Practice-ing behaviour change: Applying social practice theory to pro-environmental behaviour change

Tom Hargreaves, School of Environmental Sciences, University of East Anglia, Norwich, NR4 7TJ, UK

Email: tom.hargreaves@uea.ac.uk

Behavioural change is fast becoming the 'holy grail' for sustainable development policy. (Jackson, 2005: x)

It should be borne in mind that there is nothing more difficult to handle, more doubtful of success, and more dangerous to carry through than initiating changes. (Machiavelli, 1961: 21)

Over the last 40 years, and particularly in the sub-discipline of environmental psychology, numerous attempts have been made to construct models of human behaviour that identify its determinants in order that it may be directed into more pro-environmental channels (see either Darnton, 2008 or Jackson, 2005 for extensive reviews of such models). Specifically, such models have looked to individuals' beliefs, attitudes and values as predictors of behaviour. If only pro- or anti-environmental beliefs could be identified and modified, the models suggest, behaviour might be changed. Central to this approach has been the assumption that behaviour is 'the outcome of a linear and ultimately rational process' (Harrison and Davies, 1998: 2) of decision making undertaken by more or less rational individuals.

Perhaps the most widely used of these models is Ajzen's (1991) 'theory of planned behaviour', developed as an extension of the earlier 'theory of reasoned action' (Fishbein and Ajzen, 1975). This theory posits that behavioural intention, which necessarily precedes actual behaviour, results from interactions between an individual's attitude towards the behaviour in question, their beliefs about what others think about the behaviour – the subjective norm – and their perceived level of control over the behaviour, or perceived behavioural control. One attractive feature of this model is its openness to the inclusion of additional variables (cf. Ajzen, 1991) and, as such, through multiple applications of the model many more factors have been added, including: belief salience, past behaviour/habit, moral norms, affective beliefs, self-identity, and perceived behavioural control versus self-efficacy (e.g. Conner and Armitage, 1998; Mannetti et al., 2004). As more and more variables are added-in to such models, however, there are diminishing returns to the enhancement in their predictive capacity, and their increasing complexity renders them less amenable to practical application (cf. Jackson, 2005).

Bamberg (2003) argues that a major reason why such models have received so much attention is because they treat attitudes (and the same essentially applies for values, beliefs, needs and motivations) as 'situation invariant orientation patterns' (Bamberg, 2003: 22). In short, if the necessary cognitive components can be identified and modified, behavioural changes will cascade across contexts throughout all areas of an individual's lifestyle. More recently, however, it has been recognized that individuals do not exist in a social vacuum and, indeed, that in some cases the surrounding context overrides all of the cognitive factors included in the models (Stern, 2000). Rather than challenge the basic assumption of the models that individual decision making is 'maladaptive' (Maloney and Ward, 1973) and in need of correction, however, such insights have instead led to the gradual incorporation of various proxies for context (e.g. as social norms, social networks or surrounding infrastructures) as yet more variables in individual decision-making processes (e.g. Barr, 2003; Martin et al., 2006; Olli et al., 2001).

The persistence of such linear models of behavioural correction is perhaps partly explained because they render policy responses relatively straightforward.

Early approaches, for example, sought to provide information to fill a presumed information deficit (cf. Burgess et al., 1998; Owens, 2000) among the population and educate them to hold more (eco)rational attitudes, beliefs and values as seen, for example, in campaigns such as 'Helping the Earth Begins at Home' (Blake, 1999; Hinchliffe, 1996) and 'Are You Doing Your Bit?' (DEMOS, 2003). More recently, the contextual sensitivity of such approaches has been enhanced in social marketing interventions that seek to identify and remove real and perceived contextual 'barriers', and then to tailor messages to carefully selected audience segments in order to create new social norms that, it is hoped, will motivate individuals to adopt the desired behaviours (e.g. Barr, 2008; DEFRA, 2008).

