawi, Ibrahim A., Ndulu, Benno J. and Ndung’u, Njuguna (1997) “Debt Overhang and Economic Growth in Sub-Saharan Africa,” in Z. Iqbal and R. Kanbur (eds.), External Finance for Low-Income Countries, Washington, DC: International Monetary Fund. EURODAD (2000), “What’s wrong with the HIPC initiative?”, Bruxelles. EURODAD (2001) “Debt Reduction for Poverty Eradication in the Least Developed Countries: Analysis and Recommendation on LDC Debt,” Brussels: EURODAD. EURODAD (2001), “Debt Reduction for Poverty Eradication in the Least Developed Countries: Analysis and Recommendation on LDC Debt,” Brussels: EURODAD. EURODAD (2001), “EURODAD Debt and HIPC Initiative Update: Spring Meetings 2001”, Bruxelles. Evans, H. (1999), “Debt Relief for the Poorest Countries: Why did it take so long”, Development Policy Review, Vol. 17, pp. 267-279. Evans, Huw (1999) “Debt Relief for the Poorest Countries: Why Did it Take so Long?”, Development Policy Review, Vol. 17, No. 3, pp. 267-79. Friedman, Eric A., “Debt Relief in 1999: Only One Step on a Long Journey,” Yale Human Rights and Development Law Journal, Vol. 3 (2000), pp. 191-220. Geithner T (2001), “Delivering development after Cologne: Fashionable advocacy of debt relief as an end in itself should give way to a broader strategy for growth”, Financial Times, 9th March. Grusky, S. (2000), “Poverty Reduction Strategy Papers: An Initial NGO Assessment”, produced by Jubilee 2000: What do NGOs have to say about Poverty Reduction Strategy Papers. Gunter, Bernhard G. (1998), “Economic Structure and Investment Under Uncertainty”, unpublished dissertation, Washington, DC: American University, Department of Economics, May. Gunter, Bernhard G. (2001) “Does the HIPC Initiative Achieve its Goal of Debt Sustainability?” Paper presented at the WIDER/UNU Conference on Debt Relief, Helsinki: WIDER. Hanlon, J. (2000), “How much Debt must be Cancelled?”, Journal of International Development, No. 12, pp. 877-901. Hanmer L., Pyatt G. (1994), “Developing Country Debt and Exports: The prospects for growth in the 1990s”, Institute of Social Studies, The Hague. L. Hanmer and R. Shelton (2001), Sustainable Debt: What has HIPC Delivered?, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Harck S. (2000), “On the sustainability area as a simplifying didactic device”, Cambridge Journal of Economics, Vol. 24 (4), Oxford University Press, Oxford, pp 505-509. Hersel, Philip (1998) “The London Agreement of 1953 on German External Debt: Lessons for the HIPC-Initiative,” in EURODAD, Taking Stock of Debt Creditor Policy in the Face of Debtor Poverty, Brussels: EURODAD. Hersel, Philip (1998), “The London Agreement of 1953 on German External Debt: Lessons for the HIPC-Initiative”, in EURODAD, Taking Stock of Debt Creditor Policy in the Face of Debtor Poverty, Brussels: EURODAD. 299 P. Hjertholm (2001), Debt Relief And The Rule of Thumb: Analytical History of Hipc Debt Sustainability Targets, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Hjertholm, Peter (1999), “Analytical History of Heavily Indebted Poor Country (HIPC) Debt Sustainability Targets”, Copenhagen: Development Economics Research Group (DERG), Institute of Economics, University of Copenhagen, March. Hoffmiaster A.S., Roldos J.E. and Wickham P. (1998), “Macroeconomic Fluctuations in SubSaharan Africa”, IMF Staff Papers, Vol. 45, No. 5,. International Monetary Fund, Washington D.C. International Monetary Fund (IMF) and World Bank (2001) “The Challenge of Maintaining LongTerm External Debt Sustainability,” Progress Report, Washington, DC: IMF and World Bank. Iowusu K., Garrett J. and Croft S. (2000), “Eye of the Needle: The Africa debt report (a country by country analysis)”, Jubilee 2000 Coalition, England. Jubilee 2000 Coalition (2000), “Twenty-Two countries due to gain some debt relief in 2000, but cancellation must go deeper”, Jubliee 2000 Coalition, England. Kaminsky G.L. and Pereira A. (1996), “The debt crisis: lessons of the 1980s for the 1990s”, Journal of Development Economics, Vol 50, Elsevier Science Ltd., Amsterdam, pp 1-24. Khan, M and N. Ul Hague (1987), "Capital flight from developing countries", Finance and Development Vol. 24 (1). Kose M.A. and Riezman R. (2001), “Trade Shocks and macroeconomic fluctuations in Africa”, Journal of Development Economics, Vol 65, Elsevier Science Ltd., Amsterdam, pp 55-80. Krugman, Paul (1988) “Financing vs. Forgiving a Debt Overhang,” Journal of Development Economics, Vol. 29, No. 3, pp. 253-68. Krugman, Paul (1998), “Financing vs. Forgiving a Debt Overhang”, Journal of Development Economics, Vol. 29, No. 3 (November), pp. 253-68. Lancaster, Carol (2000) “Redesigning Foreign Aid,” Foreign Affairs, Vol. 79, No. 5, pp. 74-88. Lessard, D. R and J. Williamson (1987), Capital Flight: The problem and policy responses,Institute for International Economics, Washington, D.C. López, Ramón E. and Vinod Thomas (1990), “Import Dependency and Structural Adjustment in Sub-Saharan Africa”, World Bank Economic Review, Vol . 4, No.2 (May), pp. 195-207. Marshall A. and Pravda T. (2001), “The vicious circle: AIDS and third world debt. Report by the World Development Movement for the UN Special Session on HIV?AIDS 25 June 2001”, World Development Movement, London. Martin M. with Alami R. (2001), “Long-term Debt Sustainability for HIPCs: How to Respond to Shocks”, Development Finance International, London. McCarthy E. (2001), “Debt Relief in Africa: Is it Working? A Civil Society View: The Experience of Uganda, Tanzania and Mozambique”, Bread for the World, Debt and Development Dossier, Issue 6. Mistry, Percy S. (1996) Resolving Africa’s Multilateral Debt Problem, The Hague: Forum on Debt and Development. Ndikumana L. (2000), “Financial Determinants of Domestic Investment in Sub-Saharan Africa: Evidence from Panel Data”, World Development, Vol 28, No. 2, Pergamon-Elsevier Science Ltd., Oxford, pp. 381-400. 300 M. Nissanke and B. Ferrarini (2001), Debt Dynamics and Contingency Financing: Theoretical Reappraisal of the HIPC Initiative, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Northover, Henry (2000) “PRS – Poverty Reduction or Public Relations Strategies ?”, London: CAFOD, Policy Papers. Oshikoya, Temitope W. (1994), “Macroeconomic Determinants of Domestic Private Investment in Africa: An Empirical Analysis”, Economic Development and Cultural Change, Vol. 42 No. 3 (April), pp. 573-96. Oxfam International (1999), “International Submission to the HIPC Debt Review”, mimeo, April. Oxfam International (1999/b), “Debt Relief and Poverty Reduction: Meeting the Challenge”, UNICEF/OXFAM International Position Paper. Oxfam International (2000) Make Debt Relief Work: Proposals for the G7, London: Oxfam International. Oxfam International (2000), “HIPC leaves poor countries heavily in debt”, England, mimeo. Pasinetti L.(2000), “On concepts of debt sustainability: a reply to Dr Harck”, Cambridge Journal of Economics, Vol 24 (4), Oxford University Press, Oxford, pp. 511-514. R. Mills (2001), Debt Reduction For Poverty Eradication In The Least Developed Countries, Analysis and recommendations on LDC debt, EURODAD, London, July 2001. G. Ranis and F. Stewart (2001), HIPC: Good news for the poor?, Paper for a WIDER Conference on Debt Relief, Helsinki, August. Roodman, David Malin (2001) “Still Waiting for the Jubilee: Pragmatic Solutions for the Third World Debt Crisis”, Washington, DC: Worldwatch Institute, Paper No. 155. Sachs J.D. and Warner A.M. (1997), “Sources of Slow Growth in African Economies”, Journal of African Economies, Vol. 6, No. 3, Oxford University Press, Oxford, pp.335-376. Sachs, Jeffrey (1989) “The Debt Overhang of Developing Countries,” in J. de Macedo and R.Findlay,(eds.),Debt, Growth and Stabilisation: Essays in Memory of Carlos Dias Alejandro, Oxford: Blackwell. Sachs, Jeffrey (1989), “The Debt Overhang of Developing Countries”, in Jorge de Macedo and Ronald Findlay, (eds.), Debt, Growth and Stabilisation: Essays in Memory of Carlos Dias Alejandro, Oxford: Blackwell, pp. 80-102. Sachs, Jeffrey (2000), “The Charade of Debt Sustainability,” Financial Times, September 26, 2000, p. 17. Sachs, Jeffrey (2000), “The Charade of Debt Sustainability”, Financial Times, September 26, p. 17. Sachs, Jeffrey, Kwesi Botchwey, Maciej Cuchra, and Sara Sievers (1999) “Implementing Debt Relief for the HIPCs,” Cambridge: Center for International Development, Harvard University, August 1999 Sachs, Jeffrey, Kwesi Botchwey, Maciej Cuchra, and Sara Sievers (1999) Implementing Debt Relief for the HIPCs, Cambridge: Center for International Development, HarvardUniversity, August. Tarp, Finn and Peter Hjertholm (eds.) (2000) Foreign Aid and Development: Lessons Learnt and Directions for the Future, Copenhagen: Development Economic Research Group (DERG). United Nations Conference on Trade and Development (UNCTAD) (2000), The Least Developed Countries Report, New York: United Nations. 301 United Nations Development Programme (1999), Debt and Sustainable Human Development, Technical Advisory Paper, No. 4, Management Development and Governance Division, Bureau for Development Policy, May. United Nations Economic and Social Council (UNESCO) (2001), “Economic, Social and Cultural Rights – The Highly Indebted Poor Countries (HIPC) initiative: A Human Rights Assessment of Poverty Reduction Papers”, A Report by Fantu Cheru, 18 January. United States General Accounting Office (GAO), 2000, Developing Countries: Debt Relief Initiative for Poor Countries Facing Challenges, Washington, DC: United States General Accounting Office. World Bank (2001), “Financial Impact of the HIPC Initiative: First 23 Country Case”, Washington D.C., June. World Bank (2001), Global Development Finance, Washington, DC: World Bank. 302 9. Current political debate within international civil society on development, poverty reduction and foreign finance 1. Introduction Over the past few years, the G-8 has become one of the main interlocutors in the international community in discussion of strategies for international co-operation and development. The 2001 G-8 Summit, which takes place in Genoa next July is a further step in this direction. International civil society networks - NGOs, Third Sector organisations, interest groups, and local administrations have shown increasing interest in communicating with governments and international organisations, and in working together in new ways. During the 1990s, with the mobilisation and campaigning activities that started on the occasion of United Nations world conferences, the international civil society has become an increasingly powerful influence on international relations. They are academics, political activists, community leaders and representatives of nongovernmental organisations (NGOs) who combine intellectual and scientific rigour with concrete action in their societies and on an international level. As agreed at the G-8 in Okinawa (23 July 2000): «In a world of ever-intensifying globalisation, whose challenges are becoming increasingly complex, the G8 must reach out. We must engage in a new partnership with non-G8 0countries, particularly developing countries, international organisations and civil society, including the private sector and non-governmental organisations (NGOs). This partnership will bring the opportunities of the new century within reach of all.» The Italian Presidency made a further step in this direction by promoting the Genoa Non Governmental (GNG) Initiative. The GNG Initiative is a Forum to encourage discussion on global issues between the G-8 and NGOs. The GNG Initiative involved four Italian research institutes: Centro Studi di Politica Internazionale (CeSPI), Istituto di Affari Internazionali (IAI), Istituto per la Cooperazione Economica Internazionale e i Problemi dello Sviluppo (ICEPS), Istituto per le relazioni tra l'Italia e i Paesi dell'Africa, America Latina e Medio Oriente (IPALMO). The four institutes had the role of facilitating the dialogue and will start up the consultation process with representatives of international Non Governmental Organisations, in particular with those organisations from G-8 countries. The Forum facilitated consultation with international NGOs via the exchange of information and the creation of a mutually agreed proposal, which will include recommendations on political guidelines. CeSPI facilitated the consultation on the «Poverty Reduction Strategies» issues, ICEPS on «Finance for Development. Debt relief «, IAI on «International Governance and WTO reform» and IPALMO on «Environment and sustainable development»1. In our opinion, one of the most interesting and positive results of the initiative was that it arranged for representatives of the Italian G8 Presidency (of the Office of the Prime Minister and of the Ministry of Foreign Affairs and the Treasury), of various international organisations (World Bank, OECD, ILO, EU), scholars and the academic world, numerous NGOs and civil society associations (approximately 200 international non-governmental organisations and networks) to sit around the same table and be engaged in frank discussion. 1 The GNG Initiative has been co-ordinated by Marco Zupi, who has edited the final report in collaboration with Isabella Falautano (IAI), Karl Giacinti (ICEPS) and Alessandra Filippini (IPALMO). For any further information on it, see the GNG-Initiative Website: www.gnginitiative.net. 303 The analysis and policy recommendations of this chapter do not necessarily reflect the view of any particular NGO. This chapter is based on the reports resulted from the consultation with the NGOs who took part in special meetings and conferences organised by the GNGInitiative. The Reports cite and derive from many background studies prepared on thematic issues as well as analyses of experiences in countries, international technical reports and position papers by NGOs, who generously shared their materials. The work of the GNG-Initiative is an expression of the capacity of numerous organisations of civil society to engage in debate and comparison on the content, which is a reflection of the skill and know-how accumulated with regard to the issues of international relations. Undoubtedly, civil society organisations constitute an innovation on the scene of international relations. In the multifaceted world of civil society there co-exist specific and widespread interests, at local, national and transnational level. In the previous chapter, we demonstrated the role of a broad-based movement, politically expressive and capable of a more effective critical approach to the HIPC Initiative. At the same time, however, international civil society has a strong capacity in the costruens pars, that is they unquestionable prove the maturity of the proposals. This chapter is intended to furnish a comprehensive framework of the NGOs current positions and recommendations for the international political agenda related to debt relief. Most of NGOs agree on a "certain number" of focal points to be stressed. The sections of the present chapter, divided in three paragraphs, correspond to the main areas of political debate. However many of the issues covered in this chapter are interrelated and crosscutting. There should be a clear understanding that any issue could be addressed in its diverse interrelationships at any given point, being an integral part of an interrelated whole. Paragraph 2 will consider the proposals of NGOs for addressing the challenge of poverty reduction, paragraph 3 will explore innovative sources for financing for development including taxation and market instruments. Paragraph 4 will examine the specific issue of Debt Relief, which logically can be included in the general measures for the financing of development. From the NGOs point of view, the positions concerning debt relief/cancellation are on the top of the list for their urgency. The NGOs contributions and background papers are quoted in the bibliography annex. 2. Poverty Reduction Strategies 2.1 - G8 can play a positive role to reduce poverty One person in five across the world lives on less than $1 a day. The negative effects of the uneven distribution of the costs and benefits of the process of globalisation and liberalisation are evident. The number of people living in very bad conditions is increasing in various regions of the world. The poorest countries are failing to catch up with developed and other developing countries, and some are getting stuck in vicious circles of economic stagnation and regression. For these reasons, the commitment to halve world poverty by 2015, embodied in the international development goals, is the shared priority of international development cooperation. Development NGOs share the idea that a global poverty reduction strategy clearly needs to be mounted - with more resources, a sharper focus and a stronger commitment – only as a midterm strategy, in order to achieve the final end of poverty eradication. It is justice, not charity. Countries are free to gather and consult. So are the G8. The G8 account for the bulk of international trade, services industrial production, foreign direct investment, financial flows, 304 information and technology flows, and ODA efforts in the world. The G8 determine to a great extent the general orientations and dynamics of international affairs. However whether internationally or not, the G8 exercise, when they meet, de facto governance and leadership. Thus, their summit matters. The governance of globalisation mechanisms and institutions which have been established officially by the international community to deal with these issues are often insufficiently invested with the necessary authority or endowed with adequate resources. They are also often not conceived or organised or able to respond quickly and effectively to crises and emergencies. Thus, the role of G8 can be crucial, because of their political, economic, cultural power and influence all over the world. But G8 are neither the poor, who are the main subjects of poverty reduction strategies, nor the developing countries, who are the nations directly involved in designing and implementing strategies for poverty reduction, nor the international organisations, who serve as the multilateral agencies to provide technical and financial assistance, nor the international NGOs and private sector, who are important partners to support progress in development. NGOs consider important that G8 assume their own responsibilities, based on few and precise commitments, rather than spending time and words on useless rhetoric and ineffective spot actions. An effective co-ordination means that each actor should play its own role, avoiding any risk of over-commitment, contradiction or duplication. NGOs hope that G8 meetings resolutions and action plans prepare and consolidate and not pre-empt and contradict the meetings and action plans of the accountable representatives of the UN institutions, with due regard to existing international rules and representative institutions of the global community. The NGOs stress that the G8 should meet the commitments pledged in occasion of the UN Conferences held in the '90s, in the presence of the world public opinion. Basically, G8 should commit on some political engagement to be fulfilled and easily measured, in order to improve their performance and to serve as a benchmark and model for other rich countries. 