Fine-Tuning the Fight Against Food Waste

The complex causes of consumer food waste make it difficult for commercial actors and public policy makers to develop successful foodwaste reduction campaigns. One of the essential problems is that consumer food waste seems to be the unplanned result of divergent food-related behaviors. The current research investigates the relationship between distinctive consumer food-related lifestyle patterns and food waste. A survey with 848 consumers in a Northern European country (Denmark) suggests that segments of consumers identified by food-related behaviors have corresponding differences in food waste produced. For example, consumers’ food waste varies across different patterns of food-related lifestyle-dimensions, such as 1) cooking enjoyment, 2) food planning, 3) price orientation, 4) social relationships related to meals, and 5) food-safety concerns. The study presents possible macromarketing actions and policies targeting consumer segments to reduce food waste.


Introduction
To successfully and sustainably secure food for 9 billion people, both scholars and societal stakeholders have recommended actions aimed at avoiding food loss and food waste (FAO 2013;Foley et al. 2011;Godfray et al. 2010). Food waste can be defined as "any food, and inedible parts of food, removed from the food supply chain to be recovered or disposed (including composted, crops ploughed in/not harvested, anaerobic digestion, bio-energy production, co-generation, incineration, disposal to sewer, landfill or discarded to sea)" (FUSIONS 2014, p. 6). It thus includes all food that is discarded, even though it is appropriate for human consumption (FAO 2013). Exact numbers on food waste vary, but multiple sources suggest that about one-fourth to one-third of human food production is being lost or wasted along the food supply chain and in consumer households (FAO 2013;Kummu et al. 2012;Brautigam, Jorissen, and Priefer 2014). Food losses in the supply chain occur predominantly in developing countries, whereas food waste in consumer households occurs primarily in developed countries (Parfitt, Barthel, and Macnaughton 2010). For example, between 25% and 40% of food in the US -and even more in some single food categories (Love et al. 2015) -is lost or wasted (Cuéllar and Webber 2010;Venkat 2011). This amount is equivalent to 10% (in value) of US household food purchases (Buzby and Hyman 2012).
Food waste is problematic from a food security (Garnett 2011) and societal equity (Gjerres and Gaiani 2013) perspective and also contributes to the worrisome exploitation of our natural resource base. Agriculture, livestock and human food production have been identified as some of the most important domains regarding natural resource use and environmental degradation (Rockström et al. 2009;EC 2006). The food and agriculture sectors contribute to greenhouse gas emissions through land use change, methane emissions, water usage, and fertilizer application (Aleksandrowicz et al. 2016;Carlson et al. 2017;Lamb et al. 2016). In addition, three key systems identified as beyond safe boundaries -biodiversity, climate change, and the phosphorous and nitrogen cycle (Rockström et al. 2009) -are directly impacted by the agriculture and food sector.
Actions must be taken to reduce the large environmental impact of agriculture and to ensure food security. Recommended actions include: 1) technological mitigation, 2) the diversification and sustainable intensification of agriculture, 3) dietary shifts towards plant-based diets, and 4) avoidance of food losses and food waste (Aleksandrowicz et al. 2016;Foley et al. 2011;Garnett 2011;Reisch, Eberle, and Lorek 2013). It is difficult to identify the exact role of food waste in this process, but a study suggests that about 16% of the EU food sector's contribution to global warming (in 2011, CO 2 equivalents) is caused by food waste (Fusions 2015). Thus, food waste is a problem that hinders the agricultural and food sector's ability to sustainably feed the growing population without depleting the resources that future generations depend on while also impairing sustainable development in social, economic, and environmental capacities.
Food waste occurs along the food supply chain and is the result of multiple actors, such as institutions, supply chain actors, and consumers (Stuart 2009). Yet, in developed countries, the essential contributor to the food waste production appears to be the consumer (e.g., Buzby and Hyman 2012;Beretta et al. 2013). In the EU, consumer households account for approximately 40-45% of the food waste (EC 2010;Fusions 2015), or on average 76 kg per capita per year (EC 2010). This means that consumers waste approximately 10-25% of the food that they purchase, for example, 10-14% in Germany by value (Universität Stuttgart 2012) or 22% in the UK by weight (WRAP 2012). Therefore, altering consumer behavior is a crucial step in sufficiently 'scaling up' efforts towards sustainable development (Dietz et al. 2009;Prothero et al. 2011) and behavioral changes in the food area could prove significant (Macdiarmid et al. 2012;Reisch, Eberle, and Lorek 2013). According to Beverland (2014), sustainability of eating and diets has not yet received the macromarketers' attention that it deserves.
