Review of the literature relating to comfort practices and socio-technical systems

Transcription

1 Environment, Politics and Development Working Paper Series Department of Geography, King s College London Year 2010 Paper 35 Review of the literature relating to comfort practices and socio-technical systems Emma Hinton Department of Geography, King s College London 2010 by the Author(s) This paper is posted at King s College London, Department of Geography at

2 Acknowledgements: This report fulfils Deliverable 3 of the Carbon, Control and Comfort project. The author would like to thank the Engineering and Physical Sciences Research Council (EPSRC) and E.ON UK for providing the financial support for this study as part of the Carbon, Control and Comfort project (EP/G000395/1). Many thanks are also due to Karen Bickerstaff (King s College London), Harriet Bulkeley (University of Durham), Michelle Shipworth (University College London) and Anne Stafford (Leeds Metropolitan University) for providing constructive feedback on earlier drafts, and also to Justine Cooper (University of Greenwich) for suggesting literature to include in this review. [2]

3 Contents Abstract Introducing a socio-technical systems perspective The potential role of psychological and social factors on our interactions with energy system Social psychological approaches to understanding energy consumption Information-based approaches to behaviour change Meters, monitors and other information-based interventions The influence of social structure on individual energy consumption The parameterisation of comfort Comfort as a socio-historical artefact Comfort as an attribute Comfort as an achievement Driving change by intervening in the socio-technical system Socio-technical regimes and social structuring of technical innovation Ecological modernisation and sustainable home services Recognising the social life of things : driving change using non-human agents Comfort as an everyday practice A practice-based approach Empirical studies of comfort Concluding remarks and implications for the CCC study Appendix 1: Social-psychological studies Rational choice theory and the information deficit model of behaviour change Other social-psychological approaches References [3]

4 Abstract This paper approaches the literature on comfort and energy practices from a socio-technical systems perspective, reviewing relevant studies and recent thinking on how we study comfort and energy in the home. This constitutes a first cut at developing a theoretical position from which to explore the empirical data coming out of the CCC fieldwork. The CCC study aims to reduce domestic energy consumption through an understanding of, and interventions in, comfort practices. Whilst there is more to comfort than space heating and cooling, these two processes account for the lion s share of domestic energy use in most western societies (Shove 2003b: 396), where the aim of the CCC study is to encourage less energy intensive comfort practices rather than sacrifices in individual comfort in order to meet the desired reduction in energy consumption. The overlapping subjects of domestic energy consumption and comfort have been approached in a number of different ways within the psychological, sociological, technical and socio-technical literatures. Each of these literatures conceptualises agency differently, which has implications for the types of interventions that flow from a reading of these literatures; if we attribute agency to particular actors then this affects the types of interventions we develop and the types of results we expect them to have. For example, whilst a proponent of a socio-technical systems approach might argue that comfort can only be understood through a perspective that accounts for historical, technical and social change (Wilhite, 2009: 85), psychological approaches tend to stress the significance of individual attitudes and values, sociological approaches tend to stress the role of social structure, and technical approaches tend to stress the significance of the ability for different configurations of technologies to produce environmental conditions under which the majority of participants report feeling comfortable. In this paper, psychological, sociological and technical approaches to understanding and intervening in energy consumption and comfort practices are discussed from a socio-technical systems perspective, in order to draw out commonalities and empirical precedents that can inform our approach to and analysis of data collected during the CCC study. This paper is structured as follows. Section 1 introduces a socio-technical systems approach, and situates this alongside other approaches by distinguishing between the objects of these approaches and their attribution of agency to different human and non-human actors. Section 2 focuses on psychological and sociological approaches to understanding and driving particular types of individual-level interactions within energy systems, including a discussion of the potential for information-based interventions to result in behaviour change. Section 3 considers the parameterisation of comfort, or the ways that comfort has been constructed and institutionalised. Section 4 explores comfort as a socio-historical artefact, drawing from the socio-technical systems literature and distinguishing between two main ways of constructing comfort: as an attribute and as an achievement. This section concludes with a discussion of the types of intervention that may be appropriate to influence the socio-technical systems that comfort practices are both produced by and are involved in reproducing. Section 5 focuses on comfort as an everyday practice, including a [4]

