Geoforum 37 (2006) 906–920 www.elsevier.com/locate/geoforum Sustainable lifestyles: Framing environmental action in and around the home Stewart Barr ¤, Andrew Gilg Department of Geography, University of Exeter, Amory Building, Rennes Drive, Exeter EX4 4RJ, United Kingdom Received 7 October 2005; received in revised form 27 April 2006 Abstract This paper examines the nature of environmental action in and around the home. Given the rise of local sustainable development and the emphasis placed on individual actions for sustainability, the paper examines the role of citizens in adopting sustainable lifestyles, incorporating a range of behavioural responses from energy saving and water conservation, to waste recycling and green consumption. Focussing on the debate in geography concerning the engagement of the public in environmental action, the paper argues that despite the assertions of those who advocate a deliberative approach to engagement (see [Owens, S., 2000. Engaging the public: information and deliberation in environmental policy. Environment and Planning A 32, 1141–1148]), an approach based on a social–psychological understanding of behaviour can have signiWcant beneWts. Such an approach is being developed by geographers in a range of settings and in this paper these developments are situated within the context of existing research that has identiWed environmental ‘activists’ in terms of their values, attitudes and demographic composition. The paper aims to examine environmental behaviour in relation to two key issues: (1) the way in which environmental action is framed in everyday practices (such as consumption behaviour) and (2) how these practices are reXected amongst diVerent segments of the population to form lifestyle groups. The paper provides new insights for examining sustainable lifestyles that further our appreciation of how actions to help the environment are lived in everyday practices and framed by diVerent lifestyle groups. Accordingly, the paper oVers both academics and policy makers new insights into the potential use of focussing on lifestyle groups as a means for changing behaviour.  2006 Elsevier Ltd. All rights reserved. Keywords: Sustainable development; Environmental action; Everyday practices; Lifestyle groups; Environmental policy 1. Introduction: environmental action, sustainable development and policy Publication of the British Government’s Sustainable Development Strategy (DEFRA, 2005) has reignited the policy debate surrounding the role of individual citizens in ameliorating environmental problems caused by overconsumption and the so-called ‘throw away’ society. In line with many developed nations, the British Government’s Sustainable Development Strategy (DEFRA, 2005, p. 25) * Corresponding author. E-mail address: S.W.Barr@exeter.ac.uk (S. Barr). 0016-7185/$ - see front matter  2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.geoforum.2006.05.002 places individual actions for sustainable development at the heart of its policies for eVecting change: “Behaviour changes will be needed to deliver sustainable development. However, attitude and behaviour change is a complex subject. Information alone does not lead to behaviour change or close the so-called ‘attitude-behaviour gap’ ƒ One of the key elements of the new approach is the need to engage people close to home.” Whilst the British Government’s Sustainable Development Strategy rightly deals with the apparent complexities of behavioural change, there are two signiWcant omissions in the underlying assumptions made in the Strategy. First, there continues a conventional deWnition of environmental S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 action, which is focussed around speciWc activities, such as recycling, saving energy or conserving water. In reality, it is unlikely that individuals conform to these highly compartmentalised behavioural patterns, but that the beneWts of speciWc actions will have a positive impact on a range of environmental problems. Second, in proposing new policies to eVect behavioural change, there is an implicit assumption that messages to engage the population need to be based around information and awareness. Yet this takes little account of the ways in which diVerent lifestyle groups utilise this information or indeed how their perspective on environmental action varies (Owens, 2000). This paper therefore aims to address these two issues by using data from a large research project in Devon, south west England, to examine: 1. the ways in which environmental action is constructed in everyday life and related to everyday practices; and 2. the extent to which there are identiWable groups of individuals with diVerent behavioural properties that exemplify alternative environmental lifestyles and consequently form lifestyle groups. 2. Environmental action: values, attitudes and behaviour Geographers have engaged widely with sustainability, examining the structural and socio-political processes by which sustainability has become incorporated into social life (e.g. Eden, 1993; Munton, 1997; Gibbs et al., 1998; Blake, 1999; Bulkeley, 1999). They have also examined public attitudes towards sustainability (e.g. Burgess et al., 1998; Hobson, 2001, 2002). To these researchers, the very notion of what constitutes environmental action is problematised within a wider political discourse that has become disconnected with society more generally. Pre-determined actions set by national governments and promoted as being ‘sustainable’ are therefore constructed in ways that are not reXected in everyday social and environmental concerns. Within this context, work in geography on eVecting behavioural change has focused around what has been termed a “deliberative” model of public engagement with sustainability (Owens, 2000, p. 1141) and the ways in which individuals receive, interpret and act on environmental information in a range of discursive and institutional contexts. The deliberative model also proposes that engagement of the wider public will be forthcoming only when social and environmental problems are framed and debated within the spatial and temporal scales at which individuals are expected to take action (Blake, 1999). Accordingly, Hobson (2002, p. 113) argues that environmental action based on “voluntary information and lifestyle initiatives will constantly create ‘discursive traps’ ƒ by information presented in impersonal media”. Participation in action for sustainable development is therefore seen as contingent on a range of factors, relating to the nature of the information provided and its interpretation (Myers and MacNaghton, 1998), based upon the 907 trust with which it can be handled (MacNaghten and Jacobs, 1997; Hobson, 2001) and the complexity of scientiWc information (Eden, 1996; MacNaghten and Urry, 1998) provided by ‘experts’. Overall, as Owens (2000, pp. 1141–1142) noted, this ‘civic’ approach to examining public engagement with sustainability “ƒ is less prescriptive of information Xow and admits a wider range of understandings into the category of ‘expertise’ ” such that “ƒ what is sought here is democratic engagement ƒ moving beyond the prescribed responses to predeWned problems”. This democratic engagement has the aim not merely of identifying solutions to problems, but also to ‘re-frame’ the nature of these challenges. Most crucially, however, this approach is rooted in the belief that policy discourses are strongly focussed on the “rationalisation” of environmental action, which assumes that environmental behaviour can be changed by enhancing knowledge and awareness about environmental problems (Hobson, 2002, p. 95). Accordingly, upon learning about a speciWc environmental problem, the rationalisation argument posits that “ƒ as a result of this new knowledge, individuals will change their consumption behaviour” (Hobson, 2002, p. 96). Owens (2000) conceptualises the rationalisation argument within the context of what she terms ‘information deWcit’ models of behaviour change. Using this framework, understanding and changing public behaviours towards the environment is seen as a process of Wlling a ‘value–action gap’ (Burgess et al., 1998) with information. As Owens (2000, p. 1142) aptly puts it, surely “ƒ if people had more information about ƒ environmental risks, they would become more virtuous”. Geographers have therefore argued that information-intensive campaigns are likely to be unproductive given the fundamental assumptions made regarding the deWnition and operationalisation of environmental action. These studies, primarily drawn from the geographical literature, propose one epistemological framework that has sought to focus on the cultural and institutional contexts within which behaviour is framed. Behaviour change is therefore seen as dependant on deliberative and inclusionary processes. In contrast, evidence suggests that policy continues to approach the issues of environmental action and behaviour change from what Hobson (2002) terms a ‘rationalistic’ perspective, emphasising knowledge and awareness of speciWc environmental problems as the mode by which to eVect change and close