Who’s responsible for food waste? Consumers, retailers and the food waste discourse coalition in the United Kingdom

Drawing on empirical research, including interviews with 38 key informants, this article examines how the challenge of food waste reduction has come to be framed, interpreted and responded to in the United Kingdom, focusing on household food waste and the interface between supermarkets and households. We identify a ‘discourse coalition’ arising from collective actors central to the issue that has achieved discursive hegemony over the framing of food waste as a problem. We analyse this discourse coalition – its core storylines, actors and practices – and the conditions that have enabled its emergence. Critical accounts of sustainable consumption commonly note the ‘responsibilisation of the consumer’: or the reduction of systemic issues to the individualised, behavioural choices of the ‘sovereign consumer’. We find, by contrast, that the ‘responsibilised consumer’ is by no means the discourse coalition’s dominant framing of the problem of household food waste. Instead, its dominant framing is that of distributed responsibility: responsibility distributed throughout the production–consumption system. The article also contributes towards understanding why retailers have embraced household food waste reduction as an object of intervention without framing the issue as one of, primarily, consumer responsibility.


Introduction
Food waste has come to assume a growing significance over recent years in the public sphere, in public policy circles and in the initiatives of food retailers and producers. The economic and environmental consequences of food waste have become widely recognised internationally. Roughly, one third of food produced for human consumption is lost or wasted, with direct economic consequences put at $750 billion globally annually (Institution of Mechanical Engineers (IME), 2013). In the United Kingdom, roughly half of food waste arises in households, and of this, about 60% is deemed 'avoidable' (Waste and Resources Action Programme (WRAP), 2012). A series of widely publicised reports, from the United Nations (UN) Food and Agriculture Organisation (FAO; 2011), the IME (2013), the House of Lords (HoL, 2014) and others, have quantified the scale of global and national food waste, sought causes and apportioned responsibility.
In 2008, the WRAP, a not-for-profit agency set up by the UK government to help advise on diverting waste from landfill, published the first figures quantifying the amount of food wasted by households in the United Kingdom (WRAP, 2008). WRAP's pioneering work placed the United Kingdom at the forefront of the issue, a position widely acknowledged to hold at the time of our research in -2015(see, for example, HoL, 2014. WRAP's (2008) report produced a flurry of media coverage, informing the UK public that it is was responsible for throwing away 5.3 million tonnes of edible food a year. The Independent's front-page headline ran, 'WHAT A WASTE! THE SCANDAL OF OUR THROWAWAY SOCIETY' (Independent, 2008). WRAP's subsequent reports on household food waste continued to produce headlines and popularised the issue's environmental significance, noting, for example, that 'eliminating avoidable household food waste in the UK [would be] equivalent to taking 1 in 4 cars off the road' (WRAP, 2012). Both on the national and global stage, appetite for the issue has remained high.
This article examines how the challenge of food waste reduction has come to be framed, interpreted and responded to in the United Kingdom, focusing particularly on household food waste and the interface between supermarkets and households. Discursive constructions among key collective actors framing the issue -grocery retailers, policymakers, third sector organisations and sustainability consultancies -are explored, and a 'discourse coalition' (Hajer, 1995) is identified. Hajer (1995) defines a 'discourse coalition' as a set of 'story lines' (or narrative understandings), the actors who utter those storylines and an identifiable set of practices in which this discursive activity is produced and reproduced. We address the discursive dynamics of this coalition's framing of food waste reduction in relation to the widely acknowledged trend towards attributing responsibility to individuals as consumers (Rose, 1999). Evans et al. (2013), drawing on broader developments in waste scholarship (following, for example, Alexander et al., 2013;Gille, 2007;Gregson et al., 2007;Hawkins, 2006;O'Brien, 2008), suggest food waste is constitutive of social ordering processes at a variety of scales. Sociological studies that take household and consumer food waste as their focus (e.g. Devaney and Davies, 2016;Metcalfe et al., 2013;Watson and Meah, 2013), in drawing attention to social and cultural contexts in which food waste occurs, offer an important corrective to political manoeuvres that seek to individualise responsibilities for waste reduction. These accounts, however, do not focus on the nature and dynamics of these manoeuvres themselves. This article, by contrast, focuses on the processes through which responsibilities come to be apportioned around food waste by a range of stakeholders. Debates around the responsibilisation of the consumer are well established in the fields of ethical and sustainable consumption (e.g. Barnett et al., 2011;Maniates, 2001). Thus far, however, with the exception of Evans (2014) and Evans et al., (2017), responsibilisation has not been addressed in the specific context of food waste. Our starting point for this research was, therefore, to explore Evans' (2014) speculative account of the mechanisms through which the consumer is responsiblised in the framing of the food waste issue. Analysis of our empirical material, however, provides scant evidence in support of the individualisation of consumer responsibility for household food waste; rather, it demonstrates a dominant understanding within this discourse coalition of responsibility as distributed throughout the productionconsumption system.