There are, however, at least two main issues with addressing consumer-level food waste (FUSIONS 2014(FUSIONS , 2015. Firstly, consumer-level food waste is influenced by multiple interacting factors (Quested et al. 2013;Stuart 2009). These factors range from macro-level environmental, political, economic, technological, and legal contexts, to societal-level consumer culture and social norms, to individual-level socio-demographic characteristics and psychographic mind-sets (Aschemann-Witzel et al. 2015;EC 2010;Evans 2014;FAO 2011). Consumer-level food waste, just as waste overall (e.g., Guillard and Roux 2014), appears to be an issue that emerges from the interaction between supply chains, marketing, and society (Stuart 2009). Therefore, potential solutions to address food waste should be approached with a broader macromarketing perspective (Dolan 2002). Yet, thus far, most interventions aimed at reducing consumer-level food waste have taken a narrower micromarketing perspective, focusing on details like plate size in restaurants (Kallbekken and Saelen 2013), using forced choices in schools (Lombardini and Lankoski 2013), or providing information on sustainable options (Panzone et al. 2011). Instead, we suggest that a macromarketing approach towards consumer-level food waste behaviors might provide new, valuable insights that can drive potential interventions aimed at reducing consumer-level food waste.
A second main problem of addressing consumer-level food waste is that food waste is not necessarily a consciously chosen behavior in which consumers engage. Instead, food waste seems to be an unwanted consequence of consumption prediction errors or of inadequate practices in consumer food-related behaviors such as planning, shopping, storing, preparing, and cooking (Evans 2012;Stefan et al. 2013;Terpstra et al. 2005). Especially qualitative research studies have highlighted that consumers report feelings of guilt when throwing away foods (Comber and Thieme 2013;Graham-Rowe, Jessop, and Sparks 2014;Parizeau, Massow, and Martin 2015;Redman and Redman 2014;Watson and Meah 2013), and that consumers are embarrassed and emotionally affected by wasting food, maybe even more than by other types of waste behaviors (Gjerres and Gaiani 2013). It is useful to move beyond consumer food-waste behaviors and to study consumers' lifestyles and consumption habits concerning foods and food waste as the antecedent of wastage in order to develop potential macromarketing interventions that would be able to successfully reduce consumerlevel food waste. Such an approach acknowledges that the sum of the complex and interacting factors impacting food waste are reflected in a consumer's lifestyle (Evans and Abrahamse 2009;Plummer 1974) and combines attitudinal and behavioral indicators of consumer action (Lavelle, Rau, and Fahy 2015).
From a macromarketing perspective, all types of waste can be understood as a consequence of consumerism and overconsumption (Gjerres and Gaiani 2013). Reduction of consumption has been highlighted as a crucial element of 'real' sustainable consumption (De Coverly et al. 2008) and is necessary to reduce consumption and (re-)insert the notion of sufficiency and frugality into consumer behavior. We suggest that food waste is a particularly well-suited issue to trigger consumer reflection on consumerism overall, given the guilt and emotion that the topic creates.
In sum, the current research addresses the issue of food waste from a macromarketing perspective by clarifying how segments of consumers differ in terms of lifestyle and food waste-related indicators. The goal is to provide a basis for refining food sector procedures, targeting food sector actions, and developing public policies aimed at decreasing food waste. We do so, firstly, by identifying consumer segments based on their food (waste)-related lifestyle; secondly, by characterizing these consumer segments based on indicators of consumer food waste (knowledge, attitudes, behavior); and thirdly, by deriving recommendations for targeted actions for both food marketing and food market policies that are in line with the developed consumer segments. In our view, the present study contributes to the emerging literature on food markets, marketing and food waste (Devin and Richards 2016;Gruber, Holweg, and Teller 2016) and on targeted prevention strategies (Delley and Brunner 2017;Love et al. 2015) by applying the framework of foodrelated lifestyle to food waste.