the ‘value–action’ gap. In policy terms, environmental action is framed by existing environmental issues that confront the state, such as ‘energy’ or ‘water’ crises. The determination of how such issues can be addressed is presented to citizens both at varying spatial and temporal scales, through the mode of information transfer which encourages individuals to ‘do their bit’. The divide between these two approaches could therefore be framed solely in epistemological terms, but the perspectives adopted reXect an underlying but signiWcant divergence in methodological terms, most crucially concern- 908 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 ing how environmental action is framed, interpreted, explained and changed. As Owens (2000, p. 1143) notes, the problem that geographers have identiWed with the ‘information deWcit’ framework has partly been down to the “ƒ ‘mental models’ approach” taken by quantitative researchers which “still suVers from too rationalistic a conception of agency”. In this context, the critique challenges the notion that behaviour can be understood by examining what impact information has on action. The argument does, in many cases (e.g. MacNaghten and Urry, 1998) proceed further in many cases, to critique the epistemological framework of attempting to understand behaviour using ‘models’, which has been the dominant approach adopted by social psychologists and quantitative sociologists over the past 40 years. This is where an approach which can address the divide between these epistemological and methodological arguments can be examined. It is certainly the case that the development of social–psychological frameworks of behaviour have examined the impact of knowledge on action using quantitative methods, but 40 years of research into environmental behaviour by social-psychologists and quantitative sociologists has provided a wealth of material that should not be dismissed as taking an uncritical view of how humans act towards nature. This is a theme that has been taken forward by geographers such as Wilson (1996), Burton (2004) and Burton and Wilson (2006) in their studies of farmer attitudes and behaviour. These geographers have adopted this alternative epistemological and methodological perspective in order to address the issue of behaviour change for sustainable development. Implicit in their assumptions is an appreciation that behaviour is complex, but that the means by which to understand behaviour and potentially to eVect change is to provide policy makers with information about what the nature and structure of environmental action is, what appears to inXuence it and who the dominant participants are. Using these assumptions, research has sought to conceptualise environmental action by studying policy-deWned behaviours, but using an approach that problematises their deWnition and frames them within everyday practices. Indeed, such work seeks to understand environmental behaviour within the wider context of research that has examined a range of social behaviours using psychological frameworks to reWne understandings of behaviour. Although a range of ‘models’ of environmental action exist within this literature, by far the most signiWcant and most eVective example of this approach is the theory of planned behaviour (TPB) (Ajzen, 1991), derived from the earlier theory of reasoned action (Fishbein and Ajzen, 1975). The TPB conceptualises social behaviour as being contingent on the two major inXuences of ‘intention to act’ and ‘perceived behavioural control’ (or ‘how able’ an individual feels to act). Intentions are seen as the result of a combination of subjective norms and attitudes towards the behaviour. Although researchers have utilised this model in work on environmental action (e.g. Boldero, 1995; Chan, 1998), there has been general dissatis- faction that other factors have been neglected and this has led to a proliferation of adaptations to the TPB, which have still maintained the crucial intention-behaviour link conceptually, but which have sought to examine the role of other key variables (e.g. Lam, 1999; Taylor and Todd, 1997). The methodological assumption underlying this work is that behavioural complexity can be examined by quantitative methods (such as questionnaire surveys or observation), which use scales and measures of social and psychological constructs which have been tested, reWned and subjected to reliability analyses. This approach also argues that existing theoretical constructions of attitudes and behaviour can provide the basis for understanding and eVecting behavioural change through targeting policies which address speciWc attitudes and values within deWned social groups. Taking this alternative epistemological and methodological perspective, we return to our research questions posed at the beginning of the paper. These related to how environmental lifestyles were framed around daily practices and the diVerent lifestyle groups that might be identiWed around these practices. To address these themes, researchers from a social–psychological perspective have focussed their work on the role of three key factors that have been shown to diVerentiate between diVerent levels of commitment to environmental action, which can help us to understand how environmental lifestyles are constructed in everyday life. These factors have been conceptually developed over almost 40 years of research by social-psychologists and sociologists and can be categorised into: the situational circumstances in which individuals are placed (including socio-demographic situation), the socio-environmental values individuals hold, and attitudes towards speciWc behaviours. 3. An alternative perspective on environmental action Focussing initially on the situational factors that inXuence behaviour, considerable attention in policy circles has focussed on deWning environmental action in terms of the social composition of diVerent behavioural groups. Such socio-demographic diVerences have been noted by a range of researchers in this Weld, notably through meta-analyses (e.g. Hines et al., 1987; Schultz et al., 1995). This research has suggested that social composition can be signiWcant in diVerentiating between levels of environmental action and relates not merely to demographic proWle, but also structural characteristics (such as recycling provision) and situational factors (for example, the knowledge of environmental issues and facilities for helping the environment). These relationships are not necessarily linear (e.g. Derksen and Gartell, 1993; Guagnano et al., 1995) and debate continues concerning the role of structural measures and policy instruments in inXuencing behaviour (Gilg, 2005). Despite the focus of policy on diVerentiating according to social composition, the role of underlying social and environmental values must also be emphasised. As underlying S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 guiding principles in peoples’ lives, values have been strongly related to behaviour. O’Riordan’s (1976, 1985) spectrum of ecocentric–technocentric environmental values emphasised the role of operational values, and this was initially used in a classiWcatory framework of ‘environmentalists’ in the 1970s. Indeed, Dunlap and Van Liere’s (1978) new environmental paradigm has been developed (Dunlap et al., 2000) to demonstrate the role that relational environmental values have in diVerentiating between biocentrists (viewing humans and nature as equal) and anthropocentrists (viewing humans as dominant over nature) in behaviour towards the environment (Vining and Ebreo, 1990, 1992). More recently, Stern et al. (1995) have focussed on the role that wider social values can play in distinguishing between groups of environmental actors. Using Schwartz and Blisky (1987) and Schwartz (1992) as their guide, they argued that those most likely to undertake environmental actions were ‘altruists’ and ‘open to change’ on their continua, ranging from ‘altruism–egoism’ and ‘openness to change–conservatism’. This assertion has been partly supported more recently by both Cameron et al. (1998) and Corraliza and Berenguer (2000). A further area of work by social scientists has placed considerable focus on the identiWcation of attitudinal constructs that relate to environmental action. De Young’s (1986, 1996) seminal work on intrinsic motivations and personal satisfactions has highlighted the motivational aspects related to environmental action. He has consistently argued that participating in environmental behaviour engenders a feeling of satisfaction that yields both an inner sense of well-being alongside a belief that society is beneWting from one’s behaviour. This personalisation of environmental issues is signiWcant from both a positive motivational perspective, but also from a more negative standpoint. Researchers such as Kantola et al. (1983), Baldassare and Katz (1992) and Segun et al. (1998) have argued that internalizing environmental concerns and recognizing these as tangible threats to the self and a wider network of family and friends relates to higher levels of participation in environmental issues. These personal motivations to act evidently interplay with the beliefs an individual holds towards the Wnal outcome of their behaviour. High levels of response eYcacy, in particular a belief that a speciWc behaviour will have a tangible impact on the environment, has consistently been found to be a signiWcant factor in determining levels of environmental commitment (Seligman et al., 1979; Verhallen and Van Raaij, 1981; Samuelson and Biek, 1991; Roberts, 1996; Barr et al., 2001). Both Costanzo et al. (1986) and Stern (1992) have taken this argument further and argued that the extent to which individuals trust the environmental information provided to them, particularly by government, NGO’s and commercial organizations, is crucial to eliciting positive behavioural responses. In addition, individuals who believe that responsibility for environmental problems lies with themselves are more likely to be engaged in environmental practices (Nancarrow et al., 1996–1997; Mainieri et al., 1997). 