Our principal aim is to address how collective actors have come to frame the issue of food waste. With that in place, we seek to inform the question as to why retailers have adopted a framing of the issue as distributed responsibility rather than of the responsibilisation of the consumer. The article proceeds as follows. In the next section, we analyse the issue of food waste reduction at the interface of households and retailers. In the following section, we unpack the 'responsibilisation of the consumer' in the context of sustainable consumption and the issue of food waste. We then move on to examine the food waste 'discourse coalition' (Hajer, 1995) and the conditions which have enabled its emergence, through its shared storylines, the relationships between the actors that compose it and practices which reproduce it. We then turn to a discussion of discursive hegemony and retailers' embrace of the framing of distributed responsibility and conclude.

Retailers, consumers and household food waste reduction
In the United Kingdom, households and consumers make the largest contribution to overall volumes of waste in the food chain from production through distribution to consumption (FAO, 2011;IME, 2013;WRAP, 2008). It is perhaps unsurprising then when food waste is positioned as an 'end of pipe' problem, with responsibility for reduction primarily placed on consumers (Alexander et al., 2013).
Equally, however, retailers are routinely identified in media reports and elsewhere as bearing responsibility for food waste (e.g. HoL, 2014). Retailers are directly responsible for just 5% of food waste (HoL, 2014). However, a central bone of contention in debates around food waste reduction concerns the extent to which retailer practices are indirectly responsible, whether 'downstream' by the way in which food is marketed, 'displac[ing] waste from the supply chain into the individual consumer's fridge' (Bowman et al., 2014, p. 58), or 'upstream' in the supply chain through policies towards suppliers.
Reducing food waste in the supply chain and in store has the obvious potential of affording efficiency gains to retailers. Since the 1990s, when the European Commission began to promote the 'ecological modernisation' of production, the minimisation of waste, increases in resource efficiency and the maximisation of output have been seen to go hand in hand, in both public and corporate policy (Murphy, 2001). However, the interest of retailers in pursuing food waste reduction in their operations is not so easily explained away as simply self-interest in efficiency. Waste is built into the grocery production-consumption system in a systemic sense. Overly prescriptive aesthetic standards of fruit and vegetables are routinely cited as a frivolous cause of food waste (e.g. IME, 2013); a case of the 'dynamic relationship between retail management practices and consumer demands and influence' being central to the generation of waste (Bond et al., 2013: 11). However, an even more problematic aspect of that dynamic is consumer demand for the guaranteed availability of food stuffs -materialised in constantly restocked supermarket shelves -and short term demand fluctuations 'dependent on vagaries of weather, season, calendar and consumer trends' (Bowman et al., 2014: 66). Forecasting product volumes is highly problematic for retailers, leading to supply contracts that incentivise farmers to over-produce (Bowman et al., 2014;WRAP, 2010). Bond et al. (2013: 14) conclude that the underlying principle of retail supply chains is high throughput with waste accepted as a by-product. This suggests some uncomfortable home truths for retailers engaging with the issue of food waste reduction, rather than easy wins in corporate public relations.
Retailers' more recent focus on reducing household food waste -in some instances arguably encouraging consumers to buy less of their own productsdemands yet more of an explanation. WRAP has been central to driving both the wider food waste agenda and that of reducing household food waste specifically. WRAP has worked closely with the retail sector, and versions of its 'Love Food Hate Waste' (LFHW) social marketing campaign, launched in 2007, have been widely promoted by UK supermarkets. The Courtauld Commitment, a voluntary agreement administered through WRAP and aimed at improving resource efficiency and environmental impact in the UK grocery sector, announced in 2013 a target to reduce household food waste by 5% by 2015, building on previous commitments to reduce food waste in the supply chain.
Retailers have engaged in a number of initiatives in response. For example, technical innovations have been made in packaging to increase product longevity in the home and there have been innovations in labelling, such as simplifications of date labelling and information on storage and portion size. It is notable that more visible initiatives demonstrate retailers' recognition that marketing practice can lead to household food waste. These include changes to promotions criticised as encouraging over-purchase, such as 'Buy One Get One Free' offers, as well as product innovations designed to reduce over-purchase, such as packaged mixed fresh vegetables. Some of the retailer engagement in the area, such as providing consumers with information about recipes, cooking skills and digital tools such as online weekly meal planners, reflects a wider trend of corporate interventions into consumer practice under the auspices of sustainable consumption. There are a growing number of examples of corporate sustainability initiatives identifying the 'use' phase of products' 'life cycles' as 'hot spots' of resource use or greenhouse gas emissions. Proctor and Gamble, for example, have encouraged consumers to reduce the temperature at which they wash their clothes (see Mylan, 2016). What marks household food waste out as different from other interventions in consumer practice is that the food waste 'hot spot' is wastage of the already purchased product, the value of which has already been realised by the retailer. At first glance, then, these interventions appear contrary to the imperatives of capital accumulation.