Consumer Lifestyles
Consumer lifestyles are understood as distinct ways of living for different groups within a given society (Ganglmair-Wooliscroft and Lawson 2010;Plummer 1974). These distinctions are assumed to be anchored in the individual's value prioritiesmore than class or economic resources. These priorities are reflected in both psychographic indicators and observable consumption practices and behavior (Evans and Abrahamse 2009;Bin and Dowlatabadi 2005;Lavelle, Rau, and Fahy 2015). In addition to the individual values, attitudes, and beliefs, consumer lifestyle is often seen as being impacted by cultural influences, societal trends, and technology development. These factors can affect household characteristics and social norms, and determine such external influences as marketing and media (Ganglmair-Wooliscroft and Lawson 2010;Lawson and Todd 2002). As such, lifestyles are multi-dimensional, depend on multiple interacting factors, and can potentially evolve over time and with the individual's circumstances.
Lifestyles are related to consumer identity, given that consumers with similarities in lifestyle will also have similar mental representations of how a consumer of a certain product 'thinks, feels and does' (Reed et al. 2012). Thus, consumers project similar identity-related associations on other individuals with whom they share a similar lifestyle (Reed et al. 2012). When studying cultural practices in consumption, researchers explore the expression of underlying concepts, such as identity, and observe how identity impacts consumer lifestyles (Arsel and Thompson 2011). However, a difference exists in that identity, a self-conceptualization process, (Rosenfeld and Burrow 2017) is created more deliberately than lifestyle.

Factors Causing Consumer-Related Food Waste
Consumer perceiving 'sub-optimality' of a food item is a crucial factor for distinguishing which items are eaten or discarded (for a definition, see Aschemann-Witzel et al. 2015;De Hooge et al. 2017). Sub-optimality is related to date labelling (Wansink and Wright 2006), to visual imperfections such as shape, color, or size (Loebnitz, Schuitema, and Grunert 2015), and to minor packaging damages (White et al. 2016). In the current situation of over-supply in the food market, where retailers strive to avoid out-of-stock situations, use volume and sales promotions (Theotokis, Pramatari, and Tsiros 2012), and have a wide assortment range (Gruber, Holweg, and Teller 2016), many food products end up as 'unsellable' but not necessarily 'un-consumable'. Consumer perceptions of what is 'optimal' or not appear to have become narrower since developed countries left the times of food scarcity behind (Evans, Campbell, and Murcott 2013), not least of all due to anxiety regarding food safety (Watson and Meah 2013).
Even though consumers are increasingly aware of social and environmental consequences of their food purchases (e.g., growing interest in organic foods, local food, and ethical attributes, Willer and Lernoud 2016), consumers are relatively unaware of the high environmental impact of food productionespecially of meat (Dietz et al. 2009) -and of how the impact is aggravated when food is wasted (Williams et al. 2012). Consumers are not necessarily conscious about the extent to which food waste, in addition to the resources embedded in the food wasted, leads to assets spent on disposal and emissions from incineration and landfills (Bernstad Saraiva Schott and Andersson 2015). Often, the aspects of caring for the safety and preferences of the near family, which are important functions of food and meals, are given greater priority than consequences of food waste in day-to-day decision-making (Cappellini and Parsons 2012;Graham-Rowe, Jessop, and Sparks 2014). In this trade-off, consumers might waste food despite feeling uneasy and guilty about it (Evans 2014;Comber and Thieme 2013;Gjerres and Gaiani 2013).