909 However, eYcacy evidently relates not merely to a predicted outcome of the behaviour and one’s power to inXuence this, but also the self-eYcacy to undertake the activity. The extent to which individuals have a belief that they are able to carry out the behaviour, especially with regard to logistical elements, such as time and convenience, is a primary inXuence on behavioural commitment (Baalderjahn, 1988; Schwepker and Cornwell, 1991; Sparks and Shepherd, 1992; Chan, 1998; Barr et al., 2001). The perception of certain external factors, such as logistical concerns also relates to the role of other individuals in inXuencing behaviour. Of great signiWcance and highlighted in Fishbein and Ajzen’s (1975) original model, social inXuence and self-presentation has been tested in a range of research projects, with authors such as Berk et al. (1980), Sadalla and Krull (1995) and Lam (1999) emphasising the role of friends, relatives and other inXuential individuals in formulating and encouraging environmental behaviours, especially with regard to highly visible activities, such as consumption. Finally, there are a range of speciWc attitudinal constructs relating to particular behaviour, such as consumption attitudes. SpeciWcally here, issues relating to environmental trade-oVs and the importance of price, brand loyalty, buying locally and health and safety concerns have all been shown to inXuence environmental action (Mackenzie, 1990; Shrum et al., 1995; Roberts, 1996). This research, mainly from social-psychology, quantitative sociology and also business studies provides geographers with new insights into the factors that assist in an appreciation of environmental practices and lifestyles. It may provide a means by which to address the growing divide created by the divergence of the dominant research agenda in geography, focused on the civic and deliberative approach, and the ‘information deWcit’ model advocated by policy makers. The remainder of the paper examines how these concepts and related methodologies can be applied to an understanding of how environmental action is related to everyday practices and the potential lifestyle groups that emerge. 4. DeWning the modern environmental activist: a study of environmental action The research was undertaken from October 2001 to September 2003 and formed the major empirical element of an Economic and Social Research Council project on Environmental Action in and Around the Home (Gilg et al., 2003). It centred on a large questionnaire survey of 1600 households in Devon, a county in South West England, in the summer of 2002. The research aimed to examine the attitudes, values and reported environmental behaviour of a representative sample of the general public, incorporating the concepts reviewed above. Accordingly, the choice of a questionnaire instrument reXected the epistemological assumptions which underpinned the research and aimed to demonstrate how a quantitative methodological approach could yield generalisable data regarding the structure of environmental action and lifestyles. This sought to build on 910 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 the work of Wilson (1996) and Burton (2004) in setting a new agenda for quantitative work in geography that acknowledges the beneWts of using established measures for studying a range of social phenomena. To this end, the research selected questionnaire instruments over more conventional methods used in geography (such as focus groups and interviews) because of both the theoretical stance being taken and the desire to produce a generalisable data set comparable to existing work outside geography. Four speciWc environmental behaviours were examined (energy saving, water conservation, waste management and green consumption, in and around the home). During the process of questionnaire design, local policy makers were consulted on both the content and quality of questions posed, so as to ensure that, without compromising research goals, the results would be as useful as possible to local practitioners. After a period of consultation, a 14 page questionnaire emerged that measured 40 environmental actions, mostly on frequency scales, from ‘always’ to ‘never’. These included such activities as buying organic produce, using a water saving device in the toilet cistern and recycling household materials. In the most part, speciWc items were either amended from previous research or were drawn from the ‘recommended actions’ sections of local authority publications on environmental action. The remainder of the questionnaire posed questions relating to attitudes, values and demographic variables. The questionnaire was hand delivered to a series of preselected addresses in Plymouth, Exeter, Barnstaple and Mid Devon in order to survey a wide spectrum of environments, from inner city to rural locations. Four hundred questionnaires were delivered in each of these areas. Using a local market research Wrm, surveys were delivered to the selected household and if the occupier agreed to participate, they would collect the questionnaire two to three days later. However, if an individual declined to take part, the property next door was sampled until a response was achieved. The normal sequence would then be resumed. Households were selected randomly from the Electoral Register (Gray, 1971) in order to ensure an equal spatial distribution. 5. Results The response rate from the survey was 59%, representing those households who were pre-selected and returned the questionnaire. Seventy-nine per cent of those who agreed to participate in the survey (not from the originally selected sample) returned usable questionnaires (No. 1265). The results will be examined in relation to the two hypotheses presented in the introduction to the paper: 1. environmental action is constructed according to existing everyday practices, rather than being framed around speciWc environmentally-related practices; 2. environmental action can be examined according to lifestyle group, which will be representative of diVerences relating to socio-environmental values and attitudes. 5.1. Everyday practices: examining the structure of environmental action in and around the home To examine our Wrst hypothesis, we sought to examine how the four environmental activities that we had measured were related to each other. Of the 40 items measured in the questionnaire, 36 were measured according to frequency items (never–always). Previous research by the authors (Barr et al., 2001) found that activities such as waste management behaviour could be classiWed according to particular types of activity (waste minimisation, reuse and recycling). Using the same analytical techniques, the relationships between the four types of behaviour measured in this survey (green consumption, energy saving, waster conservation and recycling) were analysed. The links between frequency scores on a questionnaire can be examined using factor analysis. Factor analysis is a commonly used technique in social science, employed to examine the underlying structure of items in a questionnaire to assess whether there are signiWcant relationships between sets of items that can both enhance understanding of existing concepts and also structure the data for further analyses. An exploratory factor analysis was undertaken on all of the behavioural measures that were classed on 1–5 (never–always) scales. Table 1 provides the factor solution for these data, along with the amount of variance explained by each factor and the reliability of the scales formed from the relevant items. Items are recorded according to the initial behavioural classiWcation used. Three factors emerged from the analysis. A brief examination of the factors demonstrates that the Wrst hypothesis speciWed above can be supported in most part. The Wrst two factors comprise items from at least two of the conventional behavioural categories that would be expected. Factor 1 (purchase decisions) contains items relating to green consumption items, but also energy and water saving behaviours. However, it also relates to two waste management behaviour items (composting