Scope and methods
This article draws on empirical research conducted in the United Kingdom in 2014-2015. Institutional mapping was conducted to establish key actors in the framing of the issue of food waste at the retailer-consumer interface. Interviews were conducted with 38 key informants from grocery retailers, third sector organisations and campaign groups, sustainability consultancies, government and the civil service. Respondents were asked questions about how they and their organisations became involved in food waste reduction activities, their understandings of the development of the issue and relationships between stakeholders over time, initiatives engaged in and future plans, and their understanding of the roles and responsibilities of different actors, with a particular focus on households and 'the consumer'. Observation was conducted at three commercially organised conferences in which (largely non-academic) actors in the discourse coalition exchanged knowledge, understandings and experience around household food waste. Two 'multi-stakeholder' events were in addition organised by the authors: a workshop which brought together 42 actors from the stakeholder groups noted above as well as academics working in the area, and a project exhibition and panel discussion with 75 invited delegates. In addition, extensive documentary research was conducted into grey literature around household food waste reduction, such as policy and consultancy reports, as well as NGO campaign materials and corporate communications, such as company sustainability reports and consumer-facing publications.
Discursive analysis was conducted first through hand coding interview material. Open coding was followed by axial coding, through which coded segments were iteratively related to one another (Gla¨ser and Laudel, 2013). By linking frequently recurring codes across the different interviews, dominant storylines were established. Storylines were also interpreted in the light of observational and documentary data. While obviously there were nuances of emphasis, the primary components of storylines were present in some form across all interviews.
Relationships between actors and core practices were established through interviews and observational and documentary data. Simple network diagrams were developed of key knowledge exchange relationships within the coalition and analyses conducted of principle mediating actors within these networks (e.g. WRAP, the British Retail Consortium (BRC)). Core practices were further explored through documentary data (e.g. corporate sustainability reports, consultancy reports, technical literature).
The article addresses the state of play of the discourse coalition at the time the empirical research was conducted (2014)(2015). This snapshot is located, however, in the context of the period following WRAP's seminal The Food We Waste report in 2008. 1 Elsewhere, we have addressed the changing dynamics of UK food waste discourse from 2007-2015 in more depth (Evans et al., 2017).

Sustainable consumption, the responsibilisation of the consumer and food waste
Over the course of the 1990s, the conventional attribution of responsibility for environmental sustainability to producers was increasingly supplemented by a focus on the role of consumers. Individual 'behaviour change' became a dominant framing in both public policy and corporate engagement around issues of environmental sustainability (Barr and Prillwitz, 2014;Evans et al., 2013;Shove, 2010). Furthermore, this process of attribution of environmental problems to consumption and 'lifestyles' was concomitant with the increasing prominence of the figure of the 'sovereign consumer' in social, political and economic life (Slater and Tonkiss, 2001). In its more naı¨ve, or neoliberal, forms, the project of sustainable consumption came to posit agency for change with a responsibilised form of the sovereign consumer. The aggregate choices of consumers were seen as the source of 'market signals' of a more 'responsible' demand exerting influence on the supply side. Critical responses to this trend argue that responsibilising the consumer tends to reduce systemic issues to de-politicised, individualised, behavioural choices in the market place and separate out consumption choices from the framework of structural constraints that is their context (Barr and Prillwitz, 2014;Maniates, 2001;Rumpala, 2011). This framing of sustainable consumption fundamentally underestimates the extent to which individuals' autonomous action is constrained by infrastructures and socio-technical systems; access to economic, social and cultural resources; and the collective and normative derivation of most consumption (Southerton et al., 2004). Politically and ethically, it offers a 'constrained space of possibilities' defined as aggregate effects of individualised choices (Rumpala, 2011: 699).
Our entry point is Evans' (2014) speculation that there is evidence for the responsibilisation of the consumer in the framing of the food waste issue. He cites, for example, the IME's Global Food-Waste Not, Want Not report (2013) as blaming household food waste on consumers' excessive purchasing, demand for aesthetic perfection, poor understanding of 'use by' labelling and a 'culture with little understanding of the source and value of food' (IME, 2013: 27 in Evans, 2014. Similarly, the UN FAO, notes Evans (2014), 'is quite explicit that affluence, consumer attitudes and a lack of awareness are to blame', while the European Commission offers 'lack of shopping planning, confusion about . . . date labels, lack of knowledge on how to cook with leftovers' as the top of its list of causes for household food waste (p. 23). Evans thus finds the discourse of the responsibilised consumer enacted by multiple collective actors in relation to food waste, including international institutions (the UN FAO, the European Commission), nongovernmental organisations (NGOs) (WRAP, WWF), media and retailers (Evans, 2014). We similarly note the rhetoric of aggregated 'market signals' of 'responsible' demand in (at least some of) the promotional literature of the grocery sector's voluntary waste reduction initiative, the Courtauld Commitment (WRAP, n.d.). However, having empirically investigated this speculation, we suggest that Evans' (2014) account requires nuancing, in order to better understand the discursive positioning of consumer responsibility.