5 discussion of practice-based approaches and empirical comfort studies. The paper concludes with section 6, which summarises the implications emerging from this literature review for the CCC study. 1. Introducing a socio-technical systems perspective The mainstream approach to producing comfort has come to focus on measuring physiological responses to a set of environmental conditions, which may be adjusted through the tactical application of different technical systems or artefacts (discussed in section 3). This approach has come to colonise building fabrics and shape the ways that we conceptualise comfort; in this way it also structures the ways that it is possible for us to practice comfort. A growing body of literature calls for attention socio-cultural and socio-technical factors in understanding and operationalising comfort. In contrast to institutionalised conceptualisations of comfort, which consider it to be a relatively a-social category of being related to individual physiological experience, a socio-technical view of comfort sees it as being socially and historically situated (Shove, 2003a, Shove, 2003b); shaped by collective conventions, which co-evolve with technical systems (e.g. Shove, 2003b, Guy, 2006); context-dependent (Hitchings, 2009); socially constructed, contingent and negotiable, where its meaning varies with time, space and place (Chappells and Shove, 2005, Shove, 2003a); culturally negotiable (Fountain et al., 1996, Cole et al., 2008, Shove, 2006); and understood in different ways even within expert circles of architects, building services engineers, property developers, manufacturers and regulators (Chappells and Shove, 2005, Shove, 2003a). Comfort as viewed from a socio-technical systems perspective represents a distinct position that differs from other approaches such as those grounded in psychological or sociological understandings of influences on individual behaviour which sees agency distributed throughout socio-technical systems rather than residing in one actor or type of actor and emphasises the role of practices in constituting these systems. The socio-technical systems approach has evolved from the science and technology studies (STS) literature along with a number of related approaches. Each of these STS approaches stands in opposition to the approach referred to as technological determinism, which holds that technology not only develops outside of society, but can also influence it from the outside. In contrast, the approaches developed broadly within STS which include actor-network theory, social construction of technology and social shaping of technology theories see technology and society as inextricably intertwined, conceptualising the link between society and technology in subtly different ways. Actor-network theory recognises the agency of non-human actors (we could consider in this context that local and national infrastructures have agency, as do other technologies that householders may use to practice comfort) and sees these actors as being networked together, where networks are ephemeral but appear permanent through their being continually (re)produced. Social construction of technology theory proposes that human actors shape technology (rather than our actions being determined by it) and that to understand how a technology is used, we must understand the ways that the technology is embedded within its social context. Social shaping of technology studies recognise that multiple sources of influence including social, institutional, cultural and economic factors may shape the direction and rate of technological innovation. In socio-technical systems approaches, society and technical structures are considered to be part of a complex, inter-related system, where to drive change we must consider the system as a whole: including people, things and [5]

6 their immediate physical and social context. However, due to this heritage in the STS literature, the literature on socio-technical transitions tends to emphasise the role of systems and technologies at the expense of practices (Shove and Walker, 2010). It is here that the CCC study can make a contribution to the growing body of socio-technical systems-inspired literature on comfort and its relationship to energy systems. Exploring comfort and energy use from a socio-technical systems perspective represents a relatively recent turn in academic research on the subject. Other approaches focus on individual attitudes and behaviours; or the role of social structures on individual action; or the technical conditions for producing comfort in indoor environments. The object of these approaches can be caricatured as follows: psychological approaches focus on the mind of the individual; sociological approaches focus on the social body; technical approaches focus both on the individual body and its physiological responses, and on the ability of technologies to meet physiological needs. A socio-technical systems approach to comfort recognises the interplay of each of these elements, networked together into socio-technical assemblages that span multiple levels, conceptualising both human and non-human actors as having agency in terms of shaping comfort practices and constituting a wider sociotechnical regime. Practices are an important element of socio-technical systems: they are not simply the sites at which socio-technical regimes are enacted, but they also act to reproduce and sustain these regimes and continually produce and reproduce them at the domestic level (Shove and Walker, 2010). In contrast to psychological approaches that tend to locate the majority of agency in comfort-related actions within individual decision-making processes, sociological approaches that tend to conceptualise agency as shared between individuals and social systems, and technical approaches that tend to locate agency in comfort provision with technological systems and user interactions with these systems, socio-technical approaches that recognise the significance of practices in constituting the socio-technical regime locate agency in each of these social actors in addition to the practices themselves, which constitute particular forms of know-how distributed across the social body. These ideas, and their potential to contribute to our understanding of comfort practices, are discussed in more detail in the rest of this paper. 2. The potential role of psychological and social factors on our interactions with energy system One of the primary goals of the CCC study is to support less energy intensive means of providing comfort in the home. We practice comfort through interacting in socio-technical assemblages comprising arrays of human and non-human social actors and different sets of practices may be associated with different energy requirements. The majority of studies of energy consumption in the home within the social sciences literature have either focused on better understanding the psychology of individual energy consumption behaviours, or on the influence of social structure. A socio-technical approach represents an alternative take on understanding this, focusing more on the agency of non-human social actors (e.g. technologies of different kinds) organised into networks of social actors at different scales, all of which influence the ways we practice comfort in the home. This section discusses the potential for an integrated approach to understanding why we consume energy in the ways that we do that draws from all three social scientific approaches. [6]