both kitchen and garden waste). This provides good evidence to suggest that our hypothesis relating to a more practice based classiWcation of behaviours is correct, since all four of the behaviours measured in the questionnaire were present. What all the behaviours have in common is the purchase related decision implicit in each activity. This is the case with respect to the apparently odd inclusion of the composting items. Composting as a behaviour relates not merely to the activity of placing food or green items into a composter, but in most cases involves the crucial decision to purchase a composter or wormery. In contrast to factor 1, factor 2 (habits) relates to more frequent, mostly daily, actions in the home. These are referred to as ‘habitual’ because of their regularity and their relationship to everyday and seemingly ‘mundane’ practices. In almost equal measure, these items are focussed on either water or energy saving activities. In some cases the habitual nature of such behaviours could be questioned, given that certain activities would on Wrst glance appear to 911 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 Table 1 Factor solution for behavioural data Factor Variables included Label for Figs. 1–4 Variance (eigenvalue) Per cent variance Cronbach’s Alpha Purchase decisions High eYciency bulbs Energy eYcient appliances Buy organic Buy FairTrade Avoid aerosols Compost garden waste Compost kitchen waste Avoid toxic detergents Reuse paper Reuse glass Buy recycled writing paper Buy recycled toilet paper Buy locally produced foods Buy from a local store Use own bag when shopping Less packaging Use plants that need less water EN7 EN8 GC3 GC10 GC2 W10 W9 GC1 W2 W1 GC9 GC8 GC6 GC7 GC5 GC4 WA8 4.4 13.3 .83 Habits Turn tap oV when washing dishes Turn oV tap when soaping up Reduce the number of baths/showers Reduce toilet Xushes Turn tap oV when cleaning teeth Use a shower rather than bath Reduce heat in unused rooms Reduce hot water temperature Keep heating low to save energy Wait until there’s a full load for washing More clothes instead of more heating Lights oV in unused rooms Use a sprinkler less in the garden WA5 WA6 WA3 WA2 WA4 WA1 EN1 EN2 EN6 EN3 EN4 EN5 WA7 3.9 11.7 .81 Recycling Recycle glass Recycle newspaper Recycle cans Recycle plastic bottles Donate furniture to charity Donate clothes to charity W3 W4 W5 W6 W7 W8 3.5 10.5 .78 Total variance 36 The Wnal three columns of the table relate to each of the three factors. reXect major change in routine. However, it is argued here that irrespective of the frequency, once the decision has been taken to participate in such activities around the home, they are likely to become part of routine and habit. Factor 3 is however very diVerent, with a focus on recycling and the treatment of post-consumer waste items. This factor is diVerent from the second (habitual) factor because of the nature of waste management activity in the study areas. In general, recycling in Devon is a highly structured activity, usually operationalised by the use of boxes or wheeled bins collected at the kerbside each week or fortnightly. Householders must be aware of the mechanics of the scheme and may have to sort waste accordingly. As such, this type of highly structured and regulated behaviour is in contrast both to purchasing and habitual activities. The data from the factor analyses provide evidence to support the assertion that environmental action is based around daily (and weekly and less frequent) practices. This Wnding has signiWcant implications for those who seek to provide a holistic conceptualisation of environmental action, but also for policy makers, who have conventionally used a campaigning approach based on encouraging diVerent forms of sectoral behaviour (such as energy or water saving). However, there is still the fundamental issue of whether everyday practices are reXected by diVerent types of lifestyle grouping. The second part of the analysis, therefore, examines how these diVerent environmental practices are related to diVerent groups of the population and how these groups can be deWned by their socio-environmental values and attitudes. 5.2. Identifying lifestyle groups for environmental action in and around the home Our Wrst hypothesis is partly corroborated by the data presented here. However, the second hypothesis sought to examine the extent to which deWnable groups of individuals 912 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 • • • • committed environmentalists (No. 294); mainstream environmentalists (No. 412); occasional environmentalists (No. 505); non-environmentalists (No. 43). Figs. 1–4 provide information on the frequency with which individuals participated in environmental action for each cluster. Based on these groupings, a descriptive analysis of the properties of these groups follows, with a deeper analysis of the extent to which these segments may form speciWc lifestyle categories. Committed environmentalists (Fig. 1) regularly engaged with the majority of the behaviours in all of the three factors (purchase, habitual and recycling). Energy saving and waste management behaviours were the most popular, with many in the clusters stating that they always or usually Never Rarely Sometimes Usually W8 W7 W6 W5 W3 EN5 WA7 EN4 EN3 EN6 EN2 EN1 Recycling WA1 WA4 WA2 WA3 WA6 WA5 WA8 GC4 GC5 GC7 GC6 GC8 Habits GC9 W2 W1 GC1 W9 W10 GC2 GC10 EN8 GC3 EN7 Purchase decisions W4 could be classiWed according to their behaviour and the extent to which this varies with lifestyle attributes. Social scientists have developed a series of analytical techniques for classifying individual behaviours. Of these, cluster analysis provides the most suitable means by which to deal with frequency data. Cluster analysis examines the scores individuals provide on a series of questionnaire items. Each individual is paired to another with the most similar score, a process that continues until there is just one ‘cluster’ remaining. The crucial (subjective) decision is how many clusters of individuals to select. This is normally undertaken by using a dendrogram of the pairing process and ‘cutting’ the graph at an appropriate point to result in a series of groups. In the case of the current work, four major clusters were identiWed and were classiWed as follows: Always Fig. 1. Variables in cluster 1 (for item labels, see Table 1) “Committed environmentalists”. Frequency bars refer to the percentage of respondents involved in the speciWc behaviour. Never Rarely Sometimes Usually W8 W7 W6 W5 W4 W3 WA7 EN5 EN4 EN3 EN6 EN2 Recycling EN1 WA1 WA4 WA2 WA3 WA6 WA5 WA8 GC4 GC5 GC7 GC6 GC8 Habits GC9 W2 W1 GC1 W9 W10 GC2 GC10 EN8 GC3 EN7 Purchase decisions Always Fig. 2. Cluster 2 “Mainstream environmentalists” (see Table 1 for labels). Frequency bars refer to the percentage of respondents involved in the speciWc behaviour. 913 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 Never Rarely Sometimes W8 W7 W6 W5 W4 W3 WA7 EN5 EN4 EN3 EN6 EN1 Usually EN2 Recycling WA1 WA4 WA2 WA3 WA6 WA5 WA8 GC4 GC5 GC7 GC6 GC8 W2 Habits GC9 W1 GC1 W9 W10 GC2 GC3 GC10 EN8 EN7 Purchase decisions Always Fig. 3. Cluster 3 “Occasional environmentalists” (see Table 1 for labels). Frequency bars refer to the percentage of respondents involved in the speciWc behaviour. Rarely Sometimes Usually W8 W6 WE7 W5 W4 W3 WA7 EN5 EN4 EN3 EN6 EN2 Recycling EN1 WA1 WA4 WA2 WA3 WA6 WA5 WA8 GC4 GC5 GC7 GC6 GC8 W2 Never Habits GC9 W1 GC1 W9 W10 GC2 GC10 GC3 EN8 EN7 Purchase decisions Always Fig. 4. Cluster 4 “Non-environmentalists” (see Table 1 for labels). Frequency bars refer to the percentage of respondents involved in the speciWc behaviour. undertook a range of these activities. Water conservation behaviours were popular, but this was not the case across the board, with some being less popular than certain green consumer behaviours. Of particular signiWcance in this cluster was the vast majority who stated that they always or usually composted waste. Mainstream environmentalists (Fig. 2) were clearly distinguished from the committed group by their lack of composting activity; with over 60% stating that they ‘never’ composted kitchen or garden waste. In general, there were more individuals who ‘never’ or ‘rarely’ undertook a range of activities. Accordingly, there was a shift from signiWcant commitment to a more equivocal position. Occasional environmentalists (Fig. 3) were signiWcantly less likely to engage in all of the activities, but especially green consumer and certain water conservation practices. There were also larger proportions who either never or rarely recycled waste. The term ‘occasional’ seems appropriate for this cluster given the greater number of individuals who stated that they sometimes took part. Non-environmentalists (Fig. 4) stand apart in a number of ways from the previous three clusters. In all but two cases, far fewer than 50% of individuals usually or always undertook each behaviour, with most being under 20%. Those activities where there was more behavioural commitment were related mostly to ‘common sense’ actions, with 914 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 overtly pro-environmental behaviours being undertaken with little or no frequency. This group is signiWcant as it demonstrates a distinctly non-committed group of individuals who rarely engage with environmentalism. 