Our research suggests that while in food waste discourse we do find responsibility being apportioned to the consumer, the responsibilisation of the consumer is by no means, or no longer, the dominant framing of the food waste problem (see also Evans et al., 2017). We note in this context retailer initiatives addressing marketing practices. Where consumers are addressed, we witness not so much an individualisation of responsibility than an engagement with household practices. Interview and observational data demonstrate a co-production by various actors of the problem of food waste, framed as an issue of distributed responsibility. This framing is central to shared understandings and narratives through which the key actors around the issue of household food waste in the United Kingdom define the problem and potential solutions to it. In the following sections, we unpack the discursive and practical dynamics of this coalition and explore the conditions that enabled its emergence.

The food waste discourse coalition
Processes through which corporations come to address societal problems have been widely studied in organisational, innovation, transition and social movement studies (Davis et al., 2005;Elzen et al., 2004;McAdam et al., 1996;Schneiberg and Lounsbury, 2008). Generally, corporate responses to social and environmental issues have proceeded first through reactions to external, coercive pressures, primarily from campaign groups challenging corporate reputations in the public sphere and forcing regulatory change. In their classic study of institutional isomorphism, DiMaggio and Powell (1983) distinguished between three processes or mechanisms -coercive, mimetic, normative -that militate towards the uniformity of organisational forms, and by extension, we suggest, the objects and organisation of corporate intervention. Contrasted with coercive processes, mimetic isomorphism arises primarily from uncertainty, in which organisations imitate perceived successful innovations (DiMaggio and Powell, 1983). Subsequent work in the institutionalist tradition has stressed how purposive normative projects have sought to change the definition of responsibilities of actors within organisational fields, again usually following successful coercive pressures such as the contentions of social movements (Davis et al., 2005;Schneiberg and Lounsbury, 2008). While there may be complex, multi-stage dialectical processes between the build-up of issue-related pressures and corporate responses and reorientations to those issues (see Geels and Penna, 2015), almost universally, coercive pressures from social and political mobilisation precede while normative and mimetic processes follow.
Strikingly this is not the case in retailer responses to the food waste issue. The key actors concerned with food waste -retailers, policymakers, consultants, third sector organisation and NGOs -have largely not exhibited agonistic relations. WRAP's aforementioned Courtauld Commitment boasted 92% of the United Kingdom's supermarkets as signatories during its initial phase in 2005 (WRAP, n.d.). Thus, the interventions into the public sphere cited at the start of this article occurred during, and some as part of, the Courtauld Commitment (which is ongoing and now in its third phase). The coercive pressure of campaign groups has been largely absent in the dynamics through which the issue has gained its current prominence. It is also important to note that the UK government has eschewed regulatory mechanisms -another important form of coercive pressure -to address issues of food waste (Department of Food and Rural Affairs (DEFRA), 2013). 2 Asked to reflect on where the pressure for action on food waste had come from, key informants from across the actor groups concurred that, while stimulated by media attention, rather than being driven by one constituency, there has been a collaborative process: 'Lots of other organisations, UN FAO, UNEP . . . came to the same point at roughly the same time' leading to a 'sort of critical mass' (interview 6, NGO) around the issue. One retailer representative put it thus: I don't think it was pressure. It was identified as an issue, which key stakeholders cared about . . . So it wasn't pressure as such . . . it was a realisation that it was an issue. (Interview 7,Retailer) We argue that a food waste 'discourse coalition' (Hajer, 1995) has developed through interaction between collective actors invested with the issue. These actors -grocery retailers, consultancies, third sector organisations, campaign groups and policymakers -are related to one another in multiple ways, giving rise to a dense interconnected network, which we discuss below. Negotiated definitions of problems and mutual constructions of common understandings have developed between the actors around the issue of food waste.
Following Hajer (1995), a discourse coalition can be defined as a set of storylines, the actors who utter those storylines and the practices in which this discursive activity is based. In the following sections, we address each in turn, beginning with our key empirical finding, the framing of the problem of food waste as one of distributed responsibility.