Whilst Barr (2008) applauds the incrementalism of social marketing, that it 'works with' rather than against existing lifestyles and patterns of consumption, others, and most prominently Shove (2003, 2004, 2010), argue that such an approach is fundamentally flawed and in fact serves to legitimize, rather than challenge, a whole range of unsustainable social conventions. In particular, critics have argued that such approaches are excessively individualistic and fail to appreciate the ways in which, variously, social relations, material infrastructures and context are intrinsic to the performance of social practices (e.g. Bedford, 1999; Hobson, 2003; Nye and Hargreaves, 2010; Shove, 2003; Southerton et al., 2004; Spaargaren and Van Vliet, 2000), and not merely variables among many others within individuals' decision-making processes. More recently, an emerging approach based around ideas and theories of social practice has sought to address many of these shortcomings. This will form the subject of the next section.

In contrast to the undersocialized methodological individualism of the behavioural models outlined above, social practice theorists, from Giddens (1984) and Bourdieu (1977, 1990), to more recent work by Reckwitz (2002), Schatzki (1996, 2002), Shove (2010, and see Shove and Pantzar, 2005) and Warde (2005), have all sought a middle level between agency and structure. This has been found in the everyday and routine performance of social practices such as cooking, driving, washing, shopping or playing football. The practice itself, rather than the individuals who perform them or the social structures that surround them, thus becomes the core unit of analysis. As Giddens observed:

the basic domain of study of the social sciences . . . is neither the experience of the individual actor, nor the existence of any form of societal totality, but social practices ordered across space and time. (Giddens, 1984: 2)

In this view, anti- or pro-environmental actions, and more or less sustainable patterns of consumption, are not seen as the result of individuals' attitudes, values and beliefs constrained by various contextual 'barriers', but as embedded within and occurring as part of social practices (Warde, 2005). In turn, the performance of

various social practices is seen as part of 'the routine accomplishment of what people take to be "normal" ways of life' (Shove, 2004: 117).

Social practice theory thus diverts attention away from moments of individual decision making, and towards the 'doing' of various social practices and the inconspicuous consumption (Shove and Warde, 2002) they entail. Further, individuals themselves are removed from centre-stage and instead become the 'carriers' (Reckwitz, 2002) of social practices, carrying out the various activities and tasks that the practice requires. As individuals pass through life, they come into contact with, get recruited to, have 'careers' within, and occasionally defect from a wide variety of different practices. Importantly, practice theory emphasizes that it is through these engagements with practices that individuals come to understand the world around them and to develop a more or less coherent sense of self (cf. Warde, 2005). This does not, however, render individuals as passive dupes beholden to the dictates of practice, but instead conceives of them as skilled agents who actively negotiate and perform a wide range of practices in the normal course of everyday life. Bringing about pro-environmental patterns of consumption, therefore, does not depend upon educating or persuading individuals to make different decisions, but instead on transforming practices to make them more sustainable (cf. Southerton et al., 2004). As Warde (2005: 140) notes, 'the principal implication of a theory of practice is that the sources of change behaviour lie in the development of practices themselves'.

Beyond these points of relative agreement, however, 'there is no unified practice approach' (Schatzki, 2001: 2). One area of disagreement, for example, centres on defining exactly what a practice is. Here, some theorists focus on the various components or elements that make up a practice (e.g. Reckwitz, 2002; Shove and Pantzar, 2005), others on the connections between these elements (e.g. Schatzki, 2002; Warde, 2005), and still others on the position of practices as a bridge between individuals' lifestyles and broader socio-technical systems of provision (e.g. Spaargaren and Van Vliet, 2000). In this article, however, I will adopt Shove and Pantzar's (2005) empirically helpful understanding of practices as assemblages of images (meanings, symbols), skills (forms of competence, procedures) and stuff (materials, technology) that are dynamically integrated by skilled practitioners through regular and repeated performance. To provide a simple example, football involves a specific set of images (e.g. about the rules and aim of the game and the appropriate level of emotional engagement), skills (e.g. of dribbling and kicking a ball), and stuff (e.g. a ball and a goal). The links between these elements are then (re)produced and maintained by skilled practitioners in the course of a game of football. In this understanding, practices thus emerge, stabilize and ultimately die out as the links between elements are made and broken (cf. Pantzar and Shove, 2006). By implication, generating more sustainable practices calls for the links and elements of existing, unsustainable practices to be challenged and broken before being replaced and re-made in more sustainable ways. Warde (2005) offers a vital insight into how the making and breaking of these links might occur, as he shows that despite their considerable

inertia, change in practices emerges both from the inside – as practitioners contest and resist routines and conventions and as they improvise new doings and sayings in new situations – and also from the outside, as different practices come into contact with each other.