7 As viewed from a socio-technical systems perspective, our comfort practices are considered to be structured by (and also themselves structure) both collective conventions and technical systems, where all three practices, conventions and systems co-evolve. The strength of this approach is its emphasis on the socially situated and networked nature of both technologies and practices, yet it could be criticised for paying relatively little attention to the potential role for individual psychological factors in practicing comfort, or socio-cultural structure, in contributing to this coevolution of technologies and practices. Skea (2009) argues that a productive approach to understanding and driving change in domestic energy consumption would combine psychological understandings of behaviour with social scientific understandings of the ways that comfort needs are socially constructed; arguably both psychological and social scientific understandings could be further enriched by considering the ways that socio-technical assemblages can structure behaviour. The socio-technical view of comfort could be enhanced by including some consideration of the factors that might motivate an individual to undertake particular actions and consider these as part of (and partially producing) the socio-technical system. The social and psychological literatures could also be augmented by acknowledging the potential role of socio-technical systems in structuring individual attitudes and values, and social norms. A more holistic approach to understanding comfort might consider practices, attitudes, values and norms as producing and produced by the socio-technical structure, where agency takes a networked form and cannot be said to reside in one actor or type of actor. Such a view contrasts with psychological approaches, which tend to emphasise an individual s agency to act on their attitudes and values, and social science approaches that tend to emphasise the structuring effects of social and cultural norms and institutions. There is a rich body of empirical studies in the social science and psychological literature that have focused on understanding and motivating particular energy consumption and pro-environmental behaviours, but this has tended to pay relatively little attention to exploring comfort and related practices. The remainder of this section comprises two sub-sections: the first focuses on information-oriented approaches to understanding and motivating particular energy consumption behaviours, drawing from psychological studies; the second focuses on understandings of the influence of social structure on individual behaviour. Further detail and related empirical studies are included in appendix Social psychological approaches to understanding energy consumption One long-standing approach to understanding energy consumption in the home involves exploring individual psychology, typically seeking to link attitudes and values (including the influence of social factors, in some models) to individual behaviour (see Jackson (2005) for a comprehensive review of social psychological models of behaviour, with a focus on driving pro-environmental behaviour). Several scholars have distinguished between approaches that focus on individual characteristics, attitudes and values, and those that emphasise the role of social and cultural factors in shaping these attitudes and values. These two broad approaches have respectively been referred to as internalist and externalist (Jackson, 2005), or as attitude-behaviour connection models and consumer-motivation theories (Hargreaves et al., 2008). Where the former approach tends to focus on individual agency and an individual s values and norms, the latter approach recognises that individual agency may be limited by social and cultural factors including social norms and social, [7]

8 economic, demographic and political institutional contexts. Jackson sees these two approaches as being in tension, where in the first perspective, enlightened consumers are free to choose proenvironmental behaviours assuming that they possess appropriate beliefs or attitudes; in the second, consumers may be locked in to consumption choices by a variety of external conditions ranging from genetic conditioning to economic necessity, social expectation, accessibility constraints and the creeping evolution of social norms (Jackson, 2005: 24). A socio-technical stance might also recognise the role of non-human actors such as technologies of various kinds in addition to practices, as social agents within an externalist model. However, it is important to note that the distinction between internalist and externalist approaches is a generalisation and does not necessarily imply a dichotomy: there may be overlap. Whilst internalist psychological models such as Fishbein and Ajzen s seminal work (the Theory of Planned Behaviour and Theory of Reasoned Action) focus on individual psychological factors, they do not necessarily make the claim that solely individual attitudes and values explain action and acknowledge that other factors, including the social, may have a role. Within the psychology literature Stern (2000) has argued that whilst individual attitudes, values and intentions can tell us something about what people do and why they do it, that a number of factors limit individual choice: including individuals acting in social groups (e.g. in industrial complexes or in policymaking circles), infrastructural factors (which materialise previous decisions made by certain social groupings) and situational constraints, such that attitudes and values may be situationally determined rather than stable dispositions. In general, social-psychological approaches to understanding energy use (where this may overlap with comfort practices) tend to focus on individual abstracted from their structural (social, cultural or sociotechnical) contexts. Energy consumption-related actions, or intentions to act, have tended to form the focus of this corpus of literature, rather than comfort per se. The rational choice model, information deficit model and theory of planned behaviour all implicitly construct individuals as possessing almost complete agency to decide on their actions, behaviours and practices; whereas the norm activation model locates some of this agency to social systems of norms (these models are discussed in appendix 1). These types of psychological, individual-focused approaches to understanding behaviour have been criticised for exhibiting methodological individualism where social behaviour is assumed to be derivable from the aggregate of individual behaviours such that the effects of others in shaping our actions or identities, i.e. the social nature of decision-making and the influence of social structure, are ignored (Jackson, 2005). None of these approaches acknowledges the potential for non-human or material agency, although critiques have argued for attention to the effects of institutional and infrastructural context on individual agency which inevitably includes a non-human dimension. Further, psychological approaches reproduce divisions between individuals and socio-technical systems (Spaargaren, 2000). If we assume that an individual s immediate socio-technical system (or sub-system) i.e. the relational array of people and things in their domestic environment is in some way related to a suite of socio-demographic factors (perhaps resulting from tenure type, income level, family size etc), then it may be that those studies that identified associations between energy consumption actions and socio-demographic factors implicitly suggest that different socio-technical systems may support different kinds of action. According to socio-technical perspectives, sustainable practices depend less upon the diffusion of specifically green beliefs and actions or the cultivation of appropriate attitudes, and more on the ways that social and technological systems and collective conventions co-evolve [8]