Storylines
The storylines a discourse coalition draws upon represent 'condensed statements summarising complex narratives' (Hajer, 2006: 69) through which 'elements from many different domains are combined' (Hajer, 1995: 62). Storylines, as narrative structures, even in their simplest form, offer causal links that provide plausible reasons why outcomes should be expected and therefore how problems are framed and responded to (Czarniawska, 1997).Two mutually supporting storylines form the discursive basis of the coalition. As our central argument suggests, the first of these, by contrast with the responsibilisation of the consumer, posits responsibility for food waste reduction as distributed throughout the food production-consumption system, due to the complex linkages within it. This can be summarised as responsibility is distributed because 'where food is wasted is not necessarily where the causes and the drivers are' (Interview 10, NGO). While nuances obviously varied, this storyline was shared by the different actor groups and was a central organising feature of their engagement with one another. Collective actors within the discourse coalition understood food waste as arising from, and in the context of, interactions and linkages within a system of provision encompassing production, distribution and consumption: I think we would definitely see it as a system-wide issue . . . if there is a focus for our interest, it's more around what's the interaction between those different stages in the system. (Interview 1,policy) Representatives of the retail sector echoed similar sentiments: I think we have to accept that we all have some responsibility . . . where there are problems is when there's a break between one part of the food chain and another. (Interview 9,retailer) Responsibility for resolving the issue was therefore understood as distributed throughout the system, rather than simply being attributable to the activity of particular actors, whether consumers or otherwise: It's not the consumer's fault that there's food waste and it's not the retailer's fault. It's a bit of both, and looking at it through one lens is unhelpful. It's absolutely a multistakeholder issue. (Interview 15,NGO) While responsibility is understood as distributed, retailers, however, were understood to have a key role to play, as 'they sit at a pivotal point [and] have an upstream and a downstream influence' (Interview 12, NGO). Again, representatives of the retail sector appeared to acknowledge and assume these responsibilities: When you look at what are the causes of waste on farms, some of the things, it will be very little to do with [our] operations . . . But in other cases it can be down to, for example, forecasting ordering. Now, that is clearly a shared responsibility . . . Taking the whole value chain perspective of the issue is very, very important to ensure that we're not passing the waste problem down the value chain. (Interview 7, retailer) The second, mutually supporting storyline is more complex but nevertheless amounts to a normative consensus that 'food waste is morally repugnant' (Interview 14, policy). Food waste, as one of the policymakers we interviewed put it, brings to bear 'a perfect storm of issues' (Interview 24, policy) -climate change, depleting water resources and a growing global population, food security and poverty -that can only be addressed in systemic fashion. Austerity, food price inflation, the rise of food banks, concerns about food security, environmental sustainability and climate change have come together in the 'informational politics' (Clarke et al., 2007) of food waste. At the same time, the storyline continues, people have become culturally alienated from the food system. As an influential consultant put it (Interview 4), 'People are beginning to realise just how alienated society's become from the sources of food and any concept of where it comes from and what it takes to produce it'. Furthermore, despite recent food price inflation, the price of food remains cheap in historical terms: Retail has been very good at reducing the price of food. And it's driven the cost down all the way through to a level where it doesn't matter if you waste food because it hasn't got that much value. (Interview 24,policy) Together, this alienation and low costs have led to a cultural de-valuing of food. We can thus summarise the storyline as, in the face of the 'perfect storm' of issues, food waste is morally repugnant; at the same time, culturally we have forgotten the value of food. The components of the 'perfect storm' -from climate change, to food security, to food price inflation and post-crash poverty in the Global Northhave acted as an enabling discursive context in which food waste as an issue has been brought into focus as a problem in the public sphere. Within this context, however, the specificity of the existing normative, cultural repertoire of food waste is notable. Food waste commonly evokes a visceral normative reaction (Evans, 2014), which we observed among our informants. Thus, in response to questioning the company's motivations to engage with the issue, one retailer explained, '. . . our colleagues care about the issue, our suppliers cared about the issue, and our customers, through our own research, clearly care about the issue' (Interview 7, retailer).
In the following sections, we turn to the other components of Hajer's (1995Hajer's ( , 2006 definition of discourse coalition -the actors that share the storylines and the practices which form the context for their production and reproduction -and explore the practical enabling conditions of its emergence.

Collective actors
The food waste discourse coalition is instantiated through the practices of a dense network of mutually inter-related actors: retailers, sustainability consultancies, third sector organisations, campaign groups, civil society organisations and policy actors, as well as, to a degree, academics. The presence of major retailers in this network requires further consideration. Notably, one NGO respondent offered a useful counterfactual to the emergence of the food waste coalition: If you imagine a situation where significant retailers were against [the food waste] agenda, I think it would have been much harder to get some of those [food waste reduction] messages out and get consumers to act. So I think actually it's interesting to consider a situation in which retailers have done nothing or have been actively hostile in this agenda because I think it might have played out quite differently. (Interview 23,NGO) This lack of a counter-constituency is critical to the emergence of discursive hegemony over the field of possible contention. The indirect responsibility of retailers for food waste in their supply chains could be easily articulated by them in terms of consumer demand. Moreover, the very prevalence of discursive constructions of the responsibilisation of the consumer demonstrates that retailers had at their disposal an easily mobilisable defence against active engagement with household food waste reduction. Responsibilising the consumer would, for retailers, avoid an uncomfortable focus on issues of overconsumption, promotion and marketing, and shift the issue of household food waste purely onto the familiar, individualising territory of 'behaviour change'. The storyline of distributed responsibility is by no means a self-evident occurrence.