Approaches to Promoting Sustainable Behavior in a Market Context
One solution to address the issue of food waste could be trying to encourage consumers to avoid food wastage via in-store behavior as well as at home by promoting a favorable 'social idea'. In the case of food waste, this social idea consists of accepting that a consumer's food decisions have wider societal implications. This implies perceiving oneself not only as a consumer, but as a consumer-citizen (Prothero et al. 2011). Promoting such a social idea is at the core of the definition of 'social marketing' as the application of commercial marketing practices to non-commercial aims (Andreasen 2002;Kotler and Zaltman 1971;McDermott, Stead, and Hastings 2005).
Next to influencing the consumer or citizen's individual behavior in a targeted manner, social marketing also considers the consumer's behavioral context and the extent to which decision makers 'upstream' supply chain considerations ought to be encouraged to alter the consumer's choice environment (Gordon et al. 2006;Grier and Bryant 2005;Hastings and Saren 2003;Stead, Hastings, and McDermott 2007). This is well in line with theoretical models of consumer behavior and behavior change that emphasize the role of internal, individual factors as well as social influences and the criticality of a consumer' context and environment (e.g. theory of planned behavior, Ajzen 2011, and social cognitive theory, Bandura 2001). It is also in line with literature suggesting that both the individual (Lavelle, Rau, and Fahy 2015) and the macro-environmental context is key for transformation towards sustainable consumption (Prothero et al. 2011;Sunstein and Reisch 2014;Thøgersen 2014). When commercial stakeholders support actions towards a favorable social idea such as food waste avoidance, this activity might be done in the scope of the concept of corporate social responsibility (CSR, Carroll and Shabana 2010). A refined targeting of social marketing or CSR-activities in food waste avoidance can improve the efficiency of these efforts. The present research aims at contributing to this.

Current Actions in Encouraging Consumers to Avoid Food Waste
Numerous campaigners and non-governmental organizations have drawn attention to the issue of food waste in the past five years in various countries with their information and social marketing campaigns (Bloom 2010;Juul 2016;Stuart 2009;WRAP 2016). International organizations and governmental authorities (such as the European Commission, United States Department of Agriculture and US Environmental Protection Agency, and the FAO) have commissioned research on the extent of the issue, on the factors causing food waste, and on ways to tackle the problem. The United Nations (UN 2015) have the goal to decrease global food waste at the retail and consumer level by 2030 with at least 50%.
Increasing societal attention to food waste has sparked market actors to become involved in food waste reduction activities that focus on capacity building, awareness, redistribution of food destined to be wasted, or innovations in re-use (Aschemann-Witzel et al. 2017a). For example, food production and retail companies and organizations are allying up for collaborative analysis that can encourage coordinated goals and action (FWRA 2016). Societal attention has also led retailers to spearhead alteration of retail practices (Aschemann-Witzel 2018), and to create global media attention for the problem of food waste in their campaigns. For example, the abolishment of multi-item offers (such as '2 for the price of 1') by one retailer in Denmark has triggered others to follow suit; a campaign promoting sub-optimal fruits and vegetables in France has sparked worldwide media coverage (Aschemann-Witzel et al. 2017a). The success of these retailer actions depends on consumer support in reacting favorably towards the retailer and the campaign, and in adopting the related purchase behaviors. The current research thus aims at a better understanding of distinct consumer group behaviors.