2.2 - G8 commitment on adhering to the International Co-operation Consensus, a shared conceptual framework on poverty reduction All the member states of OECD’s Development Assistance Committee (DAC) have declared to share a broad understanding of poverty and its many dimensions. This International Cooperation Consensus emerged during the last ten years, affirming that poverty encompasses different dimensions of deprivation that relate to human capabilities including consumption and food security, health, education, rights, power, security, dignity and decent work. NGOs stress the importance of recognising the multi-dimensional nature of poverty, and the need to operationalise it in terms of a framework of action to reduce poverty. It implies that poverty reduction strategies should be strongly based on the high differentiation in the evolution of poverty across dimensions, regions, communities, households and individuals. Context does matter, poverty has a geography, because it is present in some places but not others. Nevertheless, there are some key principles that highlight the operative priorities. Empowerment. Poverty is not the outcome of economic processes alone. The interaction of social, political and economic processes determine poverty and power is a critical determinant. Poverty derives from a situation in which wealth is concentrated in few hands. Poor can escape from poverty if they assume the responsibility of their progress and find enough political space to do so. 305 Ownership. The importance of recipient «ownership» of development co-operation is widely acknowledged. Recipient governments and population can be said to own aid when it empowers them and serve their interests. Nevertheless, low ownership of aid is common. Donors tend to dominate the aid process and to pay inadequate attention to the poor’s own preferences. Re-distribution. Poverty reduction can be achieved if and only if all the actors interested in our future recognise the need of an effective chain between democracy, participation and social justice. There is an unprecedented growth in the inequality of the distribution of assets and income within and between countries. The combined wealth of the top 200 billionaires hit $1,135 billion in 1999, up from $1,042 billion in 1998; compared to the combined incomes of $146 billion for the 582 million people in all the least developed countries. More than three billion people earn less than three dollars a day. With poverty and inequality continuing to increase, there is a need for more fundamental change in economic and political priorities. The degree of inequality in society matters. Studies find that the responsiveness of income poverty to growth increases significantly, as inequality is lower. Initial levels of inequality in assets also determine the poverty impact of growth. Addressing these structural inequalities thus becomes a crucial component of poverty reduction strategy. Partnership. Currently, 49 countries, where 10.5% of world population lives, are identified as LDCs. LDCs are particularly ill equipped to develop their domestic economies and to ensure an adequate standard of living for their populations. The economic and social development of these countries represents a major challenge for themselves as well as for their development partners. The NGOs agree on the fact that the definition of poverty reduction strategies by the developing countries themselves is a fundamental prerequisite for sustainable development. The principle of partnership is supposed to encourage ownership of the development strategies by the countries and population concerned. In order to ensure that the process for achieving poverty eradication is defined by developing countries, the NGOs emphasise the key role of political dialogue and the necessity of providing support and capacity building in beneficiary countries. The dialogue should allow developing countries and G8 to address all issues of mutual concern and to ensure consistency and the increased impact of development co-operation. Given these key premises, in terms of commitments, the NGOs think that G8 should resist an ineffective, donor-driven approach and, rather, to commit themselves towards the implementation of strategies, defining their own responsibilities, in order to reach out common objectives. It methodologically implies that a pragmatic action-oriented donors’ commitment should come up with a clear and realistic action agenda based on some general considerations: (1) Men and women from developing countries are the main subjects to articulate appropriate policies to reduce poverty. Their leaders must be committed to reduce poverty, and donors should basically support these policies. Poverty reduction strategies mean to support poor in their sustainable livelihood strategies. (2) The donors’ development co-operation policies should support the diversity and originality of local institutional arrangements that encourage sustainable livelihoods and collective empowerment, rather than searching for a naïve unique road to development imposed from abroad. Access to development and a sustainable livelihood is an indivisible human right of all people. The means to preserve the conditions for a globally sustainable development of each individual and its community must be regarded as global public goods, independently whether the source of production and the reach of effects is local, national or global. The mutually 306 beneficial nurturing and caring of these goods must become the focus of the donors’ policy. (3) Donor countries’ development co-operation policies should reflect a participatory approach and openness to developing countries’ knowledge closely linked with people living in poverty in developing countries. (4) As a result of recession and cuts in public spending, women are increasingly represented among the world’s poor. Women are more likely to suffer poverty than men. Attention to women in development as a policy mainstreaming must increase, given the fundamental importance of gender to the elimination of poverty. But it should be translated in terms of a part of people’s everyday approach, rather than a checklist. Gender needs to become a matter of reflex, not a checklist item, nor an add-on late in the planning process. It means that great attention must be given to the impact of development to women, but also increased priority to interventions whose goal is poverty reduction through the enhancement of women’s position within society. (5) Donors’ development co-operation policies should promote investment in human resources so as to guarantee full ownership and leadership of each developing country in formulating effective national strategies to reduce poverty and to assure the sustainability of processes. In this sense, the promotion of local participation, the empowerment of the developing countries' civil society and their involvement in the specification of common guideline of poverty reduction strategies, is an important process to be supported. (6) The G8 countries’ concern with balance of power and with balance of development, should be reflected into transparency and innovative participative approach to poverty reduction, designed through international brainstorming, dialogue and concertation with non G8 countries, dialogue with and presence of national and international civil society. This is a consensus building mechanism, a welcome approach for establishing the agenda, analysing the issues and designing the action plans. Development in its fullest sense is at the centre of the agenda of poverty reduction. It must be considered as a main way for addressing the governance of globalisation. The process of global economic integration has altered the context in which most governments are thinking about policies for economic development. The globalisation model considers that national economic policymakers are accountable to foreign investors and international financial capital, which are the groups that determine whether an economy is judged a success or not. It means to empty political powers. On the contrary, poverty reduction strategy must be the way for poor people to engage the economy on their own terms and interests, not on terms set by global markets. 2.3 - From commitment to implementation of the 0.7 per cent ODA/GNP target Official Development Assistance (ODA) has an important role in achieving progress in the area of poverty reduction, by supporting the efforts undertaken by poor people in developing countries. The 1969 Report of the Pearson Commission, Partners in Development recommended an aid target of 0.7 per cent of GNP. This has remained an influential goal, though only reached today by few donors (none of them being a G7 member state). Denmark, Norway, Sweden and the Netherlands have shown that it is possible to reach the 0.7 per cent of GNP target, without inducing negative fiscal effects. Nevertheless, most of the 307 donors’ governments continue to say that aid flows remain restricted because of domestic fiscal concerns, but it is a mere pretext: the United States has a large fiscal surplus and the lowest level of ODA/GNP ratio in the DAC community (0.10 %). In 2000, Denmark reached its highest ODA/GNP ratio ever recorded (1.06%), the Netherlands (0.82%), Sweden (0.81%) and Norway (0.80%) continue to surpass this target, and Luxembourg reached for the first time, the UN 0.7 per cent target for ODA as a proportion of GNP. No other countries exceeded the average country effort of 0.39 per cent of GNP. Official Development Assistance by Member countries of the DAC fell slightly in 2000 - by 1.6 per cent in real terms. In current prices and exchange rates total net ODA flows declined from $56.4 billion in 1999 to $53.1 billion in 2000 - a fall of 6.0 per cent. Per capita aid to Sub-Saharan African countries declined from $32 per annum (1990) to $18 (2000). During the 1990s, the DAC ODA/GNP ratio has fallen to around 0.25%, well below the 0.33% average maintained in the 1970s and 1980s. This is some $20 billion per annum less than aid would have been, had the previous level of effort been maintained. In 2000, the DAC ODA/GNP ratio has fallen to 0.22%. In 2000, total aid from the non-G7 DAC countries increased by 8.3 per cent in real terms and accounted for 26 per cent of DAC Members’ ODA, compared with their 12 per cent share of DAC GNP. Total aid from G7 countries fell by 4.8 per cent in real terms. Tab. 1 - Net Official Development Assistance Flows Canada France Germany Italy Japan United Kingdom United States Total G7 countries Non-G7 countries Total DAC ODA US$ m. (2000) 1 722 4 221 5 034 1 368 13 062 4 458 9 581 39 446 13 612 53 058 ODA/GNP % (2000) 0.25 0.33 0.27 0.13 0.27 0.31 0.10 0.19 0.46 0.22 Real % change 1999 to 2000 -2.2 -13.9 5.9 -14.3 -17.9 35.6 2.7 -4.8 8.3 -1.6 Real % change 1998 to 1999 -1.9 2.3 2.2 -18.4 27.1 -10.6 2.5 7.3 4.3 6.5 Source: DAC-OECD, 2001 Among the G7 countries, Canada’s fell slightly, and reached its lowest level in 35 years (a considerable drop from the peak of 0.50% registered in 1988). A fall occurred in aid from France, due mainly to the fact that its aid to French Polynesia and New Caledonia is no longer counted as ODA. Germany increased its aid. Italy experienced a fall in ODA, due to the timing of its payments to multilateral agencies. An exceptional fall occurred in Japan’s aid, which was $2.3 billion lower than in 1999, when it included exceptional contributions to the Asian Development Bank in the wake of the Asian financial crisis. Aid from the United Kingdom rose by 35.6 per cent in real terms, more than compensating for the 10.6 per cent fall last year, with the rebound due to the timing of its multilateral contributions and its commitment to an increased budget for aid. Compared to 1999, the United States increased his aid, as a consequence of his effort to assist refugees from Kosovo, but the low level since 1998 is a result of Israel non-longer being eligible for ODA. G7 donors can not expect their commitment to the Development goals to be taken seriously, when they are signally failing to provide their share of the resources needed. 308 Basically, the NGOs recognise the importance of the 0.7 per cent of GNP target: an increased volume of the G7 ODA is considered an essential contribution towards achieving poverty reduction, by making the G7 commitment more reliable and its implementation possible. NGOs consensus emerged on the fact that G7 countries should increase, in a gradual and permanent way, their efforts to make additional official funds available, guided by the 0.7 target. The firmest commitment to positive long-term signs of an increase in aid is hoped, through the identification of the interim steps towards the goal. Any argument that the money is not available is not acceptable: in just one week, during December 1997, donors managed to pledge US$57 billion to bail out South Korea (Japan pledged US$10 billion, The US pledged US$5 billion). Public deficits in OECD countries have been reduced from 4.3% of combined GDP (1993) to 1.3% (1998), yet ODA continues to be cut. 2.4 - G8 commitment on objectives and measurable results to reduce poverty Globally, 1.25 billion people are in extreme consumption (or income) poverty and they live on less than one dollar a day; 3 billion live on fewer than two dollars a day. 1.3 billion have non-access to clean water. 3 billion have no access to sanitation; 2 billion have non access to electricity. Moreover, the number of income-poor in the developing world is again on the rise — estimated to exceed 1.2 billion in 1998 after having declined until 1996. The group of least developed countries contains the hard core of the problem of marginalisation in the world economy. Least developed countries account for 32 of the 35 countries in the lowest category of the UNDP’s Human Development Index. Projections of social indicators on the basis of the 1990s do not paint a bright picture either. In 2015 life expectancy would be only slightly higher than the 1998 average, and still well below the prevailing levels in other developing countries. If the prevailing trends continue to 2015 under-five infant mortality rates in the least developed countries average will be over 110 per thousand live births, double the rate in other developing countries in 1998. The gender gap in education will be almost the same as the levels in 1998, at 20 per cent below the other developing countries. Referring to these data, two dynamic aspects are particularly important. First, the outbreaks of financial crises have an immediate impact on the living conditions of the population in the affected areas. The improvement in levels of poverty achieved during years is rapidly being lost to the financial crisis: in East Asia, over 20 million people fell back into poverty in 1997. It means that poverty is more and more linked to the concept of vulnerability, exposure to risks and volatility people face in their everyday lives. In the context of globalisation, a decent standard of living, adequate nutrition, health care, education, decent work and protection against risks mean political and economic power, rights and capabilities to escape poverty. Second, global inequality has risen during the last years. During the 1990s, various international conferences agreed to set targets to motivate and provide focus for poverty reduction strategies of international development co-operation policies. Significant progress has been made at various international summits: the 1992 Earth Summit in Rio, the 1994 International Conference on Population and Development in Cairo, the 1995 World Conference on Women in Beijing, the 1995 Social Summit in Copenhagen, the 1996 Human Settlement Conference in Istanbul, the 1996 Food Summit in Rome. Subsequently, 309 the member states in the XXIV Special Session of the General Assembly of the UN (June 2000) committed themselves to halve extreme poverty by 2015. The effort that went into reaching agreement on commitments at each of these conferences must not be wasted. The G8 should honour agreed commitments. In Shaping the 21st Century: The Contribution of Development Co-operation, a number of specific targets agreed in UN Conferences were reconfirmed by DAC member countries, all the G7 countries included. More recently, in September 2000, in the Millennium Summit Declaration these seven International Development Goals (IDGs) were confirmed by the Heads of Government and State, in the context of the broader set of goals – including on hunger, safe water and HIV/AIDS – and in the context as well of the ultimate objective of poverty eradication. Tab. 2 - the International Development Goals2 1 Reduce by at least one-half the proportion of people living in extreme poverty in developing countries by 2015. (Copenhagen) 2 Universal primary education in all countries by 2015. (Jomtien, Beijing, Copenhagen) 3 Demonstrated progress toward gender equality and the empowerment of women by eliminating gender disparity in primary and secondary education by 2005. (Cairo, Beijing, Copenhagen) 4 Reduction of infant and child mortality rates by two-thirds by 2015. (Cairo) 5 Reduction of maternal mortality by three-fourths by 2015. (Cairo, Beijing) 6 Access through the primary health-care system to reproductive health services for all individuals of appropriate ages, including safe and reliable family planning methods, by 2015. (Cairo) 7 National strategies for sustainable development in operation in all countries by 2005, so as to ensure, by 2015, a reversal of current trends in the loss of environmental resources and the accumulation of hazardous substances. (Rio) Source: DAC-OECD, 1996 Among the International organisations, the World Bank (World Development Report 2000/2001), UNDP (Human Poverty Report 2000), IFAD (Rural Poverty Report 2001), DAC (2001 Policy Statement by the DAC High Level Meeting upon endorsement of the DAC Guidelines on Poverty Reduction) have all placed emphasis on taking concrete action to poverty reduction, referring to the IDGs. The IDGs include: by 2015, halving extreme income poverty, lowering infant, child and maternal mortality, and ensuring universal primary education and access to reproductive health services; and by 2005, achieving gender parity in education as a step towards gender equality and the empowerment of women, and implementing strategies for sustainable development as a step towards reversing the loss of environmental resources. The basic idea of global goals of measurable progress is to define a shared vision of what can be achieved, in order to move to the real implementation of poverty reduction strategies. And trend data point to an alarming gap between commitment and achieving – a gap that translates into millions of deaths. All these goals must represent a basis for dialogue with developing countries, which must set their own national goals in light of their particular circumstances. It is important to setting time-bound goals and targets, which imply the need of a long-term strategy (the 2015 horizon). The goals can be achieved, but progress in some countries and regions is great (China reduced its number in poverty from 360 million in 1990 to about 210 million in 1998), 2 Continue addressing qualitative aspects of development that are essential to the attainment of the aforementioned goals; these include capacity development for effective, democratic and accountable governance, the protection of human rights and respect for the rule of law (Copenhagen, Vienna). 310 whereas other countries have reversed their gains (many countries in Sub-Saharan Africa). Thus, a real success will require great attention to the differentiated results in all the different contexts, otherwise the outcome will be biased. A lot of studies reveal that meeting the targets depend critically on three factors: future real growth being higher than in the past (4% per capita per annum until 2015); policies that make the growth path more pro-poor; and low income inequality. Based on the assumption that the world does not lack the resources to reduce poverty and achieve the international development goals and that donors must demonstrate clear links between their development co-operation inputs and the IDG, the NGOs highlight the need for: (1) Poverty reduction is a mid-term strategy toward the long-term objective of poverty eradication; (2) Increased aid and debt reduction for health and education, including a global initiative on basic education; (3) Increased budget allocation for basic social services and for the least developed countries; (4) The phasing out of cost-recovery and user-fees in basic health and primary education systems, introduced by the implementation of the Structural adjustment programs supported by the World Bank and the International Monetary Fund; (5) Economic growth linked to redistribution in favor of the poor. In order to translate the above goals into reality it is essential that the developed countries and the LDCs jointly define objectively measurable indicators to assess the level of achievement of the expected objectives. Clear mechanisms to assess the achievement of the jointly defined objectives have to be designed and put in place. 2.5 - G8 commitment on the good practice in instrumental approaches to poverty reduction The proliferation of projects, programs, Trust Funds increases the risks of duplication and waste. NGOs underline the need to improve co-ordination. G8 are not required to introduce new programs or spot Trust Funds; they are not required to add themselves to more than a hundred bilateral, multilateral and NGOs’ agencies, each with its own procedures, priorities, planning cycles, and reporting requirements. But they can commit themselves to promote better co-ordination in sector wide approaches, supporting the central role played by the recipient governments to undertake the challenge of poverty reduction. Sector wide approaches can be useful frameworks. They are programs of action and reform in one main sector – health, education, and transport – led by the government. G8 governments are supposed to contribute funds to the common programme. But if large amount of money has been committed to sector wide approaches, then much less has been spent, because specific donor conditionalities have not been met. It is important that conditionality is reconciled with local ownership and capacity and that as the structural adjustment conditionalities failed to work, they can not remain a dominant part of the donors ODA regime. Focus on basic Social Services. The 20:20 initiative, first suggested in Human Development Report 1992 and agreed in the World Summit on Social Development in 1995, to allocate 20% of developing countries’ public spending and 20% of donors’ ODA budgets to basic social services, in order to mobilise the additional $70-80 billion a year needed from national and international sources to ensure basic social services for all. 311 Many studies have documented a shortfall in public spending of up to $80 billion a year to achieve universal provision of basic services, with around $206-216 billion required and only $136 billion being spent. This shortfall is twice the estimate of up to $40 billion at the time of the World Summit for Social Development in 1995. There is also serious discrimination in public spending on health and education – which is biased towards richer people, event though the needs remain greater for poorer people. The Social Summit set a target of reducing adult illiteracy by half between 1990 and 2000. But many regions are unlikely to reach this target. The same is true for child malnutrition, which was also supposed to be halved between 1990 and 2000. In South Asia the percentage of malnourished children remains high, and in Sub-Saharan Africa it has risen. Progress in life expectancy has also been slow. In Sub-Saharan Africa, many countries are moving backwards. In the least developed countries about a third of the population will not survive to age 40. The contributions of G7 countries for basic health care, basic education and water and sanitation was only 6.2% of ODA in 1999, less than the average contribution of all bilateral DAC donors (7.3%) and less than one third the 20% target of the 20:20 initiative. Tab 3 - G7 countries’ ODA spending for basic social services (% of bilateral aid, 1999) Canada France Germany Italy Japan United Kingdom United States Total G7 countries Total DAC Basic education 0.8 1.9 0.0 0.4 3.0 1.2 1.2 1.2 Basic health care 0.8 0.2 1.9 2.1 0.7 2.3 3.1 1.6 2.0 Water and sanitation 1.8 2.3 5.9 3.1 5.8 2.5 2.1 3.4 4.1 Sub-Total 3.4 2.5 9.7 5.2 6.9 7.8 6.4 6.2 7.3 Source: DAC-OECD, 2001 Thus, G7 countries are not fulfilling their side of the 20:20 initiative. The NGOs reaffirm the need to fulfil this proposal. Only in the context of this commitment, within the implementation of the 20:20 initiative, NGOs agree to set up new specific initiatives to face the dramatic challenge of HIV/AIDS, including matching funds to a Trust Fund by the developing countries. This is a positive development, especially if co-ordinated by UN agencies and coherent with global conventions, but NGOs refuse any spot or fragmented new initiative. The Education World Summit, hold in 2000 in Dakar, renewed the importance of the commitment in terms of basic education. Basic education is a human right, a motor for development and the foundation of responsible governance, which implies accountability to domestic constituencies. Changing the educational chances of children demands more resources and better co-ordinated and more equitable use of these resources, including ODA as well as domestic revenue. It also demands to support institutional reform and technical innovation; G7 countries should leverage improvements in policy, but avoiding eroding the State’s financial and political responsibility for basic education. 