Social practice theory, in this view, raises a series of radically different questions about how to create more sustainable patterns of consumption. The focus is no longer on individuals' attitudes, behaviours and choices, but instead on how practices form, how they are reproduced, maintained, stabilized, challenged and ultimately killed-off; on how practices recruit practitioners to maintain and strengthen them through continued performance, and on how such practitioners may be encouraged to defect to more sustainable practices.

In the rest of this article I will attempt to address some of these questions empirically through a detailed ethnographic case study of a pro-environmental behaviour change initiative. Before this can proceed, however, it is necessary to clarify my use of the terms behaviour and practice simultaneously. As may be seen from the reviews above, social practice theory represents a wholly different paradigm of social research than that found in linear models of behaviour. As such, some (e.g. Shove, 2010) argue that the terms practice and behaviour are incompatible, and that a continued focus on behaviour serves to obscure more than it reveals. This article does not fundamentally depart from this position, but it also contends that given the contemporary 'doing' of numerous 'pro-environmental behaviour change interventions', it would be unwise to ignore them altogether, and empirically misleading to call them by another name. Whilst preserving this important distinction, therefore, the aim in the rest of this article is to use practice theory's broader and more holistic perspective as a means of exploring what actually happens within behaviour change interventions and, in so doing, to try and understand the effects they have on the social practices they tackle (even if they tackle such practices unwittingly). Such analyses, it is hoped, will not only extend and improve policy and practical interventions in this important area, but also augment and develop social practice theory in an area where a more sophisticated understanding of social life and change is urgently needed.

The initial audit took place over a period of three weeks and involved the Champions team weighing all of the office's waste, taking electricity meter readings, and staying late after work to record which appliances had been left on. The headline results were then compiled by GAP to reveal that the site annually emitted 297 tonnes of CO₂ and sent 11.7 tonnes of waste to landfill, of which 58 percent could easily have been recycled.

Reflecting on the audit process in meetings and interviews, many of the Champions commented on how awkward it had felt to monitor their colleagues' normal practice:

Melissa² said 'you feel rude looking over people's shoulders and in their bins, especially if they're still in the office as it's hard to hide what you're doing', and Melanie concurred: 'you do feel rude.' They felt a bit uneasy about this, like they were 'checking up on people'. (FD, p. 19)

Such comments arguably reveal the difficulty of challenging the practical consciousness (Giddens, 1984) or even the habitus (Bourdieu, 1984) that enables practitioners to go on in everyday life in an unthinking manner. Further, the effect of 'looking over people's shoulders' may be seen as a disruption to what Goffman (1963) calls civil inattention. Rather than allowing normal practice to proceed by ignoring it, the Champions were beginning to pay attention to it and to challenge it in new ways.

When GAP first revealed the audit results to the team, they were met with many gasps and expressions of shock and surprise at the size of the impact the offices were having:

I found them [the audit results] really shocking...it's astonishing the amount of rubbish that the place actually produces...I wasn't the only person who just sat there and went 'crikey, how much do we throw away? [laughs]. We've got to do something about this'. (Melanie, FD, pp. 8-9)

As this quotation reveals, the waste products and environmental effects of normal working practices are largely invisible. The audit process was thus vital in helping to re-materialize inconspicuous consumption patterns, and also in localizing and connecting 'the environment' to everyday practice. Indeed, many of the Champions emphasized that the local nature of the audit results was important. That the results referred to their practice, as opposed to more general facts and figures about the environmental consequences of everyday life, seemed to be crucial in localizing the environment, making it real and relevant (cf. Burningham and O'Brien, 1994). Further still, the quotation also shows how the audit process served a motivating function, encouraging the Champions to 'do something about' routine practices that had previously gone unquestioned.