Food Waste and Consumer Lifestyle Research
Despite the many public policies that NGOs and food market stakeholders have undertaken to reduce food waste, it is unclear whether such actions actually reduce the food being wasted. These actions may simply be moving food waste disposal from the retailer to the consumer household (Aschemann-Witzel 2016; Devin and Richards 2016). It is imperative to understand food waste behavior as a lifestyles symptom of consumers in affluent societies, greatly marked by the cultural paradigm of consumerism (Assadourian 2010). Food consumer behavior is to be understood in relation to the consumer's motives and value orientation and in its environmental and social context (Evans 2014;Beverland 2014;Dolan 2002). Macromarketers and policy makers should take these interdependencies into account (Nyborg et al. 2016;Prothero et al. 2011;Thøgersen 2014). Exploring food waste behavior and its underlying factors with a more complex perspective of consumer lifestyles can contribute to shedding light on how food waste avoidance actions will interact with consumers in their individual contexts. According to Bin and Dowlatabadi (2005, p. 198), "the basic premise underlying consumer lifestyle research is that by understanding consumers we can design better public policies." Lorenzen (2012) even suggests that lifestyle change can be a deliberate process undertaken in response to a problem left under-addressed by current policies and practices. Thus, a consumer-lifestyle perspective can help to identify targeted actions and to align existing actions with consumer behaviors, so that social marketing or CSR activities of food sector stakeholders can be most effective.
Previous research with consumer lifestyle approaches have looked at sustainability (Poortinga and Darnton 2016) or sustainability-related issues such as energy or housing (Bin and Dowlatabadi 2005;Thøgersen 2017a;Wei et al. 2007) and food behavior (Brunsø and Grunert 1995;Buckley et al. 2005;Thøgersen 2017b). In the food area, the 'food-related lifestyle' concept has been broadly applied and cross-culturally tested (see e.g. Scholderer et al. 2004). These studies have demonstrated that consumer and citizen groups often differ in the extent to which they are interested in and find food important. Thus, consumer involvement with the topic seems to be a relevant distinction for foodrelated lifestyle. So far, looking at food waste from a foodrelated lifestyle angle is an under-researched area. A study in the UK (Mallinson, Russell, and Barker 2016) and another in Switzerland (Delley and Brunner 2017) have explored food waste behaviors via segmentation, but to the best of our knowledge, this is the first study combining lifestyle and food waste research.

Material and Methods
The research presented here was part of a larger cross-country project. The current study builds upon previously conducted research in form of a literature review, expert interviews (Aschemann-Witzel et al. 2015), focus groups, and case studies (Aschemann-Witzel et al. 2017a). Measuring actual consumer-level food waste is difficult and time-consuming, hence, such studies are available only for small population samples. However, to explore consumer lifestyles with the intention to identify consumer segments, a large dataset is needed. Therefore, we used self-reported indicators of consumers' food waste behaviors as well as an experimental choice task. Interpretation was undertaken with the necessary caution required for such data (Pham 2013), and we interpreted only differences between consumer segments, not absolute numbers (more details on the method, data, measures and additional analysis can be accessed in the supplementary file or be provided on request).

Survey and Sample
Consumers from Denmark were surveyed. A 10 to 15-minute questionnaire in national language was sent to a nationally representative online consumer panel by the company Userneeds (http://www.userneeds.co.uk/market-research). Quotas applied to the completed survey responses ensured that the sample represented the country in terms of gender, age, region of residence, income, and education (see Table 1).