250 million of children are working in the world. Exclusion from school is one of the most pervasive mechanisms for reproducing inequalities. An interesting scheme, introduced in Brasilia, Brazil Federal District, in 1994, and now implemented in some developing countries which can have wider multidimensional effect is the so called «Bolsa Escola» (scholarship for poor families). It is a yearly renewed cash transfer (minimum income) program 312 conditional on primary school attendance, targeted to poorest and most vulnerable families. In addition, children have the opportunity to attend higher levels of education through savings deposits. A Report prepared by an Advisory Group brought together by Ilo and Unctad, with the participation of representatives from some African countries and from Unicef and World Bank, has assessed the desirability and feasibility of applying such a scheme also in African countries. This scheme would be a valuable and innovative mechanism for helping government to achieve international development goals and to reduce future poverty by increasing human capital and stopping children dropping out of school. Microfinance. The Microcredit Summit’s goal of reaching 100 million of the world’s poorest families by the year 2005 has not been achieved (the Summit estimated that US$ 11.6 billion would be needed as grants and soft loans, raising the percentage of ODA going to microcredit for the poorest from the current less than 1% up to 5%). Nevertheless, NGOs recommend the fulfillment of the target, because of the main importance of microfinance as a means of commanding resources (to incur debt, i.e. to gain credit), of real empowerment of poor. Information and communication technology. The global online community has grown rapidly – from about 16 million Internet users in 1995 to an estimated 304 million users in 2000. But access to the Internet varies between regions. In 1998, more than 26% of all people living in the US were connected to the Internet, compared with 0.8% of all people in Latin America and Caribbean, 0.1% in Sub-Saharan Africa and 0.04% in South Asia. NGOs think technology is central in reducing poverty, but it means not only to create an enabling sustainable environment, a structural context to widely disseminate innovations (rather than introducing spot technologies or using digital divide rhetoric), but also reviving pro-poor technologies and support locally-owned technologies. It is not simply through the provision of PC that poor have access to new technology; technical, institutional, attitudinal and policy aspects of knowledge sharing are strictly interlinked. Promotion of decent work. Globalisation poses major challenges to the world of work. New jobs opportunities can be linked to increased uncertainty and insecurity, providing only a temporary solution to the social excluded and weakening bargaining position of workers. G8 should promote job opportunities in condition of freedom, equity, security and human dignity, that is decent work for women and men, extending social protection and promoting social dialogue, encouraging productive investment, combating forced or compulsory labour, child labour and discrimination in employment or occupation. NGOs want G8 to consider work promotion as part of an integrated development strategy, a basic human right which provides people and their families with self-esteem, dignity, equity, freedom and security. This calls for the G8 to emphasise the promotion of sustainable livelihoods, self-employment, informal sector, and micro, small and medium enterprises development. Promotion of local development. A positive response to the challenge of globalisation can be the systemic paradigm, by emphasising the «meso» dimension of the socio-economic systems. The characteristics of territories and the grade of spill-over and interaction among factor of production, institutions, organisations, technologies and social actors are very important. The nature of economic accumulation can no longer identified with the value of material and immaterial goods, rather with the one of relational goods. Thus, it is important to create suitable conditions that would stimulate the domestic growth of medium, small and micro-sized enterprises in developing countries, which are the main sources of decent work. Donors can support it through the reinforcement of productive systems at the local level. The mechanism of bottom-up partnership, based on NGOs networks and decentralised cooperation can strengthen the favourable context for local development. Support to policies for pro-poor growth. Concern for the human costs of adjustment has centered on protecting social spending and targeting it more effectively to the poor. The 313 20/20 Initiative, led by UNICEF and supported by UNDP, tries to restructure social spending to promote human development more equitably. Shifting more resources to basic social services and ensuring the quality of the services will definitely be more pro-poor than the current pattern of allocation in most countries. But there is no guarantee that these efforts alone will lead the poor to use the services. A more integrated set of interventions, both economic and social, is needed. With renewed emphasis on inequality, as an impediment to reducing poverty, growth is no longer considered the single driving force for poverty reduction. The traditional thinking was that only rapid growth mattered and that changes in inequality could make only a minor difference in outcomes. But the consensus has shifted to recognising that high inequality can be an imposing obstacle to poverty reduction, nationally and internationally In some regions with high inequality, such as Sub-Saharan Africa, there is little chance of reducing income poverty by half by 2015. In many countries the poorest receive a minuscule share of national income. Economic growth cannot be accelerated enough to overcome the handicap of too much income directed to the rich. Income does not trickle down; it only circulates among elite groups. Abandon the Structural adjustment programs. The World Bank introduced structural adjustment lending in 1979, initially as a temporary measure for developing countries with balance of payments problems and/or large debt burdens. These programs encouraged a focus on economic restructuring as a prerequisite to social welfare improvements. Structural adjustment became, and continues to be, a prominent feature of World Bank lending, supported by G7 countries. The direct and indirect results of these programs have seriously debilitating effects on the poor. The new poverty, characterised by the heterogeneity of its members, since they come from different occupational and socio-cultural backgrounds, is the direct result of structural adjustment measures. Recently, recognising the harsh consequences of adjustment for the poor, the World Bank introduced Poverty Reduction Strategy Papers as the new foundation documents for aid. Whereas in the past, country strategies were written by donor agency, the theory is that PRSPs will be prepared by developing country governments: a PRSP is supposed top be developed in a participatory way, nationally owned and to set out the government’s policy framework and agenda for tackling poverty. Donors should use PRSPs as the basis for decisions on funding. Nevertheless, NGOs’ critics continue to lament the lack of real progress towards new poverty reduction approaches, as the PRSPs centre on the continuos separation between microeconomics (put in the hand of national governments, supported by the UN agencies) and macroeconomic prescriptions (dominated by the Bretton Woods Institutions). In practice, the old structural adjustment conditionality based mindset is all too evident. Ownership is often nominal and civil society participation easily ignored: country strategies are still written and owned mainly by the Bretton Woods institutions. Like its predecessor, the New Washington Consensus seeks to make economies adjust to the distorted world market rather than to challenge the global wealth and power imbalances. The IMF and the World Bank note that their Boards will «endorse» a country PRSP, but this endorsement amounts to a veto on national approaches to poverty reduction and development. Rather than introducing new rhetoric or spot initiatives, G8 should make much progress on translating into operative sectoral approach the International Co-operation Consensus, on concentration, co-ordinating the accounting and reporting requirements, reduced aid tying or on transparent and predictable programs of ODA disbursement – all of which would demonstrate G8 commitment to government ownership. Policies for pro-poor growth should be part of any national anti-poverty plan. An examination of the distributional effects of government taxes, current expenditures and investments should 314 also figure in such plans. In some countries, making tax systems less regressive might do more for the poor than the existing targeted programs. In broader terms, NGOs ask for a shift from the Structural adjustment programs to a postWashington Consensus. 2.6 - G8 commitment on the coherence of overall policies which have an impact on poverty reduction in developing countries Daily more than $1.5 trillion is exchanged in the world's currency markets, and every year nearly a fifth of the world's goods and services are traded. Foreign direct investment topped $400 billion at the end of the 1990s. But the benefits of these global economic transactions are being spread inequitably among countries and, within them, between rich and poor. About a fifth of developing countries receive four-fifths of total private capital flows, and official development assistance, which is supposed to counterbalance the effects of market forces, is now a third lower than in 1990 in real terms - and shows no prospects of recovering. The growth of developing countries was modest during the 1990s - a result of the slowdown in world trade, declining commodity prices and heavy debt burdens. Fifty-five countries mostly in Sub-Saharan Africa and Eastern Europe and the CIS -have registered declining income. Globalisation can (and does) negatively affect poverty especially in areas such as trade and investment, finance, knowledge and technology. The challenge is to ensure that globalisation will benefit poor people. Coherence is not a mere co-ordination between Bretton Woods Institutions on their narrow range of policies. NGOs support the idea of «Systemic issues» which frames that coherence has the dimension to make local development goals and International Financial Institutions policy coherent each other. G8 should commit to set up coherent and co-ordinated policies in order to reach the international development goals. They have failed to bring their trade, aid and financial policies into line with their commitments to poverty reduction. The NGOs stress that coherence between aid goals and economic and trade policies which affect poor countries must be strengthened and, moreover, adequate strategies and methodologies should be co-ordinated among donors, and between donors and «beneficiary» governments. Programs should have a concrete impact on the reduction of poverty and social exclusion, considering all the different dimensions of poverty and promoting analysis of the constraints and potentialities of each developing country. Another important responsibility of G8 in terms of cohesion, is the strengthening of an integrated approach at international and national level, as the necessary pre-requisite for making development co-operation more efficient. At this regard the NGOs raise the problem of a more and more fragmented aid system that causes great confusion among the different actors involved in development cooperation. Effective Poverty reduction strategies should integrate economic, social, environmental and governance concerns within a comprehensive approach to development. Reducing poverty requires coherence in all government policies affecting development. Key policy areas with potentially strong poverty reduction impact include debt relief, trade, investment, agriculture, the environment, migration, health research, security and arm. The integrated set of these areas represents the bulk of international policies affecting poverty and requiring coherence. Arms trade. Arms trade is a major cause of suffering and of human rights abuses. Poor countries spend more on military expenditure than on social development, communications infrastructure and health combined. While every nation has the right and the need to ensure 315 its security, nowadays arms trade adds fuel to conflicts, wars, genocide, and epidemic diseases. Global military expenditure and arms trade is also the largest spending in the world at 800 billion dollars, annually. As world trade globalises, so does the trade in arms. In order to make up for lack of domestic sales, newer markets must be created. USA and Britain respectively do the largest and second largest businesses of arms trade in the world. Sometimes, these arms sales are made secretly and sometimes knowingly to human rights violators, military dictatorships and corrupt governments. This does not promote democracy in those nations. During the 1990s, the idea of a large peace dividend has drastically been muted. The resources assigned to ODA are equivalent to only 1/10 of armaments expenses. Rich countries’ weapons are not only used for national defense, they are used by the Turkish army against the Kurdish population, by the Indonesian army against the population of East Timor, and in civil wars on the Balkans and in Africa. Although rich countries’ governments claim not to export arms to countries that are repressing their population or that are violating Human Rights, in practice there is very little restriction on arms exports. Rich countries do not only earn money from arms exports. They also support the build up of arms industries in countries in the South by exporting technology and sometimes even whole production lines. Rich countries are providing loans and export credits to clients of the arms factories. This financial support is given by governments as well as by banks. Rich countries’ policy fails to provide full respect for international humanitarian law and falls short of establishing adequate mechanisms and procedures for States to take co-ordinated action to effectively monitor and control transfers by the States and their nationals of military, paramilitary and security equipment and services. Despite appeals from NGOs, there is no explicit obligation to prohibit transfers to forces which would most likely use them to seriously violate international humanitarian law (which sets out the rules of war). Moreover, there are virtually no provisions to address the current deficiencies in G8 States' arms control regimes, such as the failure to strictly regulate international arms brokering and licensed production agreements, or to adopt rigorous systems of certifying and monitoring end-use. There is no provision for public scrutiny over arms exports from the G8 and thus does little to foster greater transparency and accountability over the arms trade across the world. These omissions will need to be rectified in the near future if a real aim is to achieve high common standards in management of and restraint in conventional arms transfers in order to get a more coherent framework for poverty reduction strategies. First, G8 should commit to effectively discourage developing countries from high level of military spending and contrast arms trade, including the growing availability of small arms which has been a major factor in the increase in the number of conflicts (in modern conflicts over 80 percent of all casualties have been civilian, 90 percent of these are caused by small arms, and civilians are the main landmine casualties), military propaganda for arms sales, the fact that army schools set up to preserve democracy train many of the worst human rights violators and dictators in various developing countries. The European Union has tried to take a responsible step in introducing a Code of Conduct in the sales of arms, but the US seem unwilling to commit to an international agreement. Migration. In an interdependent world, poverty in developing countries increasingly affects the economic, social and political welfare of developed countries. Poverty can lead to serious global problems, such as environmental degradation, political and economic instability, and large-scale migration of people in search of a better life. Political coherence requires that G8 immigration policies will play an important role with respect to improving the living conditions of that part of the population, which may be forced into migration due to poverty. Basically, the NGOs think that migration is not a basic question of security, rather it substantially reflects development problems. The new dimensions of international migration are linked to the multidimensional nature of development and poverty: environmental political, ethnic, economic migrations. Insecurity, vulnerability, war, risk, lack of freedom 316 induce people to migrate. Also the opposite side of the coin, peopling skilled international migration and brain drains may affect development. One priority is the liberalisation of the so-called «movement of natural persons», which basically means that it should be easier and more transparent for workers from developing countries to get short-term visas to work in developed countries. Thus, migration should be strictly linked to development co-operation policies in a comprehensive approach. Migration is a direct outcome of poverty; sustainable and equal development is the way to escape from poverty. Finance and debt relief. Globalisation means much more financial-oriented exchanges in the world. The key to meeting the financing needs to support the international development goals in the poor countries does lie with domestic resource mobilisation, but also with how aid and debt relief can provide additional resources and increasing developmental private capital flows. Domestic resource mobilisation are requested for financing development goals, as the shared strategic goal is to make countries less dependent on external financial aid, but international finance institutions can not discharge their responsibility. (1) Basically, G8 countries’ implementation of the enhanced HIPC initiative should be accelerated and broadened, de-linking – if needed - debt relief from the rigid approval of PRSPs in order to accelerate the process and complementing its implementation with additional actions of poverty reduction strategies and ensuring that the benefits of this debt relief are used to reduce poverty. With the debt burden still compromising developing countries’ ability to address the basic needs of their populations, any credible strategy by G8 countries on poverty reduction needs to incorporate much more substantial debt reduction. (2) G8 countries’ support to a reform of the International Finance Institutions, which should foster pro-poor macroeconomic policies through their lending conditions. IMF policies should be reformed to provide more effective prevention against financial crises, and more even-handed burden sharing between debtors and private and public creditors through new mechanisms for orderly debt workouts. (3) G8 countries should define and implement approaches to reducing private flows volatility, particularly equity portfolio investments, and collecting new resources based on the private capital flows (see Paragraphs 3 and 4). Trade. The figures speak for themselves: the level of overall subsidisation of agriculture in the OECD countries rose from $182 billion in 1995 when the WTO was born to $280 billion in 1997 to $362 billion in 1998. World exports of goods and services expanded rapidly between 1990 and 1998, from $4.7 trillion to $7.5 trillion. But the least developed countries, with 10% of the world population, accounted for only 0.4% of global exports, down from 0.6% in 1980 and 0.5% in 1990. Sub-Saharan Africa’s share declined to 1.4%, down from 2.3% in 1980 and 1.6% in 1990. Although average tariffs are higher in developing than in developed countries, many poor countries still face tariff peaks and tariff escalation in such key sectors as agriculture, footwear and leather goods. As G8 governments negotiate global policies, they are charged primarily with pursuing national interests, so they fail to produce pro-poor policies. After the Uruguay Round, it was estimated that the new trade arrangements would lead to an increase in global income of some $212-510 billion, but a net loss of $600 million a year for the least developed countries, and $1.2 billion a year for Sub-Saharan Africa. The collapse of the agricultural negotiations in Seattle is the best example of how extremely difficult it is to change the attitude to pursue only national short-term interests. The European Union opposed till the bitter end language in an agreement that would commit it to "significant reduction" of its agricultural subsidies, which are $360 billion a year (more than 317 the African GDP). The US resolutely opposed any effort to cut back on its forms of subsidies such as export credits, direct income for farmers, and "emergency" farm aid, as well as any mention of its practice of dumping products in developing country markets. The multilateral trading system has placed the burden of adjustment on developing countries relative to countries who can afford to maintain high levels of domestic support and export subsidies. Trade development challenges vary from country to country, but most developing countries wishing to increase their participation in the global economy face a common set of obstacles. It is useful to think about