9 through ordinary practices. This is not to argue that attitudes and beliefs, or mental models of behaviour and understandings of energy and the environment, are not involved in energy use and comfort practices: however, it is to argue that they may simply be one of a range of other influential factors which may be more or less dominant, where these factors can be considered to be networked together into a more-or-less self-reinforcing system. 2.2 Information-based approaches to behaviour change Information-based approaches to behaviour change, which may take many forms and tend to be supported by policymakers and practitioners alike work on the assumption that provided with the right information, an individual will make the right choice. This type of rationale is grounded in social psychological knowledge of individual behaviour, particularly the rational choice model, the information deficit model of behaviour change, and models of individual attitudes and values such as the Theory of Reasoned Action and Norm Activation Model (discussed in appendix 1). Arguably, through focusing at the individual level and in isolation to any social or physical context, these models fail to acknowledge the influence of the organisation of these different factors (including people and things) into socio-technical systems, where effecting change (no matter how rational at the individual level) may necessitate overcoming the inertia of the system, which may be more difficult than succumbing to it. A focus on providing individuals with (better) information about their energy consumption could be accused of re-entrenching a techno-economic stance and presupposing that individuals are self-interested, knowledgeable, and economically calculative when considering energy measures where with careful monitoring and scientific control, it is possible to reproduce the technical achievements of the laboratory universally (Guy, 2006: 647). However, instead of behaving like rational actors, individual attitudes and decisions are always shaped and framed within wider social processes and, as such, are subject to change where abstraction of the opinions and outlook of energy consumers, and of design and development actors from the contexts of production and consumption, tends to isolate and freeze what are always contingent practices (Guy, 2006: 650). If individuals are enmeshed in complex sociotechnical systems that shape their experiences and practices of comfort, then comfort cannot be reduced to automated control or information provision (Jaffari and Matthews, 2009). Top-down, information-deficit approaches to behaviour change may be unsuccessful because they fail to engage with or utilise vernacular knowledges of energy consumption and the links between comfort and energy. According to this view, individuals are not empty vessels for new, technical information; instead, we are knowledgeable agents who operate according to our own vernacular knowledges. These vernacular knowledges may be worked up through personal experience of interacting with technological systems within the home, as well as being negotiated in social groupings of different kinds and at different times. Approaches that recognise the validity of lay understandings of issues, which balance these with objective truths from experts such as participatory, deliberative civic models of public engagement may be more successful in generating pro-environmental behaviour than those channelling top-down expert information (Owens, 2000). There may be some differences between vernacular and expert knowledges in the context of domestic energy systems and how they work. Kempton s studies, which have explored lay [9]