From a social practice-based perspective, the audit can be seen as problematizing the links between the images, skills and stuff of a whole bundle of practices. Quite suddenly, a critical environmental perspective was cast upon the unthinking disposal of waste and the habitual leaving-on of lights and appliances, amongst other things. In this respect, the audit appeared to provide the Champions with a certain amount critical distance from the bundles of practice in which they were normally closely entangled. Whilst they could never detach themselves from such practices entirely and 'steer [them] from the outside' (Shove and Walker, 2007: 769), the audit appeared to create a space within existing practices that enabled the Champions to question them and develop a sense of the new, pro-environmental directions they wished to travel in, a process that continued in the subsequent planning meetings.

Following the audit, discussion in the early planning meetings questioned and challenged multiple aspects of working life within Burnetts. The bosses and their cars came under particular scrutiny, as did the material fabric of the buildings on the head office site, the environmental impacts of the construction sites the company operated, as well as more mundane aspects of office practice such as recycling or turning appliances and lights off. Over the course of several meetings, the Champions gradually channelled this critical discussion towards addressing the audit results. What was interesting in these discussions was that whilst there was some focus on their colleagues' attitudes, values and beliefs, and the ways in which good communications strategies might be used to change these, this was dwarfed by the detailed attention the Champions paid to the various elements of everyday practice.

Much of the team's discussion centered on the 'stuff' and 'skills' of working practices and how they might be replaced with more pro-environmental alternatives. For example, as Figures 1 and 2 show, the Champions explored ways in which the normally unquestioned stuff of office practices could be reduced or replaced, as well as setting out detailed instructions – such as for a 'leaving the office checklist' or on how to photocopy and print double-sided – to try and introduce new pro-environmental skills around the offices. The images or meanings of existing practices were also discussed. Conventional approaches to behaviour change perhaps come closest to social practice theory here as attitudes and values might be seen as proxies for meaning. The difference, however, is that in a practice approach meaning is seen as residing within the practice rather than in individuals' heads. Throughout the team meetings the meanings of existing practices were invoked and challenged in various ways. A particularly clear example occurred when the team discussed setting targets for the reductions in environmental impacts they hoped to realize. As I recorded in my field diary:

Anna [the GAP programme manager] then got the group to commit to certain targets. This led to an interesting set of discussions...[someone] emphasized that 'we should

There is insufficient space in this article to discuss the implementation of all of the Champions' many proposals (see Hargreaves, 2008 for a more detailed discussion). Instead, this section will focus on one specific proposal that ran throughout the entire initiative and was emblematic of the challenges the Champions faced in questioning and replacing existing practices-as-performances: No Bin Day.

At the start of the initiative all employees had a bin by their desk, and by photocopiers and in kitchens there were paper/cardboard recycling bins, plastic cup recycling bins, and blue bins for confidential business waste, which was shredded before being recycled (see Figure 3). The No Bin Day proposal was to remove under-desk bins for a single day in order to de-routinize existing waste habits and re-routinize new ones (cf. Spaargaren and Van Vliet, 2000) based around the improved recycling facilities the Champions proposed to put in place beforehand. As the initiative progressed, however, it became very clear how difficult it was to challenge, let alone to replace, existing waste practices.