these challenges at three levels: the international system, national policy, and the private sector. At the international level, inadequate market access and limited capacity to participate in multilateral negotiations are perhaps the most pressing obstacles to trade development. The Uruguay Round produced a number of valuable market access opportunities, but several agreements important to developing countries have yet to be fully implemented. Tariff peaks and tariff escalation still burden a number of exports, such as textiles and agriculture, in which developing countries have comparative advantages, and anti-dumping actions and other restrictive measures have proliferated in industrial country markets as traditional import barriers have been removed. In addition, multilaterally agreed liberalisation measures have eroded the value of special and differential treatment for many developing countries. Many developing countries also lack the resources and skills necessary to participate effectively in the WTO and other international economic fora. Several dozen developing countries are unable to post a single full-time representative at the WTO, and some maintain no diplomats in Geneva at all. G8 should support policies relevant to providing enhanced access to our markets, improved rules under the WTO system, appropriate responses to environmental and social concerns, and effective participation of developing countries in international negotiations and system building. At this regard, the NGOs call for: (1) More favourable conditions of market access, including the removal of anti-dumping measures and non-tariff barriers by the G8, for major export items of developing countries can offer the potential for $700 billion in additional exports by 2005 for these countries, four times the average annual private capital inflows in the 1990s. (2) G8 Adoption of the EU’s «Everything But Arms» initiative, which is only a first step. The EU has proposed to remove all restrictions on imports from the 49 LLDCs of all products apart from armaments. From a development perspective, the principal immediate effects of EBA will be to improve, in theory, access to the EU market for the exports of ACP LLDCs and, to a much smaller extent, non-ACP LLDCs; to increase competition for industrialised countries, other ACP states and the Standard GSP beneficiaries of South Asia, Mercosur, South East Asia and the states of Eastern Europe and the Mediterranean. Neither the scale of the potential boost to LLDC exports nor the increased competition for other developing countries is likely to be large in absolute terms (but non-LDC developing countries - particularly those from the ACP group are concerned about the EU proposal, and the result will be an increase in EU dumped exports onto the world market where they will lower prices still further for ACP states). For the LLDCs, though, quite small absolute gains are likely to be relatively more important because their export base and their international bargaining power are so limited. EBA is a positive modest statement of good faith, even though the amendments introduced in January 2001, which concern the transitional periods for the phase-in of duty-free access for bananas, sugar and rice, reflects domestic agricultural concerns. Moreover, The EBA initiative 318 invalidates market regulation instruments such as the system of quotas and prices, in order to pursue a free-trade liberalisation; but the real problem which prevents LLDCs from real access to EU market is the supply capacity of exporters and the non-tariff protectionism (including anti-dumping and standard measures). (4) G8 should support effective trade capacity building and promote untying ODA to the least developed countries, in order to promote domestic providers of goods and services. While the proportion of aid tying has declined recently, it remains a significant factor affecting the quality and impact of development co-operation. Around one fifth of bilateral aid is given on the condition that it is used to purchase products and services from donors. Health research. The 1994 agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) tightens patent and copyright protection, favouring those who develop and market technology rather than society’s interest in diffusion of new technology. This agreement has negative consequences for protecting the traditional and collective knowledge of indigenous people and for the rights to health, that is negative impact on poverty reduction strategies. Biotechnology for plant breeding and pharmaceuticals has given enormous economic value to genetic materials, plant varieties and other biological resources. But industrialised countries hold 97% of all patents, and global corporations 90% of all technology and product and process patent, using traditional knowledge that communities have held for centuries. At the end of 2000, more than 36 million of people are infected with the HIV virus: more than 25 million live in Sub-Saharan Africa, 6 million in South and South-Eastern Asia, 1.5 million in Latin America. During 2000, there have been 5.3 million new infected people (only 40,000 in the Northern America and 30,000 in the EU). In the United States, people who are infected with the HIV virus can now have their lives extended indefinitely through a combination of drugs known as AIDS cocktails. The cost of these drugs is $10,000 to $15,000 a year, placing them far out of reach of the 33 million people in low-income countries. But the cost of producing these drugs is a tiny fraction of their price. An Indian generic drug manufacturer, Cipla, recently offered to provide the drugs to governments for $600 and to NGOs such as Médecins sans Frontières for $350. Copyright protection favours the «AIDS profiteers» and affects negatively human rights and corporate reputations. There is a widening health divide between rich and poor countries. This divide is being widened by the growing disjunction between international research and development of drugs on the one side, and the health needs of the poor on the other. The supply of drugs is increasingly geared towards prolonging life in rich countries, rather than saving young lives in poor countries. The development of treatment and vaccines for poverty-related diseases is dramatically under-funded. For pharmaceutical corporations, HIV/AIDS, malaria and tuberculosis, among the major causes of child mortality in poor countries, offer limited profits, and attract limited investment. Meanwhile, the enforcement of patents is raising the costs of essential drugs. At this regard, the NGOs call for stronger action: (1) to consider poor people access to seeds for food crops or to live-saving medicines as a priority of international poverty reduction strategies; (2) to give, within the implementation of the 20:20 Initiative, more ODA resources to fight against HIV/AIDS, whereas the ODA assigned to this end has decreased by 29% in the last 20 years; (3) to use market incentives, such as positive and negative tax credit schemes for pharmaceutical companies to redirect research efforts priorities, nowadays focused on the problems of rich, which is part of the under-provision of international public goods; 319 Strengthening the capacity of public and private stakeholders to engage in a dialogue on poverty reduction strategies cannot be done solely for the sake of promoting such a dialogue and based uniquely on the financial resources. NGOs stress the importance of some data: informal sectors in developing countries represent one of the most effective economies with immense potential for decent job creation and poverty reduction, whereas small and medium sized enterprises are the bulk of the most dynamic knowledge economies of industrialised countries. Globalisation requires G7 countries to support an international partnership involving the small and medium enterprises – which are not crucial in terms of foreign direct investment – and informal sector and small and micro enterprises, in order to create an enabling international environment of private sector involvement in poverty reduction and decent work creation strategies, supported by institution building, technical programs and development co-operation. 3. Finance for Development The overall process of the Financing activities for Development is receiving increasing attention due to the ongoing organisation of the International Conference on Financing for Development. The international gathering is scheduled for 18-22 March 2002 in Monterrey, Mexico, to be held at the highest political level under the auspices of the UN and the involvement of all the international financial institutions (IFIs). The General Assembly of the UN decided that this event should address national, international and systemic issues relating to financing for development in a holistic manner in the context of globalisation and interdependence. The Conference should address the mobilisation of financial resources for the full implementation of the outcomes of major conferences and summits organised by the UN during the 1990s. It is hoped that the meeting will provide a standard guideline, which is expected to critically influence future policies on financing for development of governments and IFIs. The positions of the NGOs presented in this report can be included in the general discussion, which is animating this process. Since the issue of Finance for Development is strongly linked to the strategies adopted to reduce poverty and to the international trade mechanisms, some parts of the following paragraph are overlapping with the paragraph 1 on Poverty Reduction and 4 on the WTO reform. 3.1 - Official Development Assistance There is a broad consensus among the development community that a new impetus in Official Development Assistance (ODA) is necessary if the central goal of halving poverty by 2015 has to be achieved. Of course, ODA is not the only but just one fiscal instrument to fight poverty and to promote sustainable development. However the transfer of public resources has to play a vital role because in central areas of sustainable development simple blind faith in private capital and the forces of free market alone would lead to damaging or, at the best, to ineffective results. In addition, it should be noted that ODA flows continue to represent the core of external financing in virtually all low-income countries. Faced with this situation the crisis of ODA appears dramatic3. NGOs recommend subjecting official development financing to a thoroughgoing review. With this respect the following analyses and policy recommendations have been proposed: 3 Statement by Jens Martens (2000), Overcoming the Crisis of ODA the Case for a Global Development Partnership Agreement, World Economy, Ecology and Development Association (WEED), November. See also Chapter 1. 322 (4) to introduce obligations and responsibilities of rich countries’ corporations, preventing the exploitation of lack of environmental and health controls and research standard in developing countries, in order to help shape social norms and create profit motives to promote realisation of poverty reduction strategies; (5) to support through ODA flows the set up of national health systems, reverting the negative trend imposed by the implementation of the Structural adjustment programs; (6) to build human rights safeguards into the TRIPS agreement and its implementation, because of the concern about the compatibility of the TRIPS agreement with human rights law and environmental agreement. Environment. Environmental co-operation has risen to the top of the international policy agenda. A coherent poverty reduction strategy must directly involve the environmental mainstreaming component, as poverty is not only cause of environmental degradation, but it is also worsened by unsustainable development. G8 must recognise the importance of their commitment to implement international protocols and agreements as a way to address industrialised countries’ concerns and poor interests. Full co-operation on this issue will demand substantial commitment and coherent conduct by G8, more than by other countries. G8 should require a private sector’s full participation in the fulfilment of this objective: only a shared responsibility of all the involved stakeholders will make the commitment credible. Investment. Foreign direct investment flows have boomed, reaching more than $600 billion in 1998. But these flows are highly concentrated, with just 10 countries receiving 70% of the $177 billion going to developing and transition economies. The least developed countries attracted less than $3 billion in 1998, a mere 0.4% of the total. The NGOs recognise that foreign investment can be important for economic development, at the same time they express their concern about the impact of these investment on host countries, particularly referring to the economic, social and environmental elements of the sustainable development agenda. It is important that G8 assume their own commitments rather than those of other countries. Rather than imposing particular conditionality in terms of favourable environment to investment or defining rules and norms for developing countries (such as attempted by the Multilateral Agreement on Investment negotiated by OECD and failed in 1998, which did not fit the development needs of developing countries), G8 should commit to implement a set of recommendations for responsible transnational corporate behaviour world-wide, consistent with existing legislation. Introduction of procedures and mechanisms established to promote transnational corporations’ transparency and accountability, to eliminate child and enforced labour, combat corruption, promote human rights and environmental sustainability. G8 should facilitate to incorporate stronger social, ethical and environmental standard into their management strategies. The OECD Guidelines for Multilateral Enterprises and OECD guidelines for Export Credit Agencies at the highest international standards set out some standards and benchmarks for improving the governance of FDI. Private corporation - through investment decision with huge effects on economic growth, employment conditions and environment – can help open opportunities for decent work and for reducing poverty. 2.7 - G7 commitment on governance and co-ordination involving governments, international organisations, private sector and civil societies Achieving poverty reduction for all people in the world requires action and commitment from all the groups in every society, NGOs, media and business, local as well as national government, parliamentarians and opinion leaders. 320 New partnership is needed to extend to donor relationships with the private sector and civil society in developing countries, not just to relations with their governments. The goal of these new alliances is the creation of a locally owned development strategy in which donors are able to respond more precisely to national priorities. The weakness of political power compared with economic and financial powers, as the most urgent problem to be solved in the international co-operation system and more in general in the governance of globalisation. The NGOs recognise that the governance of globalisation requires co-ordination between all the international actors involved in development co-operation as well as cohesion within all the activities implemented by a development agency. The NGOs emphasise the need to retrench the role of the International Financial Institutions (IFIs) involved in the development process and for States to retake possession of their role of leadership in proposing development policies. Politics must regain its guidance and monitoring role at multilateral level too, and this is possible by granting stronger mandates along these lines to the United Nations. In this case, the role of the United Nations, considered by the NGOs to be the only institution fit for fighting the international economic system's distortions and for promoting stronger partnership with developing countries to fight poverty, must be reaffirmed. NGOs state that in the context of globalisation an international system capable of ensuring that economic and financial integration is consistent with the objective of the fight against poverty and social marginalisation and exclusion is needed. The UN organisations should be vested with political powers not reduced to the level of mere executors of co-operation programs. In the field of development, especially the UN-ECOSOC should be supported and made more effective in its procedures, through a reform process emphasising its model character of stakeholder participation at equal levels, in order to increase coherence and consistency of operations and policies. The Intergovernmental machinery must form an integral part of the UN reform process. NGOs believe that the Under Secretary General responsible for development issues should also be Executive Secretary of ECOSOC, which has an important role to play on economic, social or environmental issues of overriding importance. NGOs support strengthening of ECOSOC, leading it to becoming a more effective body. This would involve, inter alia, the change of the ECOSOC membership: a proposal currently debated implies a membership distributed in a three-tiers way (1/3 representatives from the G7 countries, 1/3 from the most peopled countries, 1/3 from general election). Private capital, entrepreneurial capacity, knowledge, technology are crucial to promote development and reduce poverty. But private interests are not automatically effective in stimulating long-term development. The discussion on the Public/Private Partnership is the core of the current international debate on poverty reduction strategies. The NGOs affirm to welcome partnerships between public and private sectors that can contribute on making business decision makers more "development sensitive", but they disagree with the creation of spot Trust funds, which they consider a form of disguised charity. A partnership with the private sector should urge enterprise to act according to codes of conduct based on the respect of human rights and sustainable development, and in accordance with the International Development Goals. The PPP needs to be a long-term, comprehensive process-oriented approach. International private sector must be involved, not only in terms of financial resources, but in terms of coherent conduct and respect of culture, law and international declarations. 321 - Donors countries should avoid any declines in ODA and should pledge to honour existing commitments to meet the international ODA target of 0.7% GNP within a defined time frame. - In order to overcome the traditional dependency relationship between "donors" and "recipient", new forms of contractual relations between all countries should be established under the auspices of the UN. This objective could be met by a Global Development Partnership Agreement4, based on existing legal documents such as the International Covenant on Economic, Social and Cultural Rights. The Cotonou Agreement between EU and Africa, Caraibi and Pacific countries (ACP) can serve as an indication; while the proposal for a binding Anti-Poverty-Convention discussed in Geneva during "Copenhagen +5" follows the same reasoning. - It is time to rethink the quantitative target of ODA. So far the Gross National Product of the donor countries has served as the "assessment criterion" for the level of ODA and it remains to be an important political criterion. But, while this target reflects the "supply side" of ODA, it could equally think about other indicators focusing on the "demand side". This new need-based target for ODA, however, must not be regarded as an attempt to justify a further decline of ODA flows. - A large share of the current development assistance is provided in the form of soft loans. Any increase in this concessional kind of development assistance leads to a possible increase in the foreign debt of the recipient countries. It raises the general question of whether ODA should, in future, be made available in form of repayable loans at all. - There is a large consensus on situations for which only grants not loans are appropriate: natural and conflict emergencies, basic social services (basic education, health care and nutrition, reproductive health, water supply and sanitary facilities) and technical assistance in capacity building. But this list should also consider environmental measures and support for non-export-oriented agricultural production. - The policy of cost recovery which some bi- and multilateral aid agency introduced in the ‘90s in some basic social services should be completely abandoned. User fees for health services and education have a detrimental effect on access of the poor to such facilities and the funds raised by user fees have very little impact on financial stability of these services. Finally, the UN approach on this topic must be considered. In his Report to the PrepCom for the Intergovernmental event on Financing for Development5, the Secretary General states that ODA should be structure around two basic principles: "…supporting strategies that revolve around the goal of poverty reduction and that generate sustainable, equitable growth; and relying on policies and programmes that enjoy ownership by recipient countries' Governments and civil societies". The NGOs agree with the principles expressed by UN in the report: the conditionality from above (i.e. structural adjustment programs) has proved detrimental; ODA flows must be linked to poverty reduction, and above all it is necessary to strengthen ownership and empowerment of local actors to conduct development programs on their own behalf. But the NGOs are seriously worried that the instruments to achieve these goals would really involve all the institutional subjects, according to their competencies, and the civil societies. Among the national strategies for poverty reduction proposed by International 4 Ibid. Report to the PrepCom for the High Level International Intergovernmental event on Financing for Development (A/AC.257/12). 