10 understandings of domestic energy use across participants with different socio-demographic characteristics in north America, have found such a disparity (Kempton, 1986, Kempton, 1987). He found that participants routinely develop their own theories to explain the world around them [and] these theories can be useful even when they contradict conventional technical wisdom (Kempton, 1986: 75). These folk theories which provide particular ways of understanding thermostats, heat loss and comfort are distinct from (but influenced by) institutionalised theories, specifically an engineering model (focusing on heating system operation, fuel consumption and heat loss) and a physiological model (factors affecting human body temperature). Kempton identified variation in behaviours and folk models between households but also within households, and variation in whether one person or multiple people within a household controlled the thermostat. On the whole, these models were sufficiently common across participants that Kempton proposes they constitute a kind of common sense set of behaviours that circulate through word of mouth. That is, these behaviours despite being practiced by individuals are constructed and negotiated in social situations. If folk models of energy consumption are as widely adopted as Kempton s research suggests, then information-centric approaches to adjust energy consumption behaviours that rely upon institutionalised, technical information may work less well than those that speak to whichever folk understandings are at play in each household. We might think of vernacular understandings of energy consumption and links to comfort as being created through particular kinds of knowledge practices, which go on to be enacted physically in comfort practices. In addition, Kempton s studies seem to suggest that non-human actors such as technological artefacts like thermostats have some agency in the development of these folk theories, where individuals may negotiate their understandings of how heating systems work through interaction with these controls, through observations and stories about how they work, and through discussions of these stories and theories with their peers. 2.3 Meters, monitors and other information-based interventions As already noted, policymakers have favoured information-oriented approaches to generating behaviour change that supports goals such as pro-environmental behaviour and a reduction in energy consumption within the home. Information-oriented approaches are not restricted to those reliant upon awareness-raising campaigns: they may also centre on technological tools installed within individuals homes (as set out in table 1, taken from a recent report to Defra by Doble and Bullard (2008)). Table 1 separates out information and technological tools, yet arguably some of the technological tools also have an informational element, such as those listed in the active tools column. The approaches proposed by Doble and Bullard work within the existing energy infrastructure and attempt to introduce new elements into the domestic socio-technical assemblage. Different approaches attribute different elements of the socio-technical assemblage with different amounts and types of agency in the control of comfort within the home: information tools and tariffs attribute individuals with agency to control their own comfort, echoing rational actor and information deficit models of behaviour change; passive technological tools remove agency from the individual in the automation of comfort provision; whereas with active tools agency is shared between individuals and technologies. The effects of the introduction of such interventions into domestic socio-technical assemblages, and the effects on the ways that [10]

11 individuals practice comfort in their homes, tends not to be considered in accounts of such interventions; this constitutes another area where the CCC study can make an important contribution to the literature. Information tools information on awareness raising efficiency of consumer goods & buildings General Energy labels, education and procurement tailored advice information for both consumers and supply chain, better billing, total energy consumption raw data, presentation of interpreted energy consumption data, appliancelevel consumption data information on reducing emissions Generic advice on energy reductions, tailored advice (energy audits) passive tools Retrofit heating controls, remote control of supply, stand-by elimination, advanced heating controls, dynamic demand control, automated lighting, shower heads, heat recovery/ exchangers Technological tools active tools Programmable and communicating thermostat, home energy hubs, smart appliances, smart homes Tariffs Time of use tariffs, lower rates for lower security of supply, increasing block tariffs, pay as you go (prepayment) tariffs Table 1: List of potential interventions to reduce energy consumption in the home, derived from table 10 in Doble and Bullard (2008) entitled Overview of impact of different types of tools. Information-based approaches may have some value in making visible otherwise relatively invisible domestic energy consumption (Burgess and Nye, 2008) and as a result may be able to drive sociotechnical change to some extent. For example, energy labelling has supported individuals in choosing to consume energy efficient technologies and has simultaneously encouraged manufacturers and designers to produce such products (Shove, 2006). The ineffectiveness of conventional meters for making energy consumption visible was emphasised by Kempton and Layne (1994), who found in a series of in-depth interviews that individuals have limited ability to receive and analyse price and consumption data mediated by utility companies. Another study (Kempton and Montgomery, 1982) found that participants who monitored their monthly energy consumption, changes through time, comparisons between appliances and length of payback periods typically made errors in these quantifications even those participants that understood the technicalities of energy measurement through adopting cognitively efficient albeit ultimately erroneous methods of calculation. Smart meters would overcome some of these problems. In contrast to normal electricity and gas meters, smart meters present energy consumption information in a more comprehensible form. Despite Burgess and Nye s (2008) argument that the widespread installation of smart meters into domestic stock had not yet occurred because it was, at the time of writing, cost prohibitive and would pose an excessive burden on the communication networks that the widespread roll-out of two-way smart metering would rely upon to work, in October 2008 DECC announced its intention to roll out smart meters to all homes in the UK by Through this [11]