Beyond the specific point about the active involvement of retailers in the food waste discourse coalition, we note that many of our respondents have worked with one another on multi-stakeholder initiatives for waste reduction in a variety of configurations. Even when respondents had not collaborated directly, they were connected to one another via intermediaries.
Importantly, knowledge, understandings and evidence were shared freely through this network, even between competitors, enabling the mutual construction and reproduction of the shared understandings of the discourse coalition. Consultancies played an important role here as conduits of knowledge exchange -such as between academic insights and commercial practice -and brokerage between different actor groups -such as between policymakers and retailers. Consultants in the field included both for-profit and not-for-profit consultancy organisations that specialise in sustainability issues, which often developed detailed domain-specific knowledges around food waste, as well as more generic areas such as 'behaviour change' or 'sustainability communications'.
We could cite many examples of the role of consultants in knowledge brokerage. The example of the DEFRA funding of research on food waste is illustrative. DEFRA engaged with two consultancies around this project, funding the production of two evidence reviews, the first of which involved civil servants working with one consultancy, the other with both consultancies collaborating. These evidence reviews were widely circulated among coalition actors. As part of the same programme, DEFRA funded one of the consultancies to develop a consumer behaviour change programme around food waste and shopping that operated through the 'consumer panel' of one of the major retailers. A pilot project was then managed by the retailers' sustainability manager and the consultancy was engaged to write up results for DEFRA.
Also central to the development of the coalition was the role of trusted intermediary organisations, which act as vehicles for interaction and neutral forums. There was unanimity among our key informants that WRAP has played a critical role as one such, both in driving the public profile of the food waste issue and in the process of coalition building. As well as administering the Courtauld Commitment, around and through which many coalition activities took place, WRAP's consumer-focused campaign 'Love Food Hate Waste', for example, launched a new initiative across 10 UK cities in collaboration with the major supermarkets and third sector organisations (LFHW, 2014). The British Retail Consortium (BRC), a trade association, and IGD (Institute of Grocery Distribution), a research and training charity for the grocery sector, also played an important role. IGD, for example, ran a food waste media campaign supported by all the major retailers, while the BRC has brokered agreements on collective policy positions on food waste (e.g. BRC, 2015).
Intermediary organisations' importance in coalition building was stressed by retailer representatives and other actor groups. Intermediaries are of particular importance for the relationships between retailers specifically for two key reasons. First, retailers are of course competitors and work in the area of food waste, whether in terms of proprietary techniques and data or in terms of brand positioning, is not immune to competitive dynamics. Second, laws designed to prohibit price setting and cartel activities complicate and sometimes prohibit 'precompetitive' collaboration on certain activities. Intermediaries that act as 'honest brokers' are therefore crucial for retailers navigating this complex territory.

Practices
The practices within which this discursive activity took place were, first and foremost, those which actively constructed the network: attending, speaking and networking at multi-stakeholder events, including commercially organised conferences, and non-commercial roundtables and workshops organised by coalition actors. Second, were technical practices for mapping, quantifying and analysing food waste, such as product life-cycle analyses and other techniques related to the supply chain, and third, marketing and communicational practices (including social marketing addressed to consumer audiences, employee or supplier engagement activities and sustainability reporting). These practices embody and carry the dominant storylines of the coalition. Thus, retailer marketing practices have embodied the understanding of distributed responsibility that retailers can pass 'the food waste problem down the value chain' (Interview 7, retailer), evidenced by changes to promotions and new product types (such as mixed cut vegetables). Product life-cycle analyses that identify 'hot spots' of waste have also been marshalled to support changes to ordering practices that otherwise shift the burden of surplus onto the producer.
All three sets of practices share a common understanding through the 'waste hierarchy'. 3 The waste hierarchy is an order of preference of waste treatment options that aims to reduce environmental impacts by prioritising prevention, reuse, recycling and recovery over landfill. It has enjoyed wide spread support in most developed countries as a guide for waste management and was introduced into EU legislation through the 2008 Waste Framework Directive (EC, 2008;cf. DEFRA, 2011), which required member states to develop their own frameworks.
The waste hierarchy, then, can be viewed as a central organising, normative device of the discursive hegemony of the coalition. By 'device' here, we mean a method or heuristic that both organises and is carried by practice. The centrality of the waste hierarchy is evidenced by the fact that many respondents from the retail sector were quick to invoke its principles in accounting for their activity around food waste reduction. For example, Reducing waste in the first place. Absolutely. You know, even the redistribution charities will say the better outcome is no waste in the first place. They'd go out of business, but if I was to set a target and I wanted to get some headlines, I'd say my target is to give no food to charity because it's really about using it in the first place and I think actually everyone accepts that. There's no challenge to that really. (Interview 9, retailer) Hultman and Corvellec (2012) argue that EU policy on the waste hierarchy 'challenges dichotomous understandings of economy and society on the one hand, and environment and nature on the other' (p. 2414). According to Hultman and Corvellec (2012), policy around the waste hierarchy enacts a double movement between 'blackboxing' economy and environment as ontologically distinct, unproblematic categories, and problematizing this fundamental dichotomy, positing environment and economy as co-constitutive. As such we suggest the waste hierarchy represents a 'strategic ambiguity' enabling actors with potentially antagonistic discursive positions, such as environmental groups pursuing an ecological rationality and businesses pursuing an economic rationality, to enter into coalition (Wexler, 2009). Hajer (2006) notes that misunderstanding, as much as mutual understanding, can be highly functional for the creation of discourse coalition (see also Stark, 2009). We see here that the very ambiguity of this foundational understanding of the coalition accommodates otherwise antagonistic positions.