5 323 Organisations6, the PRSP seem the fittest. But since this approach is quite new, there is uncertainty on the degree of autonomy that governments from developing countries will have to draw up the papers. The NGOs welcome a recent proposal from the IMF’s General Director, Mr. Koehler, which implies the involvement of the UNDP in the evaluation of the PRSPs. However the main concern among NGOs is the real involvement of the civil societies in the PRSP process. Whatever kind of program for the exit from the debt crisis and for poverty reduction should start from the needs of the people, and the subjects of the civil society are the only one that can understand and mediate with the people. 3.2 - Trust Funds After the analysis of the ODA as a way to get major resources for development, it is important to understand the instruments that could be used as a vehicle for the allocation of these resources. Some first funding schemes have been already ventured, such as the "Global Environment Facility". Others have lately been proposed by the Italian Treasury, such as a "Multilateral Health Facility" administered by the WHO or a Special Trust Fund for education in highly indebted countries in co-operation with UNESCO. Some NGOs underline the importance of specific initiative for special cases, as it is the case for the HIV pandemic. But the participation and direct involvement of the civil society in the management of the thematic funds is the conditio sine qua non for their success. On the other hand special funds are multiplicating in the last year and there is great concern on the capacity of the international community to administer the new financial architecture created by the funds. In this context, some NGOs believe it is now possible to provide an overall framework in which a contractual global fund for the financing of development activities or the preservation of global public goods can become reality. The global conferences of the UN in the last decade have also established a consensus on the range of issues to be regarded as global public goods (GPGs): protection of the environment, education and sanitation, protection from social exclusion, gender justice, enforceable human rights and access to food. Mostly the provision of these public goods remains the task of national governments. As a matter of facts, however, national public goods are increasingly turning into international (regional e global) public goods, and their provision depends for the most part on international co-operation. This is surely the case for the global public goods of financial stability. In the Report of the coming Financing for Development Conference it is expressed an "…urgent need to review different options for financing GPGs… and to ensure that resources earmarked for GPGs concerns are additional to those geared to ongoing development assistance programmes". Referring to it, the Secretary General recommends: "…one option to consider could be to increase sector ministry budgets in donor countries to allow them to fund international cooperation linked to GPGs in their sector- while existing aid resources remain focused on the financing of national programmes." NGOs agrees with UN on this issue, but they also go a step further. Some of them propose to complement and strengthen existing bi- and multilateral efforts in development assistance by the setting- up of a contractual global fund under the auspices of the UN for the financing of GPGs7. Such a system would not substitute but complement the action taken at a government level, according to a principle of subsidiary. However, in a perspective of financing GPGs at international level, it is necessary to look beyond the funding capacity of individual 6 i.e. the UN Development Assistance Framework, the PRSP, the Comprehensive Development Framework and the Common Country Assessment. 7 A specific proposal has been advanced by an Italian platform of NGOs, see Italian NGOs Platform on FfD (2001), Towards a Contractual Global Fund for the Development of Global Public Goods, mimeo. 324 governments or intergovernmental agencies and devise a complementary global fund on a contractual, un-reversible basis. The goal of such a mechanism is to collect and distribute the surplus that the global economy is generating at the cost of externalising its negative effects across borders. By tapping a small part of these benefits, it is possible to create a global income for the financing of a fund for the preservation and development of specific GPGs. The form of a contractual global fund refers to the fact that all the nation-states are as well producers as consumers of GPGs, thought to highly different degrees. Moreover, only a contractual framework with automatic payment obligation and drawing rights according to national balances of production and consumption of GPGs, established in a common frame of evaluation, can assure the alignment of benefits. Such a system must necessarily build upon the capacity of states to provide an adequate governance and institutional framework for the pursuance of such goals, and on the adoption of incentives and internationally binding covenants. 3.3 - Taxation On the topic of innovative sources for development, including taxation, the Report of the SG on Financing for Development doesn't contain any proposal apart from a general advise on promoting national and international public/private partnership and a soft suggestion for a better analysis of national currency transaction taxes. This lack in the UN agenda of new forms of financing, including taxation, it is due to the strong opposition against international taxes by some government, particularly the US. As an example, the US tax law for the fiscal year 2000 makes all the contribution to UN conditional on the fact that UN is not engaged in any effort to implement or impose any taxation on US persons. On the other hand, NGOs explored several ways of improving taxation. Proposals have been done for the introduction of new national and/or international taxation. It has been decided to analyse some approaches to the topic, also if based on completely different criteria. Three of them are exposed here: (a) currency transaction taxation (CTT); (b) progressive income tax on the GNP of the rich countries; (c) taxation of negative externalities of production and consumption. (a) Currency Transaction Tax The idea to tax international currency transactions was firstly argued by the Nobel laureate James Tobin in 1972. Tobin proposed a CTT in order to stabilise exchange rates and to hinder short term speculation. The basic idea behind this so-called Tobin tax is simple: a small tax is levied on every currency transaction, i.e. both on buying and selling a foreign currency. This measure renders all currency transactions that speculate on minor exchange rate changes unprofitable. In particular, very short run speculations would hence be deterred. The financial crisis of the ‘90s determined the current revival of Tobin’s idea to tax international currency transaction. At present the support for a CTT is rising steadily among different institutions. Various parliaments (EU, US, France, Belgium, Canada, and Germany) have presented a draft bill for a CTT and the preparation process for the International UN Conference on Financing for Development discussed this topic. But the strongest call for a CTT, often denoted as "speculation tax", comes from the civil society. ATTAC8 is the organisation leading the movement for regulating financial markets in order to achieve a socially equitable and an ecologically sustainable development. Also the political opposition to the Tobin tax shall be noted. This political resentment was felt most 8 Association pour une taxation des Transactions financières pour l'Aide aus Citoyens. 325 strongly in the US. The following points are the main outcomes attended from the introduction of such a tax on financial transaction: - a CTT functions like a filter. It would discourage speculation by making currency trading more costly, i.e. by "throwing some sand in the efficient wheels of international finance". Because the tax rate is insensitive to the maturity of the transactions, the real tax burden would be higher the shorter the time horizon of the operation, so short-term speculative behaviour would be penalised. This implies a digressive structure of the tax. Currency transactions that have a base in real economic activity will not be hindered by the tax9. - A CTT stabilises financial markets. Assuming a decline in short-term transactions, the incentive to speculate on short-term exchange rate movements vanishes, too. This would also prevent speculative bubbles and the tax assumes a prophylactic effect with respect to crises. Especially developing countries would benefit from the reduced threat of a crisis, since this economies are particularly vulnerable to shocks and financial crises. From the point of view of development, the general increase in stability is very valuable. Both foreign trade and credit relations would become more predictable. - - Such a tax instrument would generate revenue that could be used as one possible source of finance to help meet some of the world's global economic and political challenges, such as maintaining a stable international financial system or world-wide poverty alleviation. Annual estimates of the tax revenue range from a few tens of billions to a few hundreds of billions US$ (depending on assumptions over the tax base, tax rate and types of financial instruments taxed). Thus, this globally raised revenue, largely out of the control of sovereign states, would create a truly global revenue base. Even at the lowest rate of 0.05%, the tax revenue would be more than twice as large as all current official development assistance by all industrialised countries taken together. The discussion on how the revenue should be used is still open: it has been proposed to collect the revenue in an international fund for development, ecology and social equity, or in a global fund to fight poverty. But, since part of the revenue could be used to solve national financial problems, it would increase the consensus also among budget-constrain countries. A mixture of national and international expenditure might be optimal from a strategic point of view to gather more support for the tax. As to the criticism, it has been said that such a tax would require a vast administration or would be technically unfeasible. But there is already an infrastructure, established for other purposes, which can be readily used without much effort for collecting the tax, both nationally and internationally: the system of inter-bank foreign-exchange netting and settlement10. Taking into account some of the indeed valid technical counter-arguments to the (initial) Tobin tax proposal, it has been demonstrated that a CTT can work, in practice, if it is engineered carefully11. 9 See De Brunhoff Suzanne, Jetin Bruno (2000), The Tobin Tax and the regulation of Capital Movements, mimeo. 10 See for further details Clunies Ross Anthony (1999), A Tax on Foreign-Exchange Transactions Report of a Consultation held by CIDSE in collaboration with the University of Antwerp (UFSIA), October 1999, Antwerp, Belgium. 11 A workable variant would draw upon the work of Professor Bernd Spahn of Frankfurt/Main University, who launched his proposal in 1995. This is based on a two-tier Tobin tax, levied as a national tax but introduced through an international agreement, with a minimal-rate transaction tax on all transactions (the “basic tax”), and a high tax rate (an exchange “surcharge”) that, as an anti-speculation device, would be triggered only during periods of exchange rate turbulence and on the basis of well-established quantitative criteria. 326 A CTT should not be regarded as a universal remedy for all the ills of the international financial system, it is one instrument among several for regulating financial markets. Especially with regard to the most serious speculative attacks and crises the tax is an instrument of little value. However the most important obstacle to the implementation of a CTT is the obstinate resistance of those who take advantage from the volatile financial system, i.e. private financial market players and individual governments, in particular the US administration. But NGOs are convinced that a CTT undoubtedly makes good economic sense. And for this reason they are fighting to stress that the Tobin tax is a strong means in favour of social justice for its function of redistribution- from top to bottom. (b) Progressive International Income Tax The proposal of a progressive international income tax to stabilise ODA flow comes from a study by Keith Griffin and Terry McKinley, published by UNDP Office of Development Studies. Both authors call for a new global safety net, a progressive income tax on the GNP of rich countries, the proceeds of which would be allocated to the poorer countries in line with a fixed formula. Their appeal is unambiguous: In creating a new framework for development co-operation, the objective should be to abandon the present system, where aid contributions are voluntary, the aid burden is distributed randomly and inequitable, and the aid flows are unpredictable because they are subject to annual appropriation by national parliaments. The world should move instead to a system, where contributions to the aid effort are obligatory, the burden is distributed progressively, and the annual flows are predictable. The idea of a progressive international income tax to finance foreign aid is not new, and if development aid is to have a future and be more than marginal in size, the idea should be taken seriously12. There are already precedent of such a "solidarity tax". In Germany, for instance, under the concept of financial adjustment among the federal states- the so-called "state financing offset" billions of Dollars are transferred from the economically stronger to the weaker regions each year. The EU, to name another example, has the instrument of Structural Funds to support the poorer regions and weaker economic sectors within the Union. By these means, between 2000 and 2006 an estimated 195 billion Euro will flow from the richer to the poorer sectors and regions of the EU. (c) Taxation of negative externalities on production and consumption The world economy uses global resources to generate benefits, which often do not return to the appropriate factors of their production or are consumed unequally. By tapping a small part of this benefits it is possible to create a global income for the financing of a development fund or for the preservation of specific GPGs, whose provision are clearly the task for the international community, such as the preservation of ecological and cultural resources, financial stability and equitable trade13. As already seen above, three forms of negative externalities on production and consumption shall become the sources of generating revenues for the contractual global fund: - Taxation of the use of global environmental goods: air, water, soil, forests and genetic resources. These are important factors of production that contribute to various transnational economic activities. Taxation of the transnational use of environmental commons, especially 12 13 quoted in the Statement by Jens Martens (2000), ibid. see Italian NGOs Platform on FfD (2001), ibid. 327 by transnational companies, on the basis of a global collective resource valuation, promotes the GPG of preserving and regenerating ecological and cultural capacities. - Taxation of short term financial and foreign exchange transaction promotes the global public good of financial stability. - Taxation of specific items in the trade of goods and services with particularly negative international externalities, such as: 1) goods whose production chain has an impact on global ecological balance (i.e. climate change or bio-diversity); 2) goods that can increase the chance of conflicts, such as arms; 3) products extracted in socially and environmentally sensitive areas. Indirect taxation by means of an increase in the insurance fees and export credits could also be explored. Its introduction promotes the GPG of equitable and just trade. 3.4 - Ethical Finance NGOs and academic institutions have proposed new creative sources of financing for development which are not only resorting simply to more aid, but with some use of market instruments. There are in fact different possibilities, in developed countries, for collecting private contributions and there is surely a large amount of people who would support ethically, environmentally and economically sustainable development aid projects. The precondition to assure the ethic value and the transparency of the process it is the strong involvement of both the civil societies of the countries in which the funds are raised/utilised. - The market for ethical saving is increasing in OECD countries, as it is shown by the fact that there are several 'ethic banks' and also that some commercial banks open 'ethical accounts'. Private non-profit companies with both high degrees of financial reputation and strong ethical credibility, could be the best actors for the market of ethical funds (i.e. churches and big international NGOs). - One proposal has been advanced14 to let the World Bank act as an intermediary institution between the savers in developed countries and the borrowers in poor countries. The World Bank may act as a sort of buffer between the rules of international financial markets and those that are fair for developing countries. The bank is already credible financial intermediary and manages guarantee funds, but possibly must establish its reputation also in terms of the appropriate way of employing its funds: i.e. sustainable human development oriented projects. - The OECD countries could allow their taxpayers to devote part of the taxes, or to deduct from their taxable income the money dedicated to a special fund for HIPC and/or development finance. - The microcredit and the fair trade are two milestones in the promotion of more effective "ethically correct" development policies. 24 million persons borrow money worldwide through micro credit programs, while 5 millions are involved in the fair trade. From the NGOs’ point of view, particularly micro credit, and the ethic finance in general, has a big potential and has to be intended as the finance of micro enterprises, of small producers and of local systems of production and not only as the "chance to come out from extreme poverty without assistance". The different possibilities illustrated in this paragraph: ethic funds, targeted development funds and microcredit to be financed through bond issues, tax facilities for northern taxpayers and ethical saving are not mutually exclusive. The general idea is that of testing the intentions 14 see G. Vaggi (2001), From Aid to Sustainable Finance and Trade, University of Pavia-CICOPS, April, mimeo. 328 and orientations of Northern savers, either through markets or through open mechanisms like the tax declaration. These instruments are more innovative and they have some obvious advantages over a 'compulsory' new tax for debt cancellation. 4. Debt Relief. HIPC at a glance: the need for more15 In the broader debate NGOs believe the debt has been determined by the inequity in the macroeconomic relations between the North and the South. The policies that caused the incredible revaluation of the dollar, at the end of the 70s, determined the multiplication of the payments for the debtor countries. If the amount of money paid by the debtors is calculated again using not the dollar as unit but an average value among currencies of different countries, it is clear that the debt has been completely returned, and debt has been paid many times over because of compound interest. NGOs believe that to cancel the debt is a question of justice and not of solidarity with the poor. During the year 2000 the debt relief process for the poorest countries received a much-needed boost achieving debt cancellation for some of the world's poorest countries. Thanks to the efforts of the G8 leaders, IFIs and the Jubilee 2000 campaigns, 22 countries achieved the socalled decision point in the enhanced Heavily Indebted Poor Countries (HIPC)16. But the general opinion, among NGOs, is that the present initiative concerning debt relief is not enough. As chapters 7 and 8 have stressed, a recent internal paper from the World Bank and the IMF17 states that "HIPC debt relief alone does not ensure long-term debt sustainability" (p. 24), offering that "the HIPC initiative provides a good basis for these countries to exit from rescheduling" (p. 2). This stands in marked contrast to earlier claims by creditors that the HIPC initiative would itself provide a "lasting exit" from debt problem. That is why this “official” paper is often cited by NGOs. In NGOs' opinion, there is strong evidence of the necessity of further steps for a permanent solution of the debt crisis: - the 22 mentioned countries are still spending on average more on debt servicing payments than on health care. And the projections for the future are not encouraging: the total scheduled amount spent annually by the 22 countries on debt between 2001-05 is 2.02$bn, while they currently spend 1.35$bn on the health care of their population18. - The HIPC initiative has managed to reduce the actual debt servicing burden of the beneficiary indebted countries by a grand total of just 3% in the period between 1996 and January 2001, that is the reduction of debt service effectively delivered by the HIPC Initiative19. - Zambia and Niger both face increased debt service payments after qualifying for HIPC, by 23 per cent and 32 per cent respectively20. 15 This paragraph relies upon much of the issues which are covered by chapter 4 (the history of foreign debt crises) and chapters 7 and 8 (the current multilateral HIPC initiative). Nevertheless it is treated here as a matter of recommendations, because of its political relevance and links to the other issues we present in this chapter. Obviously, a detailed analysis of these issues can be best found in the other paragraphs. 16 See part IV. 17 IMF and WB (2001), The Challenge of Maintaining Long-Term External Debt Sustainability, internal paper, April (DC2001-0013), mimeo. 18 World Bank (2001), Global Development Finance, World Bank, Washington D.C. and World Bank (2001), World Development Indicators 2000, World Bank, Washington D.C. 19 see EURODAD (2001), Debt and HIPC Initiative update, Spring meetings 2001, May 2001,. For the calculation, World Bank (2001), WB Financial Impact of the HIPC Initiative: first 22 country cases, March, mimeo. 