12 regulation, smart meters will become normalised and integrated into every home during the next decade; it is possible that energy consumption (and possibly comfort) practices may well change as a result of this technological intervention. Pre-payment meters, an existing metering technology that is already installed in millions of households across the country, may be better at making energy consumption visible than standard meters without the potential infrastructural issues associated with smart meters. Pre-payment meters may enable customers to better control the amount they are spending due to an increase in the frequency of feedback on electricity consumption and the psychological difference in paying outright for something rather than building up debt (Doble and Bullard, 2008: 67). However, prepayment meters tend not to be considered as an appropriate means of making energy visible in attempts to intervene in domestic energy consumption. It has become a truism that prepayment tariffs are inevitably more expensive than others and so exacerbate fuel poverty (e.g. Burgess and Nye, 2008, Doble and Bullard, 2008), yet this is not true for several energy providers (e.g. EDF, EBICO). Pre-payment meters are also considered to be inconvenient, necessitating trips out of the home in order to buy electricity and vigilant self-monitoring of energy consumption to ensure that meters do not run out of credit. However, this could instead be viewed positively (for those mobile enough to be able to leave the home at regular intervals, at least), following Doble and Bullard s argument that this type of activity does a particular type of psychological work. Further, visiting local shops in order to buy energy could in itself be considered a form of sustainable consumption behaviour, where using local shops and supporting the local economy are considered more sustainable than using out-of-town supermarkets for all our shopping. On the negative side, prepayment meters are often located out of sight and so they may still fail to make energy consumption readily visible. Further, it is possible that because these meters are associated with particular demographics i.e. households with low incomes, lone parents, people in receipt of benefits, those without bank accounts, and those with payment difficulties (National Energy Action 2008, reported in Doble and Bullard (2008)) that a social stigma may be attached to them. Whilst there is some evidence that the use of pre-payment meters is associated with reductions in energy consumption, there is little understanding as to the particular mechanisms by which this occurs (Doble and Bullard, 2008) just as there is little understanding of the processes by which energy monitoring devices are able to facilitate modest reductions in energy consumption (Hargreaves et al., forthcoming). Attention to the ways in which individuals interact with pre-payment meters and conventional meters may contribute to our understanding of how individuals practice comfort in more or less energy intensive ways. Despite their potential to increase the visibility of domestic energy consumption, the information provided by pre-payment meters and smart meters does not easily allow individuals to see the amount of energy that is consumed when they practice comfort in particular ways, nor does it make visible the different amounts of energy that may be consumed by different individuals in the same household at the same time but engaging in different practices. Energy monitors may go some way towards countering some of these difficulties, allowing individuals to see the energy intensity of different comfort practices that may involve specific assemblages of people and things. Whilst prepayment and smart meters provide a general indication of energy consumption, energy monitors [12]

13 offer one of the most direct feedback mechanisms on energy use and so potentially allow individuals to link their energy use to their behaviours in real time (Burgess and Nye, 2008). Whereas a technologically determinist perspective might take for granted the interactions between individuals and devices, some studies have suggested that energy monitors as new non-human residents (or guests) in people s homes may not always be seamlessly accepted into the social fabric of the home, or interacted with in predicted ways. In order to begin to develop an understanding of the ways that these devices are used in the home to produce energy savings, Hargreaves et al (forthcoming) conducted a study involving 15 UK householders trialling three different energy monitors. Energy monitors are directed at the individual level of energy consumption behaviours, and this study found that it was only isolated individuals within households typically male household members; the authors propose that female householders were less interested because they found the gadgety nature of these monitoring devices less appealing than the male participants who interacted with them and used them to adjust energy consumption. This one man and his energy monitor style of interaction was linked with household-level disputes, where the monitors facilitated new forms of surveillance between household members. The monitors were also found to appeal to younger children, some of which integrated the energy monitors into their school work, although older children were significantly less engaged. On the basis of their empirical evidence, Hargreaves et al argue that interventions should be targeted at the household rather than the individual level, since domestic energy consumption is a social and collective rather than individualised process. This suggests that energy monitors like the three trialled in Hargreaves et al s study, which tend to be used by (and appeal to) individual members of the household, need to be redesigned to work better at the whole household level. A second key finding of this study was that the degree to which these monitors were accepted aesthetically within the household, by the majority of household members, affected the degree to which they were visible within the home and, by extension, the degree to which they made energy consumption visible to household members. Monitors that were accepted aesthetically were kept on display in prominent places easily visible to the household, such that energy consumption was comparatively more visible than in other households where the monitor was moved out of sight because it was not accepted aesthetically. The design of monitors, then, appears to be an important factor in whether or not they are accepted into the home and, by extension, the impact they can have on energy consumption. A third important finding was that these monitors did not have universally positive effects, despite participants opting into participation in the trials as a result of personal commitments to saving energy, saving money or saving the environment. Whereas in some households, interactions with the monitors did encourage energy saving practices, in others they encouraged feelings of despondency and guilt, where structural constraints prevented other forms of action. The authors report that interviewees strongly criticised appliance manufacturers for making devices that were difficult to switch off completely; housing associations and local authorities for planning policies which made it difficult to install solar panels, heat pumps or small wind turbines; housing developers for not automatically installing efficiency and generation measures in new homes; and they criticised the government and politicians for failing to match rhetoric with action on energy conservation and climate change issues. Energy monitors, then, appear able to encourage limited, structurally supported (and supported by the domestic sociotechnical assemblage), individual-level behaviour change where those individuals are (generally [13]