Lifestyle Survey Measure
To study lifestyles, we chose a well-established and cross-culturally valid measure of consumers' food-related lifestyle (FRL) ). The original FRL (Brunsø and Grunert 1995) was developed in Western Europe and consists of 69 statements on 7-point disagree-agree Likert scales, covering the following five aspects of a consumer's relation to food in everyday life: 1) purchasing motives, 2) quality aspects, 3) consumption situations, 4) ways of shopping, and 5) cooking methods. The measure has been applied and adapted to different cultural backgrounds such as China (Grunert et al. 2011;Huang et al. 2015) as well as to specific food issues such as convenience food ) and obesity (Pérez-Cueto et al. 2010).
We shortened and adapted the FRL to include attitudes, interests and behaviors identified as particularly relevant for the issue of food waste. To that aim, the study built on qualitative research on consumer food waste behavior (Aschemann-Witzel et al. 2015, 2017a. Moreover, the new statements included in the survey were based on empirical research on factors affecting consumer-related food waste or behaviors known to cause or avoid food waste (Lyndhurst 2010;van Boxstael et al. 2014;Watson and Meah 2013;Williams et al. 2012;WRAP 2013), and on food and sustainability-related research containing similar survey statements (de Boer et al. 2004;Chrysochou et al. 2010;Hartmann, Dohle, and Siegrist 2013;Lea and Worsley 2008). The final 'food (waste)-related lifestyle' measure consisted of 54 statements belonging to five aspects, labeled as follows: 1) purchasing and consumption motives, 2) quality aspects, 3) consumption situations, 4) ways of shopping, and 5) ways of cooking and handling food. The items were pre-tested internally and translated and back-translated into the language study.

Food Waste Measure and Food Waste-Related Indicators
Apart from the measure of food (waste)-related lifestyle, the survey contained measures of the respondents' relation to food waste. These included 1) knowledge of the extent of food waste ('According to what you have heard or would guess: how much of . . . the world's food do you think is wasted (in % across the global food supply chain)? / . . . the foods in households are wasted (in % of the food bought)?'); 2) relative importance of the food waste issue ('How important is it to reduce food waste in comparison to . . . reducing obesity in our society? / . . . reducing environmental pollution in our society? / . . . stabilizing the economy in our society?'), measured on a scale from 1 (Much less important) to 7 (Much more important); 3) self-reported estimation of one's food waste in five food categories ('If you would try to estimate your own household, how much of the following food [Fresh fruit and vegetables, Milk and dairy, Bread and other bakery products, Meat and fish, Prepared dishes/meals] that you buy or cook ends up being thrown away at home?', expressed in %); and 4) frequency of choosing the 'optimal' product across six categories in an experimental hypothetical binary choice task.
Frequency of choosing the 'optimal' product was measured by offering respondents the choice between pictures of an optimal versus a suboptimal product from the following categories (sub-optimality in store/home in parenthesis): (brown spot) apple, (bent) cucumber, (close to expiration date/past expiration date) milk, (close to expiration date/past expiration date) yoghurt, (dented package) juice, and (some broken) biscuits. Respondents were asked, 'Imagine that you're in a supermarket, ready to select [category]. Given an identical price, which one would you choose? / Imagine that you're in your home, ready to select [category]. Which one would you choose?' and then counting how often the optimal product was chosen in the six choices. In this case, we defined suboptimal as "not perfect" in terms of date or appearance, but we did not label the options as optimal or suboptimal, as the definition of 'optimal' is a subjective interpretation by the respondent (see Table 2). As economic theory also would suggest, in the data presented, few consumers chose suboptimal food in the store given it was not reduced in price (see de Hooge et al. 2017 as compared to Aschemann-Witzel 2018). It must be noted that purchase of a suboptimal food does not mean food waste is avoided, given it might be wasted at home instead (for a discussion, see Aschemann-Witzel et al. 2017b).
It was not possible to gather data from two different sources in this survey. Therefore, common method variance (CMV) was addressed ex-ante by procedural measures through using different scale types, randomizing the sequence of the questions (Chang, van Witteloostuijn, and Eden 2010), and using multiple items for all attitudinal constructs (Fuller et al. 2016 Witteloostuijn, and Eden 2010; Podsakoff et al. 2003) indicated that CMV was not a problem in the dataset.