20 World Bank (2001), ibid. 329 Facing this situation, NGOs alert Governments and institutions around the world that without major improvements, the enhanced HIPC initiative is likely to fail in its two stated goals: delivering a permanent exit from debt crisis and releasing substantial resources for poverty reduction. In particular the international community has committed itself to the achievement of the international development goals, including halving poverty, achieving universal primary education and reducing child mortality by two thirds, by 2015. Debt cancellation alone will not ensure success, but these targets are certainly out of reach if there is not a sustainable end to the debt crisis. NGOs unanimously agree that it is time to make a new contract with clear and fair mechanisms to achieve a common solution between North and South of the debt crisis. We now present in terms of operative proposals what has been presented in chapter 8 as critical comments to the HIPC Initiative. This is a complementary part that is much more oriented in terms of practical solutions. 4.1 - Sustainability NGOs and campaigners for debt cancellation reject the IFIs narrow financial and economic conception of debt sustainability, altogether believing that good economics and social justice are inseparable. In fact if passing from the general problem of debt to the technical suggestions for its solution, NGO accept a discussion with the IFIs on the issue of sustainability. In this discussion the main problem is: which criteria should be adopted to measure the sustainability of debt? The concept of debt sustainability has been defined by the WB as follows: "… a country can be said to achieve external debt sustainability if it can meet its current and future external debt service obligations in full without recourse to debt rescheduling or accumulation of the arrears and without compromising growth"21. As we will see in detail, the main indicator adopted under the HIPC Initiative to calculate the sustainability of a country has been the "export-to-debt" ratios criterion. Only in recent years and in few, specific cases, IFIs started to consider other economic indicators of the debt sustainability, such as the debt to fiscal revenue ratios. Among NGOs, the common idea is that the current HIPC criteria are still inadequate, because countries' overall development needs - and particularly resources required to tackling poverty - have not been taken into account when assessing debt sustainability22. The main argument against the concept of sustainability used by the WB and the IMF has been developed by some Christian NGOs. CAFOD, in a document presented in the 1998, expressed the principle that debt sustainability must be seen in a broader context that incorporates the human development needs of the beneficiary countries. To this end a precise proposal has been suggested, which takes into consideration Human Development Indicators in the debt capacity analysis23. 21 A deeper analysis on sustainability conceptualisation and modelling is in part II. It is interesting to note that today the WB and IMF themselves admit that the definition of debt sustainability under the HIPC initiative "is quite narrow from a development perspective… and… does not measure the adequacy of public resources to address priority development programs after debt service has been paid", in IMF and WB (2001), ibid. (p. 12). 23 see H. Northover, K. Joyner and D. Woodward, (1998), A Human Development Approach to Debt Sustainability for the World's Poor, position paper, mimeo, which has been adopted then as a joint paper with Caritas Internationalis and CIDSE. 22 330 From the human development prospective, the focus of debt sustainability analysis should be shifted towards government revenues. In fact the debt threshold should take into consideration the governments' feasible net revenue, in other words the ability of governments to raise revenue and to fund poverty reduction programmes and to meet minimal primary services (health and education) for their population before paying the debt service24. In the traditional approach to sustainability debt service is a priority and not a residual, while according to the "bottom-up" approach described above the priority is human development. Other ways have been examined to modify the HIPC criteria. Many NGOs have argued that a specific assessment of the cost of meeting the 2015 development targets should be made and the definition of affordable debt payments (and therefore sustainable debt) based on. Oxfam, for example, calculates that future debt service for the 22 countries, at 2$bn a year, nearly matches the cost of achieving the education and health goals in these countries, at around 2.7$bn a year25. 4.2 - Eligibility criteria to receive debt relief Logic consequence of a "bottom up" approach to debt sustainability is that the eligibility criteria to receive debt relief should be re-opened. In NGOs' opinion, eligibility under the HIPC Initiative should be separated from the export rate and linked to the poverty and human development index. As said above, the criteria to define the sustainability for being eligible of a country should be based on the ability of governments to fund basic needs satisfaction for their population before paying the debt service. From this perspective, any low-income country that currently suffers from high levels of indebtedness coupled with low government revenues and widespread poverty should be eligible for such initiative. Nigeria, for example, is a case for immediate inclusion in the debt cancellation process, but also all the low-income countries, as proposed by the Italian government26. Countries that are relatively less indebted, or have relatively high government revenues, would as a result receive a lower absolute amount of debt reduction. Such an approach would be empirically consistent across countries, avoiding the arbitrary and unfair current distinction between poor countries that are HIPC and those that are not HIPCs, which is sometimes advanced by the IFIs as a reason for not proceeding further with debt reduction. A special case should be considered for the eligibility of countries afflicted by economic, natural and social shocks; in particular the HIV/AIDS emergency in the Sub-Saharan Africa and the future impact of this epidemic on the governments' ability to repay debt27. Concerning the situation of countries which are not eligible for the cancellation because of conflict or violation of human rights, NGOs welcomed the announcement by the UK government to put in a trust fund any debt service from those countries till their achievement of the Decision Point in HIPC. 24 "This is in contrast to the enhanced HIPC initiative's fiscal criterion where the setting "sustainable" ratios of debt to revenue are qualified by even more arbitrary sub-criteria". In H. Northover (2001), The Human Development Approach to Debt Sustainability for the World's Poor, background paper presented to the GNG Initiative, mimeo. 25 Oxfam (2001), Debt Relief: Still failing the poor, position paper, April, mimeo. 26 See M. Zupi (2000), Note sul tema della riduzione del debito estero dei paesi poveri, "Finance and Development" Working paper, CeSPI, Roma, Dicembre. 27 S ee National Aids Trust for more detailes on the topic. 331 4.3 - Cut-off date Under the HIPC criteria only the debt contracted before a given day (the so-called cut-off date) can be cancelled. But some countries, like Zambia and Niger, face increased debt service payments after qualifying for HIPC. Thus, NGOs call for a redefinition of the cut off date. Most NGOs hope that the G8 would invoke a new cut-off date for the Genoa summit: the day of the Koln summit, in July 1999, has been proposed. 4.4 - Cancellation of the Multilateral Debt In order to provide further debt relief to the HIPCs, the G8 have set a powerful example by agreeing on 100 per cent cancellation of bilateral debt. This example needs to be replicated throughout all the creditor countries (both the Paris Club and non-Paris Club members). For the NGOs it is now important that WB and IMF would follow the G8 with an immediate 100 per cent cancellation for the graduate HIPCs. According to Drop the Debt28, the successor to the Jubilee 2000 English Campaign, two main facts must be considered to justify the cancellation of the multilateral debt: After the effects of the enhanced HIPC initiative and the additional multilateral pledge, the major creditors to the 22 decision point countries will be the multilateral agencies, with on the top of the list the WB's International Development Association (30%) and the IMF (10%). The benefit of a 100 per cent cancellation by the World Bank and IMF alone would be around $7.2 billion in Net Present Value (NPV) terms, or about $13 billion in nominal terms. In the case of the IMF, such a cost could be easily met by using the Fund's own income stream. For the WB, the total additional cost of going up to a 100% cancellation for the 22 graduate HIPCs amounts to an additional $215 million per year for the next five years. This is eminently affordable, and could be provided from a range of resources. These are the findings of an independent analysis commissioned by Drop the Debt. 4.5 - Fair and transparent arbitration process The NGOs believe that inequity in the relations between the North and the South is the main reason determining the unbalanced process of decision making in international debt management. At present, the debtor countries have to submit their economic sovereignty completely to the conditions set by the creditors, who define the process as such and set up the rules of the process. The NGOs and international campaigns for debt relief support the introduction of an international arbitration procedure between debtors and creditors. A fair and transparent arbitration process (FTAP) must be established to this end, containing the following principles: - an impartial decision making body, independent from the involved parties; - the right of all stakeholders to be heard before a decision is being made; - the protection of the debtor's basic needs - especially the needs of the most vulnerable sectors of a sovereign debtor's society; 28 see Drop the Debt (2001), Reality Check report, London, mimeo, April. It incorporates independent studies by accountants Chantrey Vellacott (DFK) and Subhrendu Chatterji. 332 - the institution of an automatic stay once an international arbitration case is opened, in order to avoid a creditor's run on the debtor's remaining assets and to allow for an orderly procedure. These elements are part of the Chapter 9 of the US Insolvency Code, which has been proposed by the Austrian economist Kunibert Raffer for an international insolvency procedure to afford the debt crisis. The proposal has been taken up positively by the UN Secretary General in his Millennium address, by UNCTAD and by the German Federal Reserve (Bundesbank), thus reflecting the growing interest for the issue caused by the international debate among civil society organisations and academics. Other measures have been proposed: to abolish the debt when the financing has been used for assisted projects by the creditor partner who had a bad economic performance. 4.6 - From debt to development through debt-for-development swaps Italian Episcopal Conference (CEI) is conducting the debt-for-development swaps in Zambia and Guinea. This operation is now possible in Italy, thanks to a specific law on debt relief29. This law explicitly considers the cancellation of debt through conversions for sustainable development purposes. As a first step, the CEI is raising funds to finance the debt swap in Zambia and Guinea. If the Italian government doesn't cancel its credits, the Italian Catholic Conference is ready to buy the debt of these countries. The third actor involved, the local governments, should put part of its debt on a counterpart fund. Representatives of the local civil societies will then administer this fund to finance projects for poverty reduction. In fact the goal of a debt-for-development swap is not only the debt cancellation but also the obligation to destine the money freed to poverty reduction, with the direct involvement of the local civil society30. To enforce the participation of the civil society the Italian law make provision for the direct involvement of organisations that, after a fund raising, can participate in the management of the money coming from the conversion and from their own funds. In Zambia and Guinea two counterpart funds have been created with the money paid by the local government and the money collected by the Italian Church. A committee will administer this money, with representatives of the local civil society, one representative of the CEI and one of the local government. In the meantime a working group composed by representative of local associations, NGOs, and churches is selecting the projects that will be financed31. 29 Italian Law n. 209 approved on 28 July 2000. For details, see Zupi, M. (2000), ibid. For a good analysis of the debt swap mechanisms see Jurgen K. and Lambert A. (1998), Debt swaps for sustainable development, IUCN/EURODAD/SCDO, London, mimeo. 31 The Italian profit company Italsystems for Environment has proposed innovative ways for the managing of freed resources through the debt conversion, especially in the field of water resources. See for more information on the National Ecclesiastic Campaign for Debt Reduction and documents from the Italian Episcopal Conference. 30 333 Bibliography32 Abugre C. (2000), Criticism of PRSPs - Still Sapping the Poor: A critique of IMF Poverty Reduction Strategies, The World Development Movement, June (www.wdm.org.uk/). ActionAid (1998), In whose benefit? The case for untying aid. Policy briefing, April (http://www.actionaid.org/about_us/pub.html). ActionAid (1999), Good business: evaluating the impact of business-community partnerships, November, mimeo ActionAid (2001), Increasing the effectiveness of donor co-ordination. A case study of the education sector in Bolivia, Burkina Faso & Tanzania, January, mimeo. Ahamadi-Thoolen S. (2000), «The Link Between Human Rights and Democracy»; Development, Vol. 41, No. 4. Amalric F. (2001), The unbearable lightness of G7 concerns for the South: A comment on Italy's "Beyond Debt Relief", SID policy papers, SID, Rome, March, mimeo. APRODEV (2000), Building a Stronger Partnership. APRODEV Comments on the Commission's Paper: "The Commission and non-Governmental Organisations Building a Stronger Partnership, March. Arts K. (2001),WIDE positon paper on Gender aspects of the Cotonou Agreement, Network Women In Development Europe, January (http://www.eurosur.org/wide/publications.htm). Borrini-Feyerabend G., Farvar M.T., Nguinguiri J.C., Ndangang V. (2000), La Gestion Participative des Ressources Naturelles, GTZ, UICN. Bossuyt, J. Gould, J.(2000), Decentralisation and Poverty Reduction: Elaborating the Linkages, ECDP Policy Management Brief, October. CAFOD (2000), Reaction to the UK Government’s White Paper on Globalisation, December. Campagna per la Riforma della Banca Mondiale (2000), Decoder. Ambiente Sviluppo Instituzioni Finanziarie Internazionali, anno II n.6, Aprile-Giugno. Campagna Sdebitarsi (2000), Promemoria per audizione in Commissione Esteri della Camera dei Deputati su disegno di legge c. 6662, 24 Maggio, mimeo. Campagna Sdebitarsi (2001), L'arbitrato Internazionale: un elemento chiave alla soluzione del problema del debito dei paesi poveri, Roma, Gennaio, mimeo. Cassimon D. (1999), Taxing Excessive Currency Speculation to Prevent Social Crisis and Finance Global Challenges, CIDSE/Caritas Internationalis/Justice and Peace EuropeTask Group on Social Justice, January. CEI (Conferenza Episcopale Italiana)- Comitato Ecclesiale Italiano per la Riduzione del Debito Estero dei Paesi più Poveri (1999), Campagna Ecclesiale per la Riduzione del Debito Estero dei Paesi più Poveri. Documento presentato alla XLVI Assemblea Generale della CEI, Roma, Maggio, mimeo. CEPAL (2000), Equidad, desarrollo y ciudadanía, Santiago, mimeo. CEPAL (2000), La Brecha de la Equidad. Una segunda evaluación, Santiago, mimeo. Chauvin S., R. Corner, A. Chailleu, C. Pagnoulle (2000), Tobin Tax, Speculation and Poverty, Attac-Liège Thematic Group, December, mimeo. Chinnock J., Collison S. (1999), Purchasing power: aid untying, targeted procurement and poverty reduction, ActionAid, September. CIDA (1996), CIDA's Policy on Poverty Reduction, Ottawa, January. 32 The present bibliography is the result of the process of the GNG-Initiative consultation, started in January 2001, with the G8 countries' NGOs and Research Institutes. The documentation collected refers principally to the last three years: 1999, 2000, 2001. The bibliography contains also documentation of International Organisation and Institutions' reports. 334 CIDA (1998), CIDA's Social Development Priorities: A Framework for Action, Ottawa, mimeo. CIDSE (2000), Advocating for Greater Social Justice, January. CIDSE (2000), Biopatenting and the Threat to Food Security. A Christian and Development Perspective, February. CIDSE (2000), PRS - Poverty Reduction or Public Relations Strategies?; A CIDSE-Caritas Internationalis Background Paper, September, mimeo. CIDSE/ Caritas Internationalis (1997), Putting Life Before Debt, CIDSE, Rome-BruxellesNew York. CISL (2001), Prime valutazioni CISL in merito al documento della Presidenza del Consiglio: Beyond Debt Relief, policy paper presented to the GNG Initiative, April, mimeo. Clunies R. A. (1999), A Tax on Foreign-Exchange Transactions, CIDSE - University of Antwerp- UFSIA, Antwerp, Belgium, 22 October, mimeo. Commission on International Development (1969), Partners in Development. Report By Pearson, Lester B. et al., UN, New York. Comparative Research Programme on Poverty (2000), A Critical Review Of The World Bank Report: World Development Report 2000/2001 - Attacking Poverty, CROP Norway, mimeo (www.crop.org/cropweb.htm). Comparative Research Programme on Poverty (2000), The Role of the State in Poverty Alleviation, CROP Norway, mimeo (www.crop.org/cropweb.htm). Cornia G. A. (2000), Inequality and poverty trends in the era of liberalisation and globalisation, UNU/WIDER, Helsinki. Crook R. C., Sverrison A. S. (2000), To What Extent Can Decentralised Forms of Government Enhance The Development of pro-poor Policies and Improve PovertyAlleviation Outcomes? Mimeo. DAC (1996), Shaping the 21st Century: The Contribution of Development Co-operation, OECD-DAC, Paris. DAC (2001), 2001 Policy Statement by the DAC High Level Meeting upon endorsement of the DAC guidelines on Poverty Reduction, OECD-DAC, Paris, April. DAC (2001), DAC Guidelines on Poverty Reduction, OECD-DAC, Paris, April. DAC (2001), Development Co-operation report, OECD-DAC, Paris, April. De Brunhoff S. and J. Bruno (2000), The Tobin Tax and the regulation of Capital Movements, ATTAC (Association pour une Taxation des Transactions financières pour l'Aide aus Citoyens), mimeo. Debt Relief International (2001),Long-term debt sustainability for HIPCs: How to respond to shocks, January, mimeo. Demichelis D., Ferrari A., Masto R., Scalettari L. (a cura di)(2000), Debito da morire. Trentatre testimonianze sulla cancellazione del debito e i suoi inganni, Baldini e Castoldi, Milano. Dercon S. (1999), Income Risk, coping strategies and safety Nets; Katholieke Universiteit Leuve and Centre for the Study of African Economies, Oxford university, September. Dieci P. (2000), «The Growth of Inequality», Politica Internazionale, Gennaio-Aprile. Dieci P. (2001), The Challenge of Poverty Reduction and The Less Developed Countries (LDCs). Some Introductory Remarks, CISP, Jannuary, mimeo (www.cisp-ngo.org). Drop the Debt (2001), A New Deal on Debt for Genoa, making poverty reduction real. Drop the Debt briefing February, mimeo. 335 Drop the Debt (2001), Reality Check. The Need for Deeper Debt Cancellation and the Fight Against HIV/AIDS, April, mimeo. Erlassjahr 2000 (2000), International Insolvency/Arbitration Process: Arguments and Counterarguments, December, mimeo. Eurodad (2000), Global Financial Architecture and Development, Issues for NGOs, September, mimeo. Eurodad (2001), Debt and HIPC Initiative update Spring meetings 2001, May, mimeo. European Commission (2001), Communication from the commission to the Council and the European parliament; Programme for Action: Accelerated action on HIV/AIDS, malaria and tuberculosis in the context of poverty reduction, COM(2001) 96 final, Brussels 21 February (www.europa.eu.int). Eurostep (1998), Arms Export Controls. A Position Paper, March (http://www.eurostep.org/). Eurostep (2000), Coherence and Consistency of EU Policies: Proposed mechanisms for implementation; Position Paper on Coherence, February (http://www.eurostep.org/cohcons.htm). Eurostep (2000), Eurostep’s Initial Response to the European Commission’s Discussion Paper on European Community’s Development Policy, April. Eurostep (2000), Market access, agriculture, food security and reform of the World Trade Organisation, Position Paper, August. Eurostep (2000), Meeting the EU's commitments to education in developing countries: Implications for the 2001 budget; Recommendation paper; August. Eurostep (2000), Report on Workshop on Enhancing Civil Society Involvement in the Implementation of the New ACP-EU Agreement, Position Paper Geneva, 28-29 June. Foundation for Advanced Studies on International Development (2000); Agenda for international development 2000, Tokyo (http://www.fasid.or.jp). G8 Okinawa Summit (2000), meetings resolutions and action plan, Okinawa, July, mimeo. Green D., Melamed C. (2000), The rough guide to globalisation, CAFOD Briefing, June (www.cafod.org.uk/policyroughguide.htm). Hanmer L., Healey J. and Naschold F. (2000), Will Growth Halve Global Poverty by 2015? , ODI, London. Harris J. (1995), How Much Difference Does Politics make? Regime Differences Across Indian States and Rural Poverty Reduction, Development Studies Institute, LSE, London, October. Haut Conseil De La Coopération Internationale (2001), Pour Une Cooperation Adaptee Aux Besoins Des Pma. Avis Du HCCI En Vue De La Troisième Conférence Des Nations Unies Sur Les Pays Les Moins Avancés (Bruxelles, 14-20 Mai) Adopté En Assemblée Plénière Le 2 Avril. Hellinger D. (2000), United Nations Civil Society Hearings on Financing for Development, The Development Gap/ Structural Adjustment Participatory Review International Network, 7 November (www.igc.org/dgap/). Herrings R. J. (1999), Political Conditions for Agrarian Reform and Poverty Alleviation, Cornell University, prepared for the DFID Conference on 2001 World Development Report, August. Hobley M. (2000), Transformation of organisations for Poverty eradication: The Implications of Sustainable Livelihoods approaches; Draft for comments, April, mimeo. Hossain N., Moore M. (1999), Elites, Poverty and Development, A Background Paper to the World development Report 2000/, IDS, Sussex, June, mimeo. 