14 speaking) either young children or adult males and those who have a positive psychological reaction to the new visibility of their energy consumption made possible through these monitors. The monitors may not be seamlessly integrated into domestic socio-technical assemblages, where their aesthetic acceptability and fit with the rest of the material objects in the household affected their acceptance and use. Whilst these studies have not specifically explored comfort or the maintenance of comfort conditions through the monitoring of energy consumption made possible by various feedback devices, their findings provide insights into how information-based feedback devices produced and deployed as part of the CCC study may be received within the home, the degree and type of change that they may support, and the extent to which they may be integrated (or not) into everyday social life and the domestic socio-technical system. 2.4 The influence of social structure on individual energy consumption An alternative approach to understanding individual energy consumption and, by extension, the extent to which individuals may choose to adopt more or less energy intensive comfort practices - emphasises the potential for social, cultural and organisational factors to structure our actions (e.g. Owens and Driffill, 2008, Wilhite and Lutzenhiser, 1999, Kempton et al., 1992). As Jackson puts it, pro-environmental behaviour and actions depend critically on social conditions: rules, regulations, cultural norms and expectations, government itself, and the set of institutions that frame and constrain the social world (Jackson, 2008: 55). In contrast to the social-psychological approaches discussed in section 2.1, specifically those that broadly expect action to flow from intention, these approaches imply a greater role for material agency: individuals still have agency but this is mediated by structures external to the individual, which may include social and cultural norms, institutions, infrastructures and other material manifestations of social life. Here, specific behaviours can be thought of as simultaneously individualised as well as social and situated (Bulkeley and Gregson, 2009). In this vein, Kempton et al argue that individual energy consumption choices and actions may be constrained by financial or social relations of different kinds, where the poor cannot afford to reinsulate their houses, and if they are renters, they may not be allowed to. [...] And most people are reluctant to take extraordinary measures without social support from others who think the goals are important (Kempton et al., 1992: 1219). If we follow Kempton et al s line of argument we might expect that participants in the CCC study, as renters of social housing, may have limited agency to adopt certain energy efficient practices specifically those involving alterations to the building fabric and those requiring significant financial outlay. Yet because residents of social housing live in properties that are owned and managed by Registered Social Landlords (RSLs), which are subject to a suite of responsibilities and commitments (e.g. the Decent Homes Standard) towards the installation and maintenance of particular energy efficiency supporting technologies such as insulation, double glazing, heating controls and efficient heating systems, these tenants may in fact have greater access to an energy efficient domestic infrastructure than those living in privately owned or managed accommodation. At the same time, social housing tenants have limited control over the specific contractors and specific technologies and materials that are installed into their [14]