Factor Analysis and Item Reduction
We applied exploratory factor analysis (common factor analysis with oblimin rotation) for each of the five lifestyle aspects separately (see Huang et al. 2015). We conducted these analyses because we wanted to identify the correlational structure within each aspect, but retain the five aspects, as these represent five theoretically derived and empirically well-founded spheres of interaction of the consumer with his or her food; these interactions are convergent with the multidimensionality found in other lifestyle research (Ganglmair-Wooliscroft and Lawson 2010). A split-sample exploratory factor analysis applied to a random half of the data led to the same results.
Through the analysis, we determined which survey statements appeared to carry the same meaning for respondents across at least four of the five countries according to the results of the factor loadings and the lifestyle dimensions to which the statements belonged. More specifically, we retained items if they a) loaded with at least 0.32 on the respective factor (Tabachnik and Fidell 2007), and b) loaded unequivocally on a factor (by a factor loading of at least 0.2 more than any other factor loading). Furthermore, dimensions were kept if they contained at least two items, showed a sufficient reliability with Cronbach's alpha (at least 0.5 or higher, Huang et al. 2015;Kaiser 1974) or -in case of only two items in the dimension -if the correlations were significant and exceeding 0.25 (Tabachnik and Fidell 2007). This process resulted in 32 items pertaining to 13 food (waste)related lifestyle dimensions (see Table 3). These items were used in a factor analysis (fixing the factors to 13), which confirmed the dimensions of the lifestyle measure (KMO .869, Bartlett's test of sphericity p >.001).

Cluster Analysis and Segment Identification
We computed factor scores by calculating an averaged variable for each dimension (Tabachnik and Fidell 2007). We applied a two-step cluster analysis process (Punj and Stewart 1983) to determine how many and which consumer segments emerged from the variables that described respondents across countries on the food (waste)-related lifestyle dimensions. Hierarchical cluster analysis was conducted with four random samples of 100 respondents to assess the appropriate number of clusters across various random subsamples. We assessed the appropriate number of clusters through inspection of the agglomeration schedule and dendrogram. It appeared that a four or five-cluster solution was most appropriate, and we thus conducted K-Means cluster analysis with both four five clusters, finally deciding on the five-cluster solution on the grounds that it provided more meaningful and interpretable results (Tabachnik and Fidell 2007).

Segment Characterization
To describe the cluster characteristics, we used ANOVAs for multiple group comparisons and tested for differences between groups with post-hoc Games-Howell tests (as a robust test when variances are not homogenous) and with Scheffé-tests (when variances are homogenous, tested by Levene-tests). Pearson chi-square tests (two-sided) were used for variables at nominal or ordinal measurement levels. Likert scales were treated as intervals. We undertook interpretation with the necessary caution required for such data (Churchill 1979;Fusions 2014Fusions , 2015: Given we expected consumers to have difficulties assessing food waste and self-reporting their own food-waste behavior, significant differences between consumer segments were interpreted rather than described in absolute numbers.

Results
Based on the thirteen dimensions of food (waste)-related lifestyles, we identified five consumer segments (see Table 4). In describing the segments, we used the concept 'involvement/ involved' to summarize an underlying interest in or importance that is expressed when a segment scores high in the respective dimensions. The consumer segments were characterized on the food waste indicators as well as sociodemographic variables (see Table 5).

Segment 1 -the 'Cooking-involved and Spontaneous'
Segment 1 (called the 'spontaneous' for short) was characterized by an involvement with food in terms of interest in cooking and culinary experience as well as a favorable response to the dimension of taste and of price as a criterion in shopping; The segment disagreed with the 'security' dimension -indicating an interest in culinary experiences and tasting new foods. This segment was specifically characterized by neither planning food purchases nor meals in advance. Regarding sociodemographic profiles, the spontaneous were relatively younger, more often female, and had a low share of high income respondents. Concerning food waste measures (see Table 5), the spontaneous estimated their own food waste at a medium (compared to the other segments), in particularly high for fresh fruit and vegetables, dairy and prepared dishes, and they assessed the topic to be important. Of all segments, the spontaneous least likely mentioned that they use the optimal item first when at home, thus, appeared to use the suboptimal first.

Segment 2 -the 'Price versus Quality-oriented and Disliking Cooking'
Segment 2 consumers (the 