336 Human Rights Watch - International Campaign to Ban Landmines (2000), LANDMINE MONITOR REPORT 2000 Toward A Mine-Free World, August. IDS (2000); Poverty Reduction strategies: A Part For The Poor?; Policy Briefing Issue, April. IFAD (2001), Rural Poverty Report, Oxford University Press, London. ILO (2001), Decent Work in the Global Economy, Discussion Paper No. 1, Geneva. ILO (2001), Perspectives on Decent Work, 2001, Geneva. ILO (2001), World Employment Report, Geneva. IMF (2000), Retarding short- term capital inflows throught a witholding tax, IMF working papers No. 40, Washington D.C. IMF and World Bank (2000), The Impact of Debt Reduction under the HIPC Initiative on External Debt Service and Social Expenditures, Washington D.C., November, mimeo. IMF and World Bank Development Committee(2001), The Challenge of Maintaining LongTerm External Debt Sustainability, internal paper No. 13, April, Washington D.C. Institute of Economics Affair and Society for International Development (2000), Kenya at the Crossroads. Scenarios for our Future, mimeo. INZET (2000), Eu-Acp: A Fair Deal? ACP trade and poverty eradication in the post-Lomé era, Briefing Paper, May. Italian NGOs platform on FfD (2001), Towards a Contractual Global Fund for the Development of Global Public Goods. Proposal of Italian initiatives for development and social and environmental justice to the UN High Level Conference "Finance for Development" 2002, Rome, February, mimeo. Italian Parliament (2000), Legge n. 209 del 28 luglio 2000, Misure per la riduzione del debito estero dei paesi a più basso reddito e maggiormente indebitati. Pubblicata nella Gazzetta Ufficiale n.175 del 28 Luglio. Italian Presidency of G8 (2001), Beyond Debt Relief. Preliminary version, March, Rome, mimeo Italian Presidency of the G8 (2001), Beyond Debt Relief, Rome, March. Italian Presidency of the G8 (2001), Intervento del presidente del consiglio Giuliano Amato al convegno "The challenge of Global Governance and the Role of the G8", 3 April, mimeo. Jubilee 2000 (2000), Through the Eye of a Needle, the Africa Debt Report- a country by country analysis, mimeo. Kajser J. (2001), A Fair and Trasparent Arbitration Process for Indebted Southern Countries, proposal for the Finance for Development Process, Erlassjahr 2000, January, mimeo. Kajser J., L. Alain (1998), Debt swaps for sustainable development IUCN/EURODAD/SCDO, mimeo. Terreri F. (a cura di) (2001), I nuovi scenari del debito estero e l'alternativa microcredito, Microfinanza SpA, background paper presented to the GNG Initiative, April, mimeo. Kilkpatrick S., Bell R., Falk I. (1998), The Role of Group Learning in Building Social Capital; Centre for Research and Learning in regional Australia, University of Tasmania. Lustig N., Stern N. (2000), «Broadening the Agenda for Poverty Reduction: Opportunity, Empowerment, Security», Finance & Development, IMF ; Vol. 37, No. 4. Madeley J. (ed) (1999), Brussels' blind spot The lack of coherence between poverty eradication and the European Union's other policies, APRODEV, October. 337 Marshall A., and J. Woodroffe (2001), Polices to Roll-backthe State and Privatise? Poverty Reduction Strategy Papers Investigated, WDM debt policy report, April, mimeo. Martens J. (2000), Overcoming the Crisis of ODA, the Case for a Global Development Partnership Agreement, WEED Association, November, mimeo. Martone F. (2001), Senza alcuna Considerazione Etica. Investimenti privati, diritti umani ed ambiente. Raccomandazioni per una riforma ambientale delle agenzie di credito all'esportazioni con particolare riferimento alla SACE, Campagna per la Riforma della Banca Mondiale, Febbraio. Maxwell S., Kenway P. (2000), New Thinking on Poverty in the UK: Any Lessons from the South?, ODI Poverty Briefing, London, November. Medici Senza Frontiere (2001), Dossier «Campagna per l’accesso ai farmaci essenziali», Roma. Moore M., Leavy J., Houtzager P., White H. (1999), Polity Qualities: How Governance Affects Poverty; IDS, Sussex, September. Mothers' Union (2001), Eliminating World Poverty for Today's Children - Tomorrow Citizens. A Six Pont Plan For Eliminating Child Poverty, February, mimeo. Nedervee Pieterse J. (2000), «After post-development», Thirld World Quarterly, Vol. 21, No. 2. New Humanity (2000), World Youth Fund. The source of the proposal: the experience of the Economy of Communion, May, mimeo. Northover H. (2001), The Human Development Approach to Debt Sustainability for the World's Poor, CAFOD, the GNG Initiative background paper, April, mimeo. Northover H., K. Joyner and D. Woodward (1998), A Human Development Approach to Debt Sustainability for the World's Poor, CAFOD (Catholic Agency for Overseas Development), position paper, mimeo. Olsen G. R. (2000), Public Opinion and Development aid: is there a link, CDR Working Paper, Copenhagen, December. Ottaway M., Carothers T. (2000)Funding Virtue: Civil society aid and democracy promotion, Carnegie Endowment for International Peace, mimeo. OXFAM (2000), Missing the target: the price of empty promises, July, mimeo. OXFAM (2000); Growth With Equity is Good for the Poor; June, mimeo. OXFAM (2001), Education: tackling the global crisis, April, mimeo. OXFAM (2001), Making PRSPs work. The role of poverty assessment. April, mimeo. Oxfam International (2001), Debt Relief: Still failing the poor, position paper, April, mimeo. Oxfam International (2001), Making PRSPs Work: The role of poverty assessments, position paper, April, mimeo. Paul J. A. and J. Garred (2000),, Making Corporations Accountable, Global Policy Forum, background paper for the UN FfD process, December, mimeo. Pouliquen L. (2000), Infrastructure and Poverty, December, mimeo. Promocion del Desarrollo Popular, FONAES Sedesol (2000), Vida Digna y Sostenible. Talleres regionales, local, global y mundial: surge un sistema sinérgico de intrecambio de valores 2000-2001, mimeo. Ramonet, I. (1998), «The politics of hunger», Le Monde Diplomatique, November. Reality of Aid (2000), - An Independent Review of poverty reduction and development assistance: Annual report - Poverty Inequality and Aid: Rhetoric and Reality, London. 338 Rebelo S. (1998), The role of Knowledge and Capital in Economic Growth, North-western University, September, mimeo. Rete Lilliput (2001), Per una Politica ed un'Economia del Bene Comune, January, mimeo. ReteLilliput (2001), Per Una Politica ed Un'Economia del Bene Comune, Mimeo. SID (1998), «Sustainable Livelihood: Communities as seeds of change», Development, Vol. 41, No. 3, September. SID (2001), Dossier «Building New Coalitions Around the Sustainable Livelihoods Approach», Rome, mimeo. Sinha S., Lipton M., Yaqub S., Church J. (1998), Undesirable Fluctuations, Risk and Poverty. A Review, Sussex University, mimeo. Staples S. (1998), Protecting War Militarism and The Multilateral Agreement on Investment (MAI), PEACEWIRE, September. Swiss Coalition of Development Organizations (1998), From Debt to Development, the Swiss Debt Reduction Facility, Berna, mimeo. T. Carothers T. (1999), Aiding Democracy Abroad. The learning curve, Carnegie Endowment for International Peace, mimeo. The Development Gap (2000), Statement Of The Development Gap On The Proposed Multilateral And G7 Debt-Reduction Plan Rejection of IMF Policies and Expanded Role in Debtor Countries, mimeo (www.igc.org/dgap/). The Global Alliance for Vaccines and Immunization (GAVI) (2000), Presentation, Washington D.C., mimeo. The World Development Movement (2001), Polices to Roll-back the State and Privatise? A debt policy report, April. UK Government (2000), Eliminating World Poverty : Making Globalisation Work for the Poor . White paper on international Development - Presented to the Parliament by Secretary of State for International Development, by Command of her Majesty, December. UN (2000), Recent Development in the Debt Situation of Developing Countries, report of the Secretary General, New York, 26 September, mimeo. UN (2000), Report of the Secretary-General to the Preparatory Committee for the HighLevel International Intergovernmental Event on Financing for Development, (A/AC.257/12), 18 December, mimeo. UN Department of Economic and Social Affairs (2000), Resources for Social Development: Additional and Innovative Measures, Clunies-Ross, DESA Discussion Paper n°11, New York UN, Joint Statement of the co-chairmen at the conclusion of the second substantive session of the preparatory committeeon financing for development. 9 March 2001 UNCTAD (2001), The Least Developed Countries Report, Geneva. UNCTAD (2001), Trade and Development Report, Geneva. UNDP (2000), Human Development Report 2000, New York. UNDP (2000), Human Poverty Report 2000: Overcoming human poverty, New York. Vaggi G. (2000), Debt Sustainability and the Financing of the HIPC Initiative, University of Pavia-CICOPS, Occasional paper series, January, mimeo. Vaggi G. (2001), Trade and Sustainable Finance for Development, University of PaviaCICOPS, paper presented to the annual conference of the Italian Association for the History of Economic Thought, Lecce 24-26 May, mimeo. 339 Van Hees T. (2000), Open Letter Eurodad to our ministers and representatives in the boards of the world bank and IMF, Eurodad, Bruxelles. van Reisen, M. (2000), EU Global Player The North-South Policy of the European Union, Terre des hommes – Germany, The Transnational Institute, WEED and Eurostep, May. Wahl P. and P. Waldow (2001), Currency Transaction Tax- a Concept With a Future. Chances and Limits of Stabilising Financial Markets Through the Tobin Tax, WEED working paper, February, mimeo. Welch C. (2000), Structural Adjustment Programs & Poverty Reduction Strategy, Foreign Policy in Focus, Vol. 5, No. 14 April. Whitehead L., Gray-Molina G. (1999), The Long Term Politics of Pro-Poor Policies, Nuffield College, Oxford, August. WHO (2000), Report on Infectious disease; Remove obstacles to healthy development 1999, Geneva. Wood A. (1999), What Role for the Multilateral Institutions, Donors, and NGOs in the New Framework for Poverty Eradication? Bretton Woods Project Briefing, November. Wood A. (2000), New Development Tools or Empty Acronyms? The reality behind the Comprehensive Development Framework and Poverty Reduction Strategy Papers, Bretton Woods Project, September. Wood A. (2000), The ABC of the PRSP An introduction to the new Bank and Fund Poverty Reduction Strategy Papers, Bretton Woods Project Briefing, April. World Bank (1990), World Development Report 1990, Oxford University Press, New York. World Bank (2000), World Development Report 2000/2001-: Attacking Poverty, Washington D.C. World Bank (2001), Financial Impact of the HIPC Initiative: first 22 country cases, March, Washington D.C. World Bank (2001), Global Development Finance 2000, Washington D.C. World Bank (2001), World Development Indicators 2000, Washington D.C. Yaqub S. (1999), How Equitable Is Spending on Health and education?, Background Paper to World Development Report 2000/1, Sussex University, September, mimeo. Zupi, M. (2000), Note sul tema della riduzione del debito estero dei paesi poveri, "Finance and Development" Working paper, CeSPI, Roma, Dicembre. 340 Conclusions Debt per se is not bad so long as it is properly invested and managed to bring about rapid economic growth and social development which would not have occurred had a country relied solely on its own resources. Loan contraction should enhance capacity to pay, and debt levels should be kept within manageable limits in order to maintain credit worthiness. When contracting additional loans, it is imperative for counties to adopt debt management policies, which ensure that the marginal returns to investment remain within the country. The provision of large investment capital suggested in the growth theories of Harrod (1939), Domar (1946) and Rostow (1956) has very often motivated proposals for foreign aid as means of bridging the resource gaps created by ambitious production targets in the developing world. External debt is regarded as one of the main economic ills facing Sub-Saharan Africa in its current stage of development. Foreign loans and aid, which spurred growth during the sixties and early seventies, have turned into a nightmare of debt, constantly forcing countries to initiate relief measures. Since 1980 total external debt has grown faster than any region, mounting to a crisis level which has since drawn international attention. The crisis has taken a heavy toll in human terms, depriving countries the necessary external resources needed for growth and social development. Lamentably, debt burdens are continuing to rise under weak economic performance and uncertainty increase in net transfers. Furthermore, debt burdens are also mounting at the time countries are undertaking difficult policy reforms under Structural Adjustment Programs. While present lending is linked to policy reform initiatives, there are still formidable obstacles facing Sub-Saharan countries which preclude many from achieving the rapid growth needed for debt reduction. Sub-Sahara Africa total external debt now stands at a staggering level of $223 billion. This amount is negligible compared with over $2.5 trillion owed by all Third World countries. However, its implications are serious for SSA region that is experiencing high population growth, high debt burden indicators, low and stagnating per capita income and human development as well as faltering GDP and export growth rates. SSA per capita debt of about $400 outpaces the region’s per capita income of $340. A large share of total debt (65%) is owed to official creditors at concessional rates while the remaining 35% is owed to private sources. A higher proportion of bilateral aid (loans and grants) is tied in one form or the other. Donor nations still prefer to channel most of their aid to former colonies. The region’s commercial debt not only trades at very low rates in the secondary market but have also received less attention as a remedy of Africa’s debt problems. Although debt burden is not uniform across countries, low-income African nations have been hit hardest because of their relatively weak economic position. Several internal and external factors have been blamed for the causes of African debt crises, especially the OPEC oil shocks and weak commodity prices which brought about drastic changes in international lending and borrowing policies of many African countries. International efforts to curb the region’s crippling debt have not been very successful in bringing debt levels of many countries to manageable levels. While these measures continue to provide short-term palliatives, countries are obliged to cope with short-term nonnegotiable payments to the revolving fund.of IMF. Long term solutions to Africa’s debt problems calls for a concerted and coherent effort by donor nations, the World bank and African countries to bring about substantial reduction in African debt and inducing favourable trade, financial, security environments. Commercial debt is a small proportion of the region’s total debt and does not immediately constitute a threat to the international financial system. Unless sufficiently addressed, Africa’s continued debt overhang raises fears of the region being trapped in a “vicious cycle” of debt and permanent reliance on external assistance. The debt-export ratio is also a “liquidity indicator” relating total external debt to exports. Debt-GNP ratio is a “solvency indicator” which evaluates a country’s ability to service debt on the basis of its national income assuming convertibility of currencies. SSA countries’ situation is veryy bad considering both these two indicators. And paradoxically, as a matter of fact, Africa’s military expenditure as a percentage of GDP was the highest in the world for most of the 1970s and 1980s. Unwarranted military expenditures were 341 not only uneconomical, but caused internal havoc and destabilized many countries, including among others Angola, Ethiopia, Mozambique, Somalia, Sudan and Zambia (for Africa as a whole, the ratio for 1977 was about 4.7% and dropped only to 4.3% in 1987) 1 . In sum, it is true that, at a general level of investigation, there are some important common characteristics of African indebtedness. Sub-Sahara Africa debt differs fundamentally in many respects from those of other Third World countries, especially the highly indebted countries of Latin America and the Caribbean (LAC). External debts of African countries are bilateral and multilateral in nature, contracted mainly on concessional terms - low interest rates with long repayment and grace periods. They are mostly public loans guaranteed by respective governments for various public development projects. In contrast, Latin American debts are from private commercial banks in creditor nations, acquired at market rates of interest with short repayment and grace periods. Although governments guaranteed most of the loans, the major customers were private corporations and individuals in Latin America who borrowed to finance private business ventures. Hence, the two types of debts can be distinguished as “publicly guaranteed versus privately guaranteed loans. Sub-Saharan Africa is less indebted compared to LAC but it is the most debt burdened region of the Third World. Its current total external debt, which is equivalent to the combined debts of Brazil and Argentina, constitutes about 11% of Third World debt. In contrast, LAC is the most indebted region of the world holding about 30% of Third World debt, approximately ten times that of SSA. Although the debt-service ratios of both regions have declined since 1988, those of LAC have been higher largely because of the high interest rates on their debts. Because of strong export performance of LAC, the debt-export ratios has continued to decline since 1987 and are currently below those for SSA. The debt-GNP ratios for LAC have also been declining since 1987 and are much below those of SSA. Though countries like Nigeria borrowed heavily from private commercial banks, capital flight has not been the major reason for Africa’s debt problems. The main cause of Sub-Sahara Africa debt problems lies more with the interaction between internal and external factors. At the beginning of the third Millennium, marked as it has been by enormous political, technological, and economic upheavals that are no more than vaguely delimited by the keyword “globalisation”, we are faced with the question of how to permanently overcome poverty on a world-wide basis. Globalisation has been given different definitions by different people. In any case, it has generally been assumed that globalisation has helped spur economic growth throughout most of the world. However, in terms of the practical experiences in many SSA countries, the definition relevant to our circumstances is another. The official data for the last two decades (19802000) remind us a different story. Economic growth has slowed dramatically, especially in the less developed countries, as compared with the previous two decades (1960-1980). For example: • From 1960-1980, output per person grew by an average, among countries, of 83%. For 19802000, the average growth of output per person was 33%. • Eighty-nine countries – 77%, or more than three-fourths – saw their per capita rate of growth fall by at least five percentage points from the period (1960-1980) to the period (1980-2000). Only 14 countries – 13% – saw their per capita rate of growth rise by that much from (19601980) to (1980-2000). The Center for Economic and Policy Research (CEPR) has recently released a report titled, “The Emperor Has No Growth: Declining Economic Growth Rates in the Era of Globalisation” written by Mark Weisbrot, Robert Naiman, and Joyce Kim of CEPR. The report illustrates how globalisation and the policies of the IMF and the World Bank, have failed to bring about economic growth in the developing countries in which they promote their macroeconomic policies. It concludes by asserting that we should be very cautious in pretending that someone has the 1 Source: U.S. Arms Control and Disarmament Agency (1989), World Military Expenditures and Transfers, Washington, D.C. 342 necessary expertise or answers to the difficult and often country-specific problems of economic growth and development. The uncertainty that pervades financial markets in the early months of 2002 should shift to cautious conclusions on the natural and immediate benefits of globalisation, particularly referring to SSA countries. Some main conclusions emerge from our study. First, the problem of highly indebted countries is not a new one, both in theory and in practice. From the two Greek city-states who defaulted on loans from the Delos Temple in the fourth century BC to Mexico’s default on its first foreign loan after independence in 1827 to Haiti’s 1890s high ratio of debt to exports and to the German crisis of post-war repayments in the first half of the XX century, debt servicing difficulties have been a feature of the world economy throughout history. And we should remind that during the 1800s, Britain extended huge loans to the United States to finance industrial development and expand its capitalistic state. The United States later turned into a creditor country after World War II, and through the Marshall plan aid provided loans that helped rebuild Europe from the ravages of war. In both these cases, debt settlements were made only after long negotiations that resulted in concessional rescheduling, payment in kind and eventually large amounts of debt forgiveness. Thus, debt crises are not new