Discussion
The existence of a discourse coalition does not inevitably lead to discursive hegemony in a particular policy domain. However, our research does suggest that the food waste coalition has achieved discursive hegemony. Hajer (2006) offers 'a simple two-step procedure for measuring the influence of a discourse': the extent to which people use it to conceptualise a particular domain ('discourse structuration', in his terms) and the extent to which it becomes institutionalised in organisational practices (p. 70). When there is both structuration and institutionalisation, a discourse coalition becomes 'hegemonic in a given domain' (Hajer, 1995: 59). We suggest both criteria are fulfilled in the case of the food waste discourse coalition. The dominant storylines of the coalition were shared by all our key informants (a strong representative sample of actor groups), including both retailers and their potential antagonists in civil society and campaign groups. The storyline of distributed responsibility reflects negotiated definitions of problems and mutual constructions of common understandings. Furthermore, this central storyline, together with the central common understanding and heuristic device of the food waste hierarchy, are instantiated through a set of organisational, technical and communicational practices, or put another way, are institutionalised by retailers and third sector or civil society organisations (such as WRAP and Zero Waste Scotland).
It is necessarily beyond the scope of this article to answer the question why have retailers embraced household food waste reduction as an object of intervention, and done so without primarily framing the issue as one of consumer responsibility. However, the preceding analysis warrants and permits discussion of the enabling conditions of this unexpected discursive hegemony.
First, we note that the efforts of actors in the discourse coalition towards household food waste emerged against a backdrop in which wider issues of waste, foregrounded to households through recycling initiatives, had reached a relative level of maturity: in terms of normative consensus, infrastructure, popular understanding and household recycling practices. WRAP's work on recycling at the household and Local Authority level was again crucial in the United Kingdom. As more than one respondent remarked, conventions around household recycling have become ingrained, but 'we just hadn't thought about food' (Interview 12, NGO).
Second, we should acknowledge the maturity of the corporate sustainability agenda in the supermarket sector, with major retailers setting agendas and demonstrating high levels of consumer engagement around sustainability (Evans et al., 2018;Lehner, 2015;Oosterveer, 2012). The UK context furthermore offers examples of institutionalised multi-stakeholder cooperation, with the sector's engagement with long established collaborative fora such as the Ethical Trading Initiative (founded 1998) and the Marine Stewardship Council (founded 1996), that bring companies and campaign groups together. We stress again in this context the important role of trusted intermediaries, such as WRAP, and of specialist sustainability consultancies as knowledge brokers, which facilitate the exchange of understandings among actors and the co-production of distributed responsibility as a framing of the issue.
Third, it can be argued that the oligopolistic tendency of the UK grocery market (Harvey, 2007) militates towards the rapid normative and mimetic transfer of issues and understandings through densely woven stakeholder networks, even in the face of the intense competition characteristic of the retail sector. Schurman and Munro (2009) demonstrate this effect in a comparative case study of anti-genetically modified (GM) food movements in the UK and US contexts. They argue that the 'economic opportunity structure ' (p. 193) of retail market organisation in the United Kingdom militated towards the success of the anti-GM campaign.
We suggested above that the lack of a counter-constituency was critical for the emergence of discursive hegemony. We stress again that there is nothing inevitable in retailers coming out strongly in support of the issue of food waste reduction. To reiterate, first, food waste is built into grocery supply chains in a systemic sense (Bond et al., 2013), and second, retailers have marginal direct 'in store' responsibility for food waste -about 5% across the whole production-consumption system (HoL, 2014). Moreover, as for active engagement in household food waste reduction, this appears against retailers' immediate financial self-interest. However, we can speculate a recursive relationship between the process of the formation of the discourse coalition and the cognitive framing of the issue on the part of retailer representatives. Hajer (1995) argues storylines are more probable to secure discourse hegemony if they are both cognitively acceptable for their proponents and other potential partners in coalition and provide proponents with some strategic advantage within a given domain. With respect to the first condition, we have elsewhere (Evans et al., 2017) noted a shifting terrain of sustainable food politics over the period in which the coalition developed (roughly 2008-2014), moving away from an individualisation of responsibility. Furthermore, distributed responsibility reflects a state of the art of commercial research in the field (itself informed by academic problematizations of consumer responsibility). The retailer representatives we interviewed are individual mediators between the storylines and understandings of the discourse coalition and retailers as strategic collective actors. Further research would be needed to uncover organisational decision-making processes to fully explain retailers' orientation with the discursive hegemony, but we can suggest two senses in which the storyline of distributed responsibility is 'positionally acceptable' (Hajer, 1995) to retailer's strategic orientations.