5.3. Who are the ‘environmentalists’ ƒ and the ‘non-environmentalists’? These four distinctive clusters are signiWcant and they provide a behavioural snapshot of environmental lifestyle groups in Devon in the early 21st century. The question that arises from these data is what values and attitudes deWne these lifestyle groups? The following sections describe these attributes in detail. The reader should note that demographic characteristics are not dealt with here, but will be the subject of forthcoming work as the complexity of the social composition of the groupings cannot be dealt adequately in the space provided here. 5.3.1. Social and environmental values This research measured a series of social value constructs (Schwartz, 1992) relating to two dimensions, notably ‘altruism–egoism’ and ‘openness to change–conservative’. In addition, values relating to the new environmental paradigm (Dunlap and Van Liere, 1978) and operational values (O’Riordan, 1985) were also measured. Stern et al. (1995) have demonstrated that there may be conceptual links between these dimensions of social and environmental values. To examine the value structure, social and environmental values were placed into a factor analysis to examine the empirical links between the diVerent constructs. Tables 2 and 3 provide the relevant results. Table 2 demonstrates that the four factors which emerged related closely to Schwartz’s original conception of social values, whilst three environmental value factors were evident (Table 3), pertaining roughly to biospherism, anthropocentrism and an ecocentric–technocentric dimension. The biocentric and anthropocentric factors partly comprised a range of items that comprised Dunlap et al.’s (2000) New Ecological Paradigm scale, whilst the ecocentric–technocentric factor was smaller and more distinct, relating to O’Riordan’s (1985) operational values. In all three cases, higher scores on a factor indicate a pro-environmental position. To investigate whether these related to environmental action, the diVerences between each behavioural group and each factor were investigated. The means for each behavioural cluster and the Mood median statistic used to test for diVerences are presented in Tables 2 and 3. Table 2 clearly shows that for both the altruism and openness to change factors, no signiWcant diVerences existed according to behavioural group. Accordingly, the evidence provided by previous research (such as Corraliza and Berenguer, 2000) does not Wnd support in these data. In large part this may be because the data were so heavily skewed, with large majorities in both cases stating that they found all of the constructs ‘important’. This is not the case with regard to the conservative and egoism factors. Table 3 clearly demonstrates signiWcant diVerences according to behaviour in these instances. Interestingly, in contrast to Stern et al.’s (1995) assertions, those least committed to environmental action (cluster 4) were those most likely to disagree with this factor. However, they were also those who were most likely to place greater emphasis on personal wealth. These data indicate some key trends with regard to social values. In particular, there would appear to be a clear divide between non-environmentalists and the remaining sample. Those least committed to environmental action were more likely to value wealth, social power and be less Table 2 Social value factors and Mood median statistics showing the diVerences between the four clusters Factor Variables included Means Cluster 1 Cluster 2 Cluster 3 Cluster 4 Altruistic Loyalty Honouring parents Equality Social justice Enjoying life Helpful 26.3 26.5 26.4 Openness to change Varied life Exciting life Curious 11.5 11.3 Conservative Social order Obedience Authority Unity 15.5 Egoism Wealth Social power InXuential 8.7 Total variance Variance (eigenvalue) Per cent variance Mood’s median 26.6 3.1 20 5.8 (p > 0.05) 11.6 11.6 2.5 16 2.6 (p > 0.05) 15.5 15.3 14.9 2.1 13 7.8 (p < 0.05) 9.0 9.1 9.7 1.8 11 9.8 (p < 0.05) 59 Individuals were asked how important each value was to their own life, rating each from 1 (very unimportant) to 5 (very important). Each factor mean is a composite score of the items in the factor and divided by the number in the sample. 915 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 Table 3 Environmental value factors Factor Variables included Means Cluster 1 Cluster 2 Cluster 3 Cluster 4 Variance Per cent Mood’s (eigenvalue) variance median Faith in growth: There are no limits to growth anthropocentrism for nations like the UK Modifying the environment seldom causes serious problems Science will help us to live without conservation Humans were created to rule over nature 15.4 14.7 14.8 13.3 2.3 24 12.6 (p < 0.05) Spaceship Earth: biospherism The balance of nature is delicate and easily upset The Earth is like a space ship, with limited room and resources Plants and animals do not exist primarily for human use One of the most important reasons for conservation is to preserve wild areas 16.6 16.3 16.1 15.1 2.2 22 10.2 (p < 0.05) Ecocentismtechnocentrism Technology will solve many environmental problems Exploitation of resources should be stopped 6.8 6.9 6.8 6.2 1.2 12 3.9 (p > 0.05) Total variance 58 Individuals were asked to rate their agreement with each statement from 1 (strongly disagree) to 5 (strongly agree). Each factor mean is a composite score of the items in the factor, all coded in a pro-environmental direction and divided by the number in the sample. committed to unity (a belief in the importance of consensus and social cohesion). These data do not provide evidence that there is a clear altruistic–egoistic continuum of values on which the environmentalist can neatly be placed. Rather, given the nature of the questions posed and the social desirability of answering questions in certain ways, a skewed pattern has emerged that only reXects the (signiWcant) divergence between non-environmentalists and those who are relatively committed. In summary, social values do not generally appear to be related to environmentalism, but this relationship is skewed and may need further investigation, as highlighted by Burton and Wilson’s (2006) recent research which suggests that behavioural context is signiWcant in determining which value dimensions are expressed and the strength to which these are advocated. In regard to environmental values, a far clearer and uniform pattern emerges. The Faith in Growth factor (Table 3) represents both relational and operational values and provides a consistent pattern, whereby committed environmentalists disagree most strongly with such an anthropocentric/ technocentric position, whilst non-environmentalists are most likely to agree with such a position. Although the data are skewed in a broadly pro-environmental direction, it is evident that there is a trend from committed to non-environmentalists for stronger disagreement. This is also the case with the Spaceship Earth factor (biospheric/ecocentric) shown in Table 3. In all cases, committed environmentalists take the strongest biospheric/ecocentric position, with nonenvironmentalists taking a weaker stance. This pattern is not present in the Wnal factor (Table 3), which may be accounted for by the more deWnitive statement wording used. Overall, committed environmentalists are more likely to hold biospheric and ecoentric values and despite the skewing of the data, there are signiWcant diVerences between this group and all other categories. Social values, and particularly environmental values, are clearly diVerent according to behavioural category, with committed environmentalists more likely to hold values that encapsulate an emphasis on unity (social cohesion and order) with less emphasis on personal wealth. They believe in the intrinsic value of nature and in limits to growth. Accordingly, previous research linking environmental values to behaviour (e.g. Stern et al., 1995) is supported in part, giving credence to the argument that those not involved in environmental action tend to believe that humans have greater value than nature and that technology provides a means by which to resolve environmental dilemmas. Such a Wnding enables social researchers to appreciate the eVect of embedded values in everyday lifestyle choices and may assist in the development of further lifestyle adjustment programmes. 5.4. Environmental attitudes The relationship between environmental attitudes and behaviour encapsulates a wide range of research and we hypothesised that these would show clear diVerences according to behavioural group. The study examined a range of environmental attitudes which measured all of the various constructs provided in the literature review above. 