and debt is not bad by itself. Even though development economics emerged after the II World War to confront the causes of poverty in developing countries and to define strategies for economic progress, nevertheless we showed how old and profound is the relationship between foreign finance and theory of development process. Since the beginning of the theoretical debate, it was clear how complex is this relationship. In particular, we emphasised the importance of fiscal components of external debt problem, the so called three-gaps model, which is considered a peculiarity of SSA case. But we also stresses the complexity of factors that interact, as demonstrated by the fungibility argument and the Dutch disease. The complexity and the presence of counteracting effects seem to be the real nature of development and debt linkages, when we analyse in detail economic literature. Schematic and simple relations must be abandoned in favour of case-by-case analysis. Second, the idea of foreign debt sustainability, conducting the discussion in economic and financial as well as political terms, is much more complex and ambiguous than it may appear. And this relates to the question of debt relief. In fact, it is very different to consider sustainability from the lenders point of view (how much debt relief can they afford?) than from the borrower countries (how much debt repayment can they guarantee?). Sustainability is complex as complex is the map of lenders: we know that the most important component of SSA foreign debt burden is long-term debt outstanding, that the next important category is short-term debt outstanding, followed by the use of IMF credit, which became more and more important in the 1980s, when structural adjustment and enhanced structural adjustment facilities became important components of flows to SSA. And another characteristics of the structure of the SSA debt is the changing pattern of its creditors. It can be generally be said that bilateral debt is the most important component of the total debt. Most of the increase in the 1980s reflects the impact of repeated rescheduling with interests being capitalised and compounded rather than flows of new money. Although, in general, the three types of flows (bilateral, multilateral and private) grew dramatically in the 1980s, the grow in multilateral flows was relatively large. And contrary to common belief, it is interesting to note that private debt is a serious problem, even more serious in the 1990s in the West and Central Africa. The most significant component of such debt is suppliers’ credit and bank credit covered by Export credit agencies, such as the Italian case demonstrated. A larger share of the official debt is on concessional terms. The level of concessionality is the highest for East and Southern Africa, but there has been a growing trend towards non- concessionality. Such non- concessionality is the highest for West and 343 Central Africa. Besides, the SSA debt has increasingly been characterised by the importance of interest and principal arrears in shaping its level. Beyond prevailing rhetoric of international initiatives, which affirm to make external debt sustainable (as the current multilateral initiative says), differences among countries are important, the nature and composition of debt are important, specific history and context do matter, the interests and motivation and bargaining power of different lenders and debtors are crucial. There is no “scientific” reason to define a general threshold of debt sustainability or to identify 41 countries eligible to debt relief initiatives (as the current HIPC initiative does), rather than concrete, legitimate political will. The implications of approaching debt sustainability from a development and human rights bias, rather than from the fiscal bias of the present debt cancellation proposals are impressive. So far, the international community has proposed to write off about $100 billion of the $2465 billion developing country debt. A “rights based approach” would require the writing off of perhaps $500 billion in debt owed by 66 countries that cannot afford to pay their full debt service and still meet development and human rights targets to which the international community is already committed. Using historic precedents would require the writing off of $900 billion or more. Third, the characteristics of African problems stress the importance of contextual and specific national and historical aspects, within the same thematic issue of debt crisis. Heterogeneity does matter. It can be argued that Africa’s external finance problem is the result of the structure of its trade in the world economy in general and its place as a commodity producer in particular. The general African economies’ performance since the time of political independence can only be classified as dismal. There has been a secular deterioration of their terms of trade, dependency on foreign capital (debt and aid) has also grown at alarming rates exasperated by stagnation in exports, the level of investment is extremely low, physical and social infrastructures are deteriorating rapidly. These are aggravated by political instability and frequent wars, and they are inter-linked to wide corruption, wrong economic policies, serious responsibilities of inadequate local élites. However, again, specific context and differences do matter a lot. The origin of external debt can be traced to the willingness of the debtor country to borrow and the lenders to lend. From the debtor country’s perspective the need and decisions arised from given contexts, and it is also important to keep in mind that external debt was also driven by the pressure and willingness of lenders to lend – particularly the IMF and the World Bank, in SSA case -. Fourth, as a direct consequence of the preceding point, we need to go beyond the averages. Beneath the aggregate numbers exists a variety of experiences. And it does imply relevant consequences in terms of econometric analyses. Most work in the growth literature relies simply on cross-country regressions. However, we don’t consider cross-country regression as a reliable method of empirical argumentation relating to African development. We think that the most compelling evidence on these issues can come from careful and patient in-depth case studies of individual countries. Weak theoretical foundation and specification of functional forms for the relationship, poor quality of data bases and errors and biases of measurement, inappropriate econometric methodologies are common problems. In the context of relationships that have both a temporal and cross-sectional dimensions, there is the problem that the estimated impact from a cross-section need not be the same as that from time-series data. Regressions and their conclusions are strongly dependent on the period, sample of countries, and variables chosen. Thus, we adopted different techniques, in order to demonstrate the importance of the heterogeneity across countries and the interactions across channels of capital inflows in terms of development. In fact, our basic idea is that the relationship between external debt – and, in general, foreign capital flows – and economic development is complex and non-linear. It does clearly depend on the specific country’s context (a cross-section component) and on the specific past history (a time-series component). Moreover, and we think that it does represent the bulk of this work and the implicit mainstreaming of all the chapters, we 344 disagree with the conventional wisdom, which attributes to some kind of foreign capital inflows to be good or bad in itself for development. An implicit recommendation of this conventional wisdom is that a large share of FDI in total capital inflows is a measure of something good happening in the economy, as well as a large share of debt is bad. We think that the real effects are associated with the nature, size and composition of each capital flow, in addition to the cross-section and timeseries components, and above all with the interaction component among different kind of foreign capital inflows. We believe that foreign capital inflows enter multiplicatively rather than additively into a development equation (again, the specific analysis of Italian experience, in terms of aid and debt relief policies, very clearly and dramatically confirmed it). It has direct policy implications. First, it is hard to argue that the rise in the share of external debt is an indication of good health. But this does not mean that the rise in debt is bad in itself. Hence, there is no reason to say that, in the next future, debt must be totally replaced by grants – as it was recently and authoritatively affirmed by the US Congress - or that in perspective, FDI should be the bulk of foreign capital inflows, for the sake of private sector promotion. Fifth, from the beginning we admitted that “class” is a concept to be reconsidered, because the important changes of the organisation and division of labour and the very specificity of developing countries’ context should imply a reconsideration of the concept of class as it was categorised by Marx. But, concerning the appropriateness of the units of investigation, we stress that countries are not necessarily the best units to analyze economic growth, particularly in the new context of globalization. The importance of both the local development and territorial context from one side, and de-localization of economies emphasize other dimensions and borders than national ones. Particularly in Africa, we can use concepts as enclaves, export-oriented and cash-crops sectors. We have also to remind the importance of informal sector and dual economies. Even concerning external debt, we have to recognise that we have a map of multiple actors (multilateral, bilateral, private lenders, governments, countries, institutions, organisations, people,…), interests and preferences (political, strategic, economic, environmental,…), each of one having different time horizons (long-term or short-term horizons). We seriously question whether the country entity is the most appropriate unit and dimension to be investigated. Sixth, in terms of political implications, debt relief seems to demonstrate the emergence of international political consensus and capacity to implement a win-win strategy, which will satisfy debtor countries as well as most of the constituencies based in developed countries. But apart from the rhetoric widely spread all over the world on debt cancellation, international politics offers no quick panacea for the myriad social interests, which are involved in external debt and aid game. We stress the importance of holistic visions of change and development, as those which can be referred (even if they are not explicitly assumed as a system) to international civil society. Our study tried to provide a systematic analysis of the economic rationale of external debt for development of poor countries, to analyse debt crisis in SSA and to review the extent, the effectiveness and the future prospects for current debt relief initiatives (HIPC Initiative, above all) and also to explore the vision, expressed as a set of recommendations, of international civil society, which has been seriously involved in the search for solutions of debt crisis. We presented an original synthesis of a lot of separate political requests from civil society organisations that can be reduced at coherent and comprehensive approach. This is particularly important as, in the context of a multi-players game such as debt-game is, we think that those organisations, which can be called international civil society and are deeply interested in the globalisation issues, can not be reduced to the grotesque image of the “no-global people” (a terms implying the negative idea of those who simply raise objections, who have their destruens pars without any concrete vision and proposal to submit to political arena). We proposed an integrated vision, which would like to be an indirect expression of the capacity of numerous organisations of civil society to engage in debate and comparison on the content, which is a reflection of the skill and know-how accumulated with regard to the issues of 345 international relations. Undoubtedly, civil society organisations constitute an innovation on the scene of international relations. In the multifaceted world of civil society there co-exist specific and widespread interests, at local, national and transnational level. We emphasised the theoretical foundations of a vision belonging to an important constituency, a broad-based movement, politically expressive and capable of a more effective critical approach to the HIPC Initiative 2 . At the same time, and this is what we stress, international civil society has a strong capacity in the costruens pars, that is they unquestionable prove the maturity of the proposals. These proposals are an important political counter-part to the economic analyses we presented. It is not necessarily based on the theoretical assumptions proposed in the economic literature, but it represents an effective lobbying campaign to be taken in account, which should be seriously assessed. The international civil society has become an increasingly powerful influence on international relations; its political contribution to international finance and development co-operation has to be considered. Seventh, and linked to the international civil society’s proposals, to analyse the relationship between debt and development in Africa as well as current debt relief initiatives is a way to analyse development co-operation, within the main context of finance for development. This is the bulk of Italian case-study (presented in chapter 6, Appendix 2), and is the final result of our work. It is important to link debt relief initiatives to aid policies, trade regimes, financial systems and national economic and social policies. Aid and debt effects are a complex matter, and selectivity in foreign capital allocation is a difficult task, if defined in general terms. The aid and debt effectiveness depends on the external and climatic environment: the worse this environment, or the more vulnerable the recipients countries, the higher the aid effectiveness. Many SSA countries continue to undergo both trade and current account deficits; and debt service provides a major additional impediment. In fact, both the trade balance and the current account have always been negative, but the trade deficit is usually much smaller than the current account deficit. This is not just a trade problem; even with a trade account on balance or in the positive, this group of countries would have had negative current account, mainly because of the external debt service. SSA countries have a negative, large and increasing net factor income. Aid has substantially increased from the eighties to the nineties; but is smaller than debt service, and figures seem to indicate that aid has been used to service the debt, at least partially. Nevertheless, everyone repeats that aid has to be fully employed to undertake a process of social and economic transformation instead of being used for repaying past debt. Apart from aid, no forms of external finance have substantially contributed to the needs of external finance of these countries, neither commercial loans, nor FDI, nor Portfolio Investments as can be seen from the composition of the flows to SSA countries in the nineties. We have to adopt a political economy approach, being concerned with the interface of politics and economics – the influences on economic policy rather than political systems per se. Thus, the focus must be on how aid (and the associated relations with donors), especially in relation to debt relief, interacts with the political factors that influence the policy process. The evidence of the 1990s gives ground for cautious considerations. There is no simple answer to this question, given the real-world complexity of recent African history. the Dollar and Kraay3 view, that “anyone who cares about the poor should favor the growth-enhancing policies of good rule of law, fiscal discipline, and openness to 2 In a very broad sense, some doubts on the real effectiveness of the HIPC initiative are spread all over the costituencies. «The enhancement of the initiative for highly indebted poor countries will reduce the indebtedness of those countries that meet the scheme’s conditions. However, most will still be left with a large debt burden, which could become unmanageable should there be large fluctuations in prices. The IMF and World Bank admitted as much in a recent paper. A large write-off of IMF and World Bank debt will happen only if there is pressure from the institutions’ shareholders», Financial Times, 26 April 2001. 3 Dollar, David and Aart Kraay (2000), “Growth is Good for the Poor”, World Bank, Development Research Group, mimeo, March 346 international trade” as well as the opposite view that the growth reforms induced increasing inequality, denying benefits to the poorest 4 , are wrong. Freedom and democracy – as Julius Nyerere wrote – are not commodities which can be just lifted from a shelf and given life, in a easy way. Moreover, after two decades of economic and social crises in SSA, to increase output and exports diversification needs time and investment, to increase the saving ratio through tax reform and a more accountable credit and financial system also requires time and investment, to achieve high rates of both capital accumulation and human development requires time and investment. Given the above considerations finance for development will remain an absolute necessity, even if all the foreign debt should be cancelled, which is not the case yet. Obviously, this overview and these recommendations are a stage in a process, not a final product. They are presented here in a spirit of discussion, debate, correction and improvement. Now, the crucial question is: is there some room for hope that the development failures of the past decades, combined with new interest in the concept of global public goods (from international health and security, all the way to financial stability) provide strong arguments in favor of SSA development? When world leaders met in September 2000 to commit to the Millennium Development Goals, a horizon of optimism lit up. Only 18 months later, approaching the Global Summit on Financing for Development taking place 18-22 March 2002 in Monterrey, Mexico, hopes were fading that the world leaders take their commitments seriously. The Monterrey Consensus that heads of states and governments signed after two years of debate and negotiations, did not contain any concrete commitment to raise the finances needed to half the number of the world’s poor by 2015, as envisaged by the Millennium Development Goals. It also failed to address the many structural problems that prevent the realisation of a more equitable and inclusive economic globalisation. Instead, it just repeats the unproven promise of the last two decades that further trade and investment liberalisation will enable the private sector to take care of the world’s poor. But evidence is mounting that this idea fails to deliver wealth and economic justice for all. At the beginning of 2002, international NGOs demand for comprehensive solutions to external debt problems of poor countries and they are also proposing a reformed procedure between debtors and creditors which would not only overcome this structural and ethically questionable imbalance, but would also be likely to lead to a more efficient debt management: (1) A neutral decision making body, which is independent from both parties involved; (2) The right of all stakeholders to be heard before a decision is made; (3) The protection of the debtor’s – in this case the most vulnerable sectors of a sovereign debtor’s society – basic needs, before debts are collected. (4) The institution of an automatic stay, once an international »insolvency” or »arbitration” case is opened, in order to avoid a creditors’ run on the debtor’s remaining assets and to allow for an orderly procedure. Nowadays, this is one of the most debated issues, but it does make sense if and only if it is part of a more general global development partnership agreement, extended to the main aspects of international relations. A lot of international high level conferences and summits risk to be journalistic events with abstract outcomes. The important commitments must not remain pure lip-service, they have to be translated into concrete political actions. The financing for development challenge offers a crucial opportunity to give a signal of global change. Governments and international organisations have to move beyond the !agreed language” and to take credible steps towards a new North-South partnership, based on more resources of better quality than in the past and more coherent policies. 4 Forsyth, Justin (2000), Letter to The Economist, June 20. 347 To conclude, a sort of anecdote. When we were going to conclude our work, we discovered and read two papers recently published, which shared the same idea we have assumed in a part of our econometric analysis. These are the work of Ricardo Hausmann5 , from the Harvard University (concerning the critique to the rhetorical assumptions on the benefits due to FDI), and of Henrik Hansen6 (on the importance of interactions between debt and aid), from the University of Copenhagen. This fact simply confirmed to us that the room for original ideas is – as it always is much more constrained than we supposed at the beginning of this research, two years ago. And, directly or indirectly, what we think and write is basically the result of what we listened, read and learnt. 5 Development Centre Seminars (2001), Foreign Direct Investment Versus Other Flows to Latin America, Oecd- Iadb, Paris. 6 H. Hansen (2001), “The Impact of Aid and External Debt on Growth and Investment: Insights From Cross-Country Regression Analysis”, WIDER Conference on Debt Relief, Helsinki, August. 348

Log In

International Financing for Development Co-operation On the linkages between foreign capital inflows and development in poor countries

International Financing for Development Co-operation On the linkages between foreign capital inflows and development in poor countries

Related Papers

RELATED PAPERS