15 homes, since these decisions tend to be made by the RSL, removing a degree of agency from the tenant. Kempton et al also flag up the importance of a socially supportive environment to drive particular energy-related behaviours, which may be mediated by NGOs as part of a wider social movement. This type of argument has also been made by scholars studying one UK-based NGO s behaviour change programmes, Global Action Plan (GAP), which sets out to cultivate supportive social groups in local areas to drive change (e.g. Hargreaves et al., 2008, Hobson, 2003, Hinchliffe, 1996, Maiteny, 2002). Local social networks may also influence individual energy consumption behaviours, in addition to (or instead of) membership of wider social movements mediated by NGOs. The influence of local social and cultural norms on energy consumption practices, in relation to particular material artefacts, is illustrated in Wall and Crosbie s (2009) study of the uptake of different lighting technologies in the home. They found that some participants chose inefficient lighting because it had been seen in others homes and appeared desirable, where participants negotiated their own personal experience with this kind of social influence and information made available through their personal social networks in order to make decisions. Wall and Crosbie s study illustrates the range of factors that influence energy consumption decisions in the home: particular types of bulbs were associated with particular qualities (in terms of the quality of light, the aesthetics of the bulb, the length of time bulbs take to come on fully, their perceived cost and reliability), where their respective desirability or acceptability in the home was negotiated through social interaction in addition to the ability for particular technologies to fit within the existing domestic technical infrastructure (where, for example, most CFLs are incompatible with dimmer switches, and may stick out of some light fittings which may be deemed unattractive). Some argue that institutional, infrastructural factors limit our ability to act in certain ways. Our ability to adopt energy efficient practices is constrained because we live in a built environment constructed by powerful, embedded supply-side interests (Skea, 2009: 76). The built environment, comprising the electricity grid, heating infrastructure and drinking water systems, underpins and to some extent structures our everyday consumption (Spaargaren, 2000); as Ropke argues, practices co-develop with changes in production technologies, supply chains, transport infrastructure, exchange institutions, retail systems (2009: 2496). From a socio-technical perspective, the built environment and the full gamut of infrastructural and institutional structures are considered to have co-evolved with social practices, and are socially and historically situated. Whereas a century or more ago domestic dwellings and their inhabitants would have developed and deployed entirely different sets of practices in order to access resources such as energy and water, apparently ratchet-like socio-technical change has led to the co-evolution of alternative, energy-intensive practices and built environments, which to some extent lock us into particular practices. To address such lock-in, Jackson (2005) argues for social rather than individual change, a view echoed in the Sustainable Consumption Roundtable report, I will if you will (2006), in which he also had a hand. Yet this appears to pose a chicken-and-egg situation: which must come first, individual or collective change? And how can they both occur simultaneously, if each depends upon the other? The individual and the social are not mutually exclusive: individual behaviour plays a role in (re)producing social structure, such that the social can be changed when individuals or groups exercise agency (Cole et al., 2008). Successive, local and incremental changes to the social structure [15]

16 in this way can go on to influence individual behaviour, since individuals make choices but always within specific social, cultural and geographical contexts, according to the agreed rules of particular social practices which, of course, will change as actors experiences, interests and capabilities change (Burgess et al., 2003: 276). The notion of consumer lock in appears to gloss over the potential for individual agency, instead constructing individuals as more-or-less passive entities trapped within sociocultural frameworks that they cannot resist, let alone break free from. However, studies indicate that individuals do exercise agency (at both the individual and the social level). Individuals may actively resist attempts to coerce them into undertaking particular behaviours: for example, in their study of efforts to cultivate sustainable waste reduction habits, Bulkeley and Gregson (2009) found that some social housing residents resisted their landlords and interventions that were perceived as being associated with them. The relationship between social landlord and tenant may, therefore, influence the extent to which participants in the CCC study are receptive to undertaking particular actions. In order to resist the influence of social and technical structures, individuals must have strong internal motivations to undertake these actions (Maiteny, 2002); just as it is unreasonable to construct all individuals as being the passive victims of consumer lock-in, it is unreasonable to construct all individuals as active agents ready and able to resist structural factors. Whereas lock in tends to be framed negatively, it may also have positive consequences: if, for example, external factors such as regulation or economic instruments lock us into particular pro-environmental behaviours (Owens and Driffill, 2008), then this can potentially encourage a re-evaluation of the self such that individuals may consider themselves to be more environmentally-oriented as a result (Jackson 2005). Whilst it may be possible, in this way, to change behaviours without changing attitudes, this behaviour change may be relatively ephemeral especially if it is not rooted in, or driven by, meaningful experience (e.g. Maiteny, 2002). Giddens (1984) structuration theory provides a means of moving beyond the structure-actor dualism. Here, practices are the basic ontological units for analysis (Ropke, 2009: 2491). According to this view practices, rather than signs or abstract structures, are key to both constituting and understanding the social realm (Gram-Hanssen, 2010: 153). Actions are considered to be processes rather than discrete entities, and the social is constituted by and through interactions between individual agency and structural constraints and opportunities. Structuration theory has been influential in the development of practice theory (Gram-Hanssen, 2010) (discussed in section 5.1). Here, everyday practices, such as those of the household, are not simply mundane acts but rather daily habits and practices that constantly create and recreate social ordering (Hobson, 2003: 103). That is, through the reproduction of practices across time and space, individuals contribute to the production of social order (Ropke, 2009). If this is the case, then it may be more productive to explore collections of practices or lifestyles than to focus on individual attitudes (Spaargaren, 2000). According to structuration theory, we may engage as individuals with practices in two ways: through our practical and discursive consciousnesses. Routine, everyday activities are rooted in the practical consciousness, which does not require us to consciously reflect on our actions; we access a discursive consciousness that draws on our knowledge and experience when we consciously decide to undertake an action or practice. Since practices enacted by individuals in their everyday lives go on to produce the social structure, individuals may be able to [16]