916 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 These were measured on Likert agreement scales and were also factor analysed in order to derive empirical dimensions. There is not space in this paper to examine these in full detail, but a series of 15 factors emerged, measuring logical concepts that related to those identiWed in previous research. A summary of the content of each factor is provided in Table 4. As with environmental values, the key question focused on the extent to which those who were more committed to environmental action held diVerent attitudes to those who were less committed. A series of Mood median tests were computed to examine whether there were signiWcant diVerences between the scores on each factor and the behavioural groupings. These are presented in Table 4, along with the mean scores for each cluster. With the exception of one factor (Brand loyalty), there was a statistically signiWcant diVerence between the behavioural groupings for each factor. Given that the scales were coded in order to show a pro-environmental attitude in all cases, higher scores (and therefore means) represent a more pro-environmental position. Although means are not the most eVective way of showing such information, they do provide an eVective technique for comparison across a large number of clusters and factors. Although variation is evident, a general trend emerges which demonstrates that those with more pro-environmental attitudes were in clusters 1 and 2, with those holding less pro-environmental positions being in clusters 3 and 4. These results provide some important Wndings. First, the greatest diVerences, indicated by higher Mood median statistics (all p < 0.05), indicate that concern and commitment, moral motivations, logistical factors and green consumer beliefs provide scope for diVerentiating between diVerent types of lifestyle groups. Four key themes emerge from these diVerences. In the Wrst instance, there are clearly high levels of concern, which are illustrated by the personal responsibility and moral obligation felt by respondents to help the environment. Clearly, personalisation of the environmental problem and a genuine moral obligation to act personally is a motivator for environmental action. A further theme to emerge relates to logistical factors and the ease with which individuals felt they could be pro-environmental. A belief that helping the environment takes less eVort and is worthy of a time commitment is evidently signiWcant. In addition, there is the perceived social acceptance of the behaviour, enabling individuals to take part with minimal personal sacriWce to their self-presentation. This supports the notion that those less pro-environmental individuals perceive a certain social stigma about participating, reXecting an underlying social attitude which places environmentalism outside the mainstream of socially desirable activities. Finally, the fourth theme to emerge relates to the speciWc importance of green consumer beliefs, with those more concerned with health and safety issues and the importance of local produce being more pro-environmental. A second trend in the data relates to the signiWcant, if less sharp, diVerence between factors pertaining to outcome beliefs, price, comfort and environmental rights. Two core themes emerge from these data. In the Wrst instance, the eYcacy of a given activity is evidently highest amongst those who are committed environmentalists, indicating that being conWdent in the ability of one’s personal actions to eVect a tangible outcome environmentally is signiWcant in motivating behaviour. A second theme reXects the wider Table 4 Environmental attitudes Factor Salient composite variables (expressed in pro-environmental direction) Concern and commitment Environmental concern Time to help the environment Self-eYcacy in helping the environment Personal responsibility for environmental problems Moral motives Moral obligation to help the environment Social acceptance of environmental behaviour Self-presentation of environmental action a positive quality Outcome beliefs Beliefs in eYcacy of actions Price Willingness to pay more for environmental products Satisfactions Various satisfactions from environmental action Logistics Convenience of helping the environment Green consumer Importance of health, safety and local issues attitudes to helping the environment Comfort Willingness to sacriWce home comforts to help the environment Environmental rights Belief in strong environmental rights Awareness of norm Awareness of the normality of environmental actions Trust and responsibility Trust in environmental information Importance of personal responsibility Extrinsic motivation Rejection of extrinsic motives to help the environment Personal instinct Trust in one’s own instinct regarding the environment Brand loyalty Willingness to change brand labels to help the environment Personal threat Threat of environmental problems to the self Means Mood median test Cluster 1 Cluster 2 Cluster 3 Cluster 4 27.7 26.5 25.1 22.0 69.6 (p < 0.05) 18.7 18.6 17.2 15.2 62.3 (p < 0.05) 17.0 11.3 18.9 10.6 15.8 16.7 10.8 19.1 10.0 15.3 16.0 10.0 18.6 8.9 14.4 14.9 9.5 18.7 7.8 13.4 45.1 (p < 0.05) 32.5 (p < 0.05) 9.4 (p < 0.05) 78.1 (p < 0.05) 84.9 (p < 0.05) 7.0 10.2 6.2 9.0 6.9 10.0 6.3 9.5 6.4 9.3 5.8 9.0 5.4 8.1 5.1 9.1 42.9 (p < 0.05) 37.0 (p < 0.05) 23.5 (p < 0.05) 12.3 (p < 0.05) 6.8 3.8 2.6 3.9 6.6 3.7 2.7 3.8 6.4 3.6 2.7 3.7 6.4 3.9 3.0 3.5 10.7 (p < 0.05) 12.4 (p < 0.05) 3.5 (p > 0.05) 17.8 (p < 0.05) Each factor mean is a composite score of the items in the factor, all coded in a pro-environmental direction and divided by the number in the sample. S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 debate surrounding personal sacriWces on the one hand and willingness to pay for environmentally friendly products on the other. Those who tended to be less willing to pay for environmentally friendly products were less inclined to help the environment, a theme that is demonstrated with regard to the willingness of individuals to sacriWce home comforts. A third trend in the data demonstrates some weak diVerences, pertaining to trustworthiness of environmental information, satisfactions and brand loyalties. The data suggest that there are diVerences between the behavioural groups according to the trustworthiness of information, but that these do not reXect behavioural commitment in the more linear fashion shown above. Accordingly, committed environmentalists may be less inclined to trust information provided than non-environmentalists. Indeed, signiWcant diVerences were reXected between the groups and scores on the satisfactions factor, even though there is little discernible diVerence between committed and non-environmentalists. Finally, there was no signiWcant diVerence between the willingness to change brand loyalty and behavioural commitment. These data suggest that the dominant themes in the literature relating to attitudes towards environmental action are generally supported. There are strong diVerences in behavioural commitment according to environmental concern, social acceptance, convenience and green consumer motivations. However, the inXuence of satisfactions and trust in environmental information is questionable. Overall, the environmentalist is a highly concerned individual, motivated by a range of issues, who is conWdent in the outcome of their actions and Wnds helping the environment relatively simple and socially desirable. 6. Discussion At the outset of this paper, two key issues were outlined, relating to new forms of environmental practice and the identiWcation of lifestyle groups for environmental action. Through a series of analytical procedures, the paper has provided an examination of the diVerent behavioural responses of individuals to the new environmental agenda focused around sustainability and has given an insight into the proWle of the modern environmental ‘activist’. The data have provided some corroborative evidence for existing research, whilst providing some new avenues for future investigation. The Wrst point to be drawn from the data relates to the construction of environmental action. It is indicative of the current research climate, alongside a similar trend in policy, that research on environmental action has become sectored into dealing with speciWc issues such as recycling or energy saving. This research provides evidence that environmental behaviour transcends these somewhat compartmentalised boundaries and should be placed in an holistic context which recognises links between speciWc modes of behaviour. The Devon study has demonstrated that environmental action may be deWned by three behavioural types that 917 conform to speciWc activity domains. These can be characterised as purchase related activities, such as various forms of green consumption, which cover a wide variety of actions, including the purchase of fairly traded produce, organic foods, environmentally benign products and energy saving appliances. The next behavioural type reXects habitual activities, undertaken within the home as part of a daily routine. Such behaviours do not generally reXect major shifts in behaviour, but rather adjustments to lifestyles. Finally, recycling behaviour constitutes a highly structured and mechanised behaviour, with individuals