As soon as the No Bin Day proposal was mentioned, it ran into problems. First, David suggested that it may breach Burnets' contract with its cleaners as it would

leave them less work. Next, when the proposal was sent to Brian (non-Champion) the Facilities Manager, who was responsible for waste management at the site, he raised a number of objections: that waste in the wrong bin could breach data protection laws, that the lack of a bin for food waste could lead to hygiene problems and that placing new recycling bins in corridors could be a fire hazard. Whilst these may seem to be fairly trivial issues, in an interview with Brian he proudly pointed out the Facilities Management team's excellent health and safety record and suggested that, far from being anti-environmental, he had a 'duty of care' to all employees that forced him to consider a range of legal obligations:

People forget that they have to be clean to a certain standard, they have to be lit to a certain standard, they, you know, and all the things that Facilities Management do. They have to choose the right chairs, you know, and the right desk heights, and the right equipment you know... I've got a duty to protect [them] here. (Brian, interview, FD, p. 49)

As these initial objections show, bins are involved in and support a number of different practices, serving a wide range of different purposes at the head office site, a point Brian expressed in an email to Peter, one of the Champions, by arguing that it is 'someone's right to expect [a bin] as part of a normal office' (FD, p. 112).

As the Facilities team's objections mounted, the proposal faced its first serious challenge from within the Champions team at a meeting in May. Here, Louise argued that taking people's bins away was 'an invasion of privacy' (FD, p. 109) and referred to the CHANGE programme (a parallel initiative being run at Burnetts), which was 'all about encouraging people to choose the right thing, and then we're taking bins away and not offering them a choice' (FD, p. 109). Further, she imagined a scenario in which the Chief Executive was with a high-profile client who'd sneezed and had a dirty tissue: 'is he supposed to say, "oh just go to the bin at the end of the corridor?"' (FD, p. 110). Rather than abandon the idea altogether, she suggested the Champions should offer people a choice of whether or not to relinquish their bin.

These kinds of arguments continued and the Champions progressively watered down the proposal until, in Sally's words, it ultimately 'fizzled out' (FD, p. 146) in late July. The Champions ended up offering employees the choice of relinquishing their bin, an option no-one selected, and in conjunction with the Facilities Management team, rather than removing bins they in fact distributed additional desktop recycling trays (see Figure 4) to all employees. Brian also mentioned that these recycling trays had the added advantage of reducing fire risk by limiting the amount of paper people kept on their desks and ensuring it was stored flat.

Despite the active and committed effort of the Champions team, operating in an ostensibly supportive organization, as the No Bin Day proposal progressed, the manner in which under-desk bins mesh with other practices, professional and normative standards, legal regulations and even social interactions was gradually revealed, and the proposal was blocked at almost every step. The disagreements

between the Champions and Facilities Management team further reveal that some groups of practitioners, or communities of practice (Wenger, 1998), have vested interests in preserving the precise bundles of practice that make up the status quo. In this particular dispute, Facilities Management should be seen as a powerful group within Burnetts, possessing a proud health and safety record (an understandably important issue within the construction industry e.g. DTI, 2006), and able to back this up with legal arguments. By contrast, the Champions team were a newly formed, informal and voluntary group of practitioners who had little more than the normative power of the environment on which to make their case. Perhaps unsurprisingly, this argument proved too weak.

These vital issues, and the at times surprising connections between practices, would be missed in more conventional and functionalist analyses of pro-environmental behaviour change processes that focus exclusively on individuals' explicitly 'environmental' attitudes and values (cf. Hobson, 2002; Ungar, 1994). At the same time, this example has offered at least three ways in which social practice-based analyses might be extended. First, it suggests a need to look beyond single practices and towards the relations within and between whole bundles of practice that co-exist in particular domains of everyday life (cf. Warde, 2005). Second, it has

illustrated the close relationship between practices and the power and social relations that they support and uphold and which, in turn, ensure that those practices are maintained, stabilized and reproduced. These social aspects and (micro)politics of practice are often neglected (cf. Røpke, 2009) but it would appear they deserve closer attention, particularly in studies concerning attempts to change practices. Finally, this example has shown empirically how social order exists in and is supported by practices (Reckwitz, 2002: 251). By logical implication, changing practices means changing the social order and, as was seen in the Champions long-running dispute with the Facilities Management team, is a process that may well result in both winners and losers. Unsurprisingly, therefore, and as the Champions team gradually discovered, current arrangements of practice are likely to be strongly protected, leaving relatively little room for manoeuvre for those who wish to introduce changes.