We see no imperative to unmask the supposed real economic motive of retailers' engagement with the issue, as calculative agents in the market. Retailers are socially embedded actors operating in the context of the normative expectations of the contemporary discourses of sustainability and social responsibility, however imperfectly. Equally, this is a strategic context within which distributed responsibility does resolve the apportioning of responsibility for food waste in an acceptable manner for retailers, where otherwise they might be criticised, given their critical role in exerting influence upstream to producers and downstream to consumers (Harvey, 2007;Oosterveer, 2012). More crucially, we suggest retailers' engagement with household food waste should be seen in the context of brand management (see Lehner and Halliday, 2014). As Dixon (2007) points out, supermarkets seek to establish themselves as trusted lifestyle authorities and create positive brand connotations by proffering solutions to customers' everyday problems. Household food waste offers a site in which retailer brands may offer productive engagement in prudent household management.
The preceding analysis suggests that retailers have been offered a fertile social context in which to intervene. Food waste can be seen as a medium through which to engage consumers in otherwise often complex and distant sustainability agendas (see Lehner, 2015). Moreover, it is a medium -unlike other sustainability issues, such as household energy use -that offers visceral and affective engagement often lacking in the subject matter of corporate sustainability initiatives. The uncontroversial, proximate, normative relation to wasting food makes the issue unusual in the opportunities it affords retailers for affective engagement around sustainability with their brands. In contemporary brand management, the consumer is envisaged as the active co-producer of brand value (Vargo and Lusch 2004;Zwick et al., 2008). Viewed as such, household provisioning practices become the productive site for consumer engagement.

Conclusion
To conclude, we note the suggestion made by Evans et al. (2013) that the emergence of political and cultural interest in food waste might be generative of new political and ethical possibilities. In light of the preceding analysis, it seems credible that taking notice of waste and acknowledging our interconnections with it (Hawkins, 2006) call forth new ways of governing sustainable consumption. Our entry point for this article was Evans' (2014) findings that, in an earlier phase, the food waste issue was framed through the lens of the responsibilised consumer. We have argued, by contrast, that the current, discursively hegemonic, framing of the issue is one of distributed responsibility. However, rather than simply contradicting Evans' (2014) position, following Barnett et al. (2011), we acknowledge that the figure of the responsibilised consumer has been mobilised in the process of driving the issue of food waste up the agenda. We see 'the consumer' here, however, not so much as a locus of sovereignty and agency but as a rhetorical figure and point of identification (Barnett et al., 2011). The consumer in no way disappears from the picture -we have highlighted retailers' attempts to intervene in household practice, for example -but this does not amount to a programmatic attempt to individualise responsibility (for further discussion of this point, see Evans et al., 2017). The figure of the responsibilised consumer has not been deployed by retailers to displace their responsibility as central intermediaries in the food production-consumption system -despite the opportunities to do so afforded by both the existing normative repertoire around wasting food and discourse of individualised consumer responsibility.
Optimistically, the discursive hegemony of distributed responsibility may suggest signs of the exhaustion of the responsibilised consumer as a frame for corporate and public policy on sustainable consumption. At the very least, it cautions that critical discussions of sustainable consumption should not take for granted the responsibilisation of the consumer across different empirical domains. Rather there is a need to focus empirically on processes of responsibilisation in different production-consumption systems, and to address the variegated dynamics of contention around different aspects of sustainable consumption.
Finally, whether the discourse coalition around food waste in the United Kingdom is an enduring phenomenon, or exhibits characteristics transferable across political and economic contexts, remains to be seen. There is nothing inevitable about the configuration that characterises the current situation in the United Kingdom. We note, for example, that the waste hierarchy, which we argue is an enabling device of the discourse coalition (and which is central to EU waste policy), is not un-contestable on ecological grounds. Van Ewijk and Stegemann (2014), for example, conclude that the waste hierarchy is an insufficient foundation for waste and resource policy to achieve absolute reductions in material throughput in the economy. We might ask then, with Gille (2007), what is excluded from the representation, conceptualisation and enactment of food waste within the current coalition? It is these exclusions that have the potential to destabilise the observed discursive hegemony and open the issue of food waste into a wider field of political contention around sustainable consumption.
3. The 'waste hierarchy' can be theorised as a 'general understanding', a component of practices, in Schatzki's (2002) terms: that is an understanding common to multiple practices that is expressed through their performance. This is a quite distinct sense from that of Hajer's (1995) concept of 'storyline'.