sorting and cleaning materials for recycling collection. These fundamental diVerences in daily environmental practices emphasise the ways in which sectored approaches do not transcend the lived experiences of individuals and thus may not engage fully with the means by which individuals can change behaviour. A second point to emerge from the data relates to the level of behavioural commitment, both between types of behaviour and individuals. In general, sustainable purchasing behaviour was least popular, with few individuals engaging in organic or FairTrade purchases. Recycling, as a highly structured and long-standing environmental action, was the most popular activity. However, the cluster analysis provided a clear interpretation on the diVerent forms of lifestyles that were apparent, reXecting committed environmentalists through to non-environmentalists. These groups provide a valuable insight into the everyday lives of individuals and how environmental action varies between such groupings. Both committed and mainstream environmentalists were highly committed to a range of environmental activities, although the former were keen composters, with the latter rarely doing so. In general, these two groups represent a substantial segment of the population who are willing to engage in environmental activities and thus they have adjusted their lifestyles accordingly. They are willing to engage in most forms of sustainable consumption, although not with the regularity of recycling or habitual behaviours. Conversely, the lifestyles of occasional and especially nonenvironmentalists are reXected in a rejection of most environmental practices, including recycling. Only those habitual actions which do not necessarily have an environmental motivation are undertaken with some regularity. Clearly then, these data provide a window on contemporary lifestyles that enable us to plot a range of environmental actions across behavioural groups. The third and most useful element of the data is how researchers can use these behavioural clusters to construct a picture of the types of individual engaged in a range of environmentally-related lifestyles. The research suggests that there is evidence to support Stern et al.’s (1995) assertion that Schwartz’s (1992) value continua are of utility when examining environmental behaviour. The data in this paper suggest that there are diVerences in behavioural groupings according to a number of value constructs, relating speciWcally to the importance of unity, wealth and social power. These data may imply that there are some divisions between 918 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 how values are constructed and the ways individuals behave towards the environment. However, they do not provide the sort of conclusive evidence that would be needed to assert a clear link between social values and environmental action. Accordingly, more research is needed to provide data on this issue. One of the problems in the current work may have been the measurement scales used and the heavy skew in the data that was apparent, where overwhelmingly positive social values were expressed. Where clear conclusions can be drawn is in respect of environmental values. The data clearly showed that biospheric and ecocentric values were held by those who were committed environmentalists, with anthropocentric and technocentric values held by non-environmentalists. This is a signiWcant Wnding, as it corroborates much of the evidence from work on both relational and operational values (Dunlap and Van Liere, 1978; Dunlap et al., 2000; Thompson and Barton, 1994). The two largest factors included both anthropocentric/techocentric as well as ecocentric/biospheric items, indicating that relational and operational values are intrinsically linked, as would be anticipated. This Wnding provides a useful means by which to summarise the impact of environmental values, which for those least interested in environmental behaviour, are dominated by a belief that humans are dominant over nature and that technological solutions are the key to resolving environmental dilemmas. A fourth point that emerges from the data relates to attitudes. The environmentalist clearly has a positive, conWdent and responsible attitude towards environmental protection. Environmentalists are conscious of the eVects of environmental problems and thus feel morally obliged to participate in behaviours that they believe will have a tangible eVect on the global and local environment. This supports considerable research suggesting that social–psychological constructs such as moral obligation, responsibility, social desirability and response eYcacy are signiWcant motivators for environmental action. However, the data do not support the hypothesis that satisfactions are relevant to environmental action (De Young, 1986, 1996) or that trusting given sources of information is eYcacious (MacNaghten and Urry, 1998). A Wnal point brings us back to the debate raised in the second section of the paper, which highlighted the diVerence between the dominant geographical discourse (emphasising a ‘civic’ approach to understanding and changing behaviour) and the simple, rationalistic ‘information deWcit’ models of behaviour. We proposed that an alternative to both of these perspectives would be to dig deeper into the methodological advances that have been made outside the discipline of geography, towards social psychology, quantitative sociology and business research. We have presented an analysis that genuinely attempts to understand and Wnd ways to understand behaviour, but within what geographers have on the surface, rejected as a ‘rationalistic’ approach. We do not accept this overall conclusion, given the valuable insights that have been provided by a wider range of disciplines. This paper has clearly demonstrated that Owens’ (2000, p. 1143) comment that critics of the approach argue that “ƒ the ‘mental models’ approach still suVers from too rationalistic a conception of agency and from a methodological individualism which abstracts human subjects from their social context” is misplaced. Indeed, our approach here implicitly involves social context and agency in a non-judgemental framework. In addition to this signiWcant point, the research here demonstrates that there is clearly room for developing understandings of behaviour and behavioural change, which can make a contribution to policy. Accordingly, we would argue that geographers can learn a great deal from the work of other disciplines in the engagement of citizens with sustainability. 7. Conclusion With the onset of sustainable development, individuals have come to symbolise the move from global collective action, to local and personalised responses to environmental issues. This shift, which can be seen in tandem with the move to individualism indicative from the late 1980s onwards, has individualised and privatised environmental action, such that environmental action has been integrated into everyday life, in some cases with relatively minor adjustments to lifestyles. Such a move to a new environmental activism has resulted in a diverse set of behavioural responses, which range from a strong commitment to environmental actions, from composting to buying energy saving light bulbs. In other cases, there have been few adjustments, focused mostly on habitual activities. Accordingly, there is scope for research that examines these diVerent types of lifestyle and which explores in more depth the motivations, attitudes and values of these diVerent groups. This study has shown how environmental action has become embedded within everyday experiences and lifestyles in and around the home. However, geographers need to examine how such lifestyles relate to other areas of interest, in particular at work and during leisure activities. There is scope to examine the diVerent behavioural settings that mould alternative responses to diVerent environmental situations. For example, does environmental behaviour change, or even become inverted, on holiday or at work? How do value systems relate to these diVerent circumstances and can geographers make a contribution to the widening debate regarding lifestyle changes? One thing is certain, that the governmental pressure on individuals to engage with environmental practices will continue to grow and it is therefore crucial for geographers to map and synthesise these changes in everyday lives. Acknowledgement The authors would like to thank the Economic and Social Research Council for Wnancial assistance in undertaking this research (Grant No. R000239417). S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 References Ajzen, I., 1991. The theory of planned behavior. Organizational Behavior and Human Decision Processes 50, 179–211. Baalderjahn, U., 1988. Personality variables and attitudes as predictors of ecologically responsible consumption patterns. Journal of Business Research 17, 51–56. Baldassare, M., Katz, C., 1992. The personal threat of environmental problems as predictor of environmental practices. Environment and Behavior 24, 602–616. Barr, S., Gilg, A.W., Ford, N.J., 2001. A conceptual framework for understanding and analysing attitudes towards household waste management. Environment and Planning A 33, 2025–2048. Berk, R.A., Cooley, T.F., La Civita, C.J., Parker, S., Sredi, K., Brewer, M., 1980. Reducing consumption in periods of acute scarcity: the case of water. Social Science Research 9, 99–120. Blake, J., 1999. Overcoming the ‘value–action’ gap in environmental policy. Local Environment 4, 257–278. Boldero, J., 1995. The prediction of household recycling of newspapers: the role of attitudes, intentions, and situational factors. Journal of Applied Social Psychology 25, 440–462. Bulkeley, H., 1999. Valuing the Global Environment: Publics, Politics and Participation. Unpublished PhD Thesis, University of Cambridge, Cambridge. Burgess, J., Harrison, C.M., Filius, P., 1998. Environmental communication and the cultural politics of environmental citizenship. Environment and Planning A 30, 1445–1460. Burton, R.J.F., 2004. Reconceptualising the ‘behavioural approach’ in agricultural studies: a socio-psychological perspective. Journal of Rural Studies 20, 359–371. Burton, R.J.F., Wilson, G., 2006. Injecting social psychology theory into conceptualisation of agricultural agency: towards a post-productivist self-identity. Journal of Rural Studies 22, 95–115. Cameron, L.D., Brown, P.M., Chapman, J.G., 1998. Social values and decisions to take proenvironmental action. Journal of Applied Social Psychology 28, 675–697. Chan, R.Y.K., 1998. Mass communication and proenvironmental behaviour: waste recycling in Hong Kong. Journal of Environmental Management 52, 317–325. Corraliza, J.A., Berenguer, J., 2000. Environmental values, beliefs and actions: a situational approach. Environment and Behavior 32, 832– 848. Costanzo, M., Archer, D., Aronson, E., Pettigrew, T., 1986. Energy conservation behaviour: the diYcult path from information to action. American Psychologist 41, 521–528. Department of the Environment, Food and Rural AVairs, 2005. Securing Our Future: delivering UK sustainable development strategy. The Stationary OYce, London. Derksen, L., Gartell, J., 1993. The social context of recycling. American Sociological Review 58, 434–442. De Young, R., 1986. Some psychological aspects of recycling. Environment and Behavior 18, 435–449. De Young, R., 1996. Some psychological aspects of reduced consumption behaviour: the role of intrinsic motivation and competence motivation. Environment and Behavior 28, 358–409. Dunlap, R.E., Van Liere, K.D., 1978. The ‘New Environmental Paradigm’. Journal of Environmental Education 9, 10–19. Dunlap, R.E., Van Liere, K.D., Mertig, A.G., Jones, R.E., 2000. Measuring endorsement of the new ecological paradigm: a revised NEP scale. Journal of Social Issues 56, 425–442. Eden, S., 1993. Individual environmental responsibility and its role in public environmentalism. Environment and Planning A 25, 1743–1758. Eden, S., 1996. Public participation in environmental policy: considering scientiWc, counter-scientiWc and non-scientiWc contributions. Public Understanding of Science 5, 183–204. Fishbein, M., Ajzen, I., 1975. Belief, Attitude, Intention and Behaviour: An Introduction to Theory and Research. Addison-Wesley, Reading, MA. 919 Gibbs, D., Longhurst, J., Braithwaite, C., 1998. ‘Struggling with sustainability’: weak and strong interpretations of sustainable development within local authority policy. Environment and Planning A 30, 1351– 1365. Gilg, A.W., 2005. Planning in Britain: Understanding and Evaluating the Post-war System. Sage, London. Gilg, A.W., Ford, N.J., Barr, S., 2003. Environmental action in and around the home. Final report on ESRC Project R000239417, ESRC, Swindon. Available from: <www.esrcsocietytoday.ac.uk>. Gray, S., 1971. The Electoral Register: Practical Information for Use When Drawing Samples, Both for Interview and Postal Survey. OPCS, London. Guagnano, G.A., Stern, P.C., Dietz, T., 1995. InXuences on attitude–behavior relationships: a natural experiment with curbside recycling. Environment and Behavior 27, 699–718. Hines, J.M., Hungerford, H.R., Tomera, A.N., 1987. Analysis and synthesis of research on responsible environmental behaviour: a meta analysis. Journal of Environmental Education 18, 1–8. Hobson, K., 2001. Sustainable lifestyles: rethinking barriers and behaviour change. In: Cohen, M.J., Murphy, J. (Eds.), Exploring Sustainable Consumption: Environmental Policy and the Social Sciences. Elsevier, Amsterdam. Hobson, K., 2002. Competing discourses of sustainable consumption: does the ‘rationalisation of lifestyles’ make sense? Environmental Politics 11, 95–120. Kantola, S.J., Syme, G.J., Nesdale, A.R., 1983. The eVects of appraised severity and eYcacy in promoting water conservation: an informational analysis. Journal of Applied Social Psychology 13, 164–182. Lam, S., 1999. Predicting intentions to conserve water from the theory of planned behaviour, perceived moral obligation and perceived water right. Journal of Applied Social Psychology 29, 1058–1071. Mackenzie, D., 1990. The green consumer. Food Policy (December), 461– 466. MacNaghten, P., Jacobs, M., 1997. Public identiWcation with sustainable development: investigating cultural barriers to participation. Global Environmental Change 7, 5–24. MacNaghten, P., Urry, J., 1998. Contested Natures. Sage, London. Mainieri, T., Barnett, E.G., Valdero, T.R., Unipan, J.B., Oskamp, S., 1997. Green buying: the inXuence of environmental concern on consumer behaviour. The Journal of Social Psychology 137, 189–204. Munton, R., 1997. Engaging sustainable development: some observations on progress in the UK. Progress in Human Geography 21, 151–159. Myers, G., MacNaghton, P., 1998. Rhetorics of environmental sustainability: commonplaces and places. Environment and Panning A 30, 333– 353. Nancarrow, B.E., Smith, L.M., Syme, G.J., 1996–1997. The ways people think about water. Journal of Environmental Systems 25, 15–27. O’Riordan, T., 1976. Environmentalism. Pion, London. O’Riordan, T., 1985. Future directions in environmental policy. Environment and Planning A 17, 1431–1446. Owens, S., 2000. Engaging the public: information and deliberation in environmental policy. Environment and Planning A 32, 1141–1148. Roberts, J.A., 1996. Green consumers in the 1990’s: proWle and implications for advertising. Journal of Business Research 36, 217–231. Sadalla, E.K., Krull, J.L., 1995. Self-presentational barriers to resource conservation. Environment and Behavior 27, 328–353. Samuelson, C.D., Biek, M., 1991. Attitudes toward energy conservation: a conWrmatory factor analysis. Journal of Applied Social Psychology 21, 549–568. Schultz, P.W., Oskamp, S., Mainieri, T., 1995. Who recycles and when? A review of personal and situational factors. Journal of Environmental Psychology 15, 105–121. Schwartz, S.H., 1992. Universals in the content and structure of values: theoretical advances and empirical test in 20 countries. Advances in Experimental Social Psychology 10, 221–279. Schwartz, S.H., Blisky, W., 1987. Toward a psychological structure of human values. Journal of Personality and Social Psychology 53, 550– 562. 920 S. Barr, A. Gilg / Geoforum 37 (2006) 906–920 Schwepker, C.H., Cornwell, T.B., 1991. An examination of ecologically concerned consumers and their intention to purchase ecologically packaged products. Journal of Public Policy and Marketing 10, 77– 101. Segun, C., Pelletier, L.G., Hunsley, J., 1998. Towards a model of environmental activism. Environment and Behavior 30, 628–652. Seligman, C., Kriss, M., Darley, J.M., Fazio, R.H., Becker, L.J., Payor, J.B., 1979. Predicting summer energy conservation from homeowners’ attitudes. Journal of Applied Social Psychology 9, 70–90. Shrum, L.J., McCarty, J.A., Lowrey, T.M., 1995. Buyer characteristics of the green consumer and their implications for advertising strategy. Journal of Advertising 24, 71–82. Sparks, P., Shepherd, R., 1992. Self-identity and the theory of planed behavior: the role of identiWcation with ’green consumerism’. Social Psychology Quarterly 55, 388–399. Stern, P., 1992. What psychology knows about energy conservation. American Psychologist 47, 1224–1232. Stern, P.C., Dietz, T., Guagnano, G.A., 1995. The new ecological paradigm in social–psychological context. Environment and Behavior 27, 723–743. Taylor, S., Todd, P., 1997. Understanding the determinants of consumer composting behaviour. Journal of Applied Social Psychology 27, 602–628. Thompson, S.C.G., Barton, M.A., 1994. Ecocentric and anthropocentric attitudes toward the environment. Journal of Environmental Psychology 14, 149–157. Verhallen, T.M.M., Van Raaij, W.F., 1981. Household behaviour and the use of natural gas for home heating. Journal of Consumer Research 8, 253–257. Vining, J., Ebreo, A., 1990. What makes a recycler? A comparison of recyclers and nonrecyclers. Environment and Behavior 22, 55–73. Vining, J., Ebreo, A., 1992. Predicting recycling behaviour from global and speciWc environmental attitudes and changes in recycling opportunities. Journal of Applied Social Psychology 22, 1580–1607. Wilson, G., 1996. Farmer environmental attitudes and ESA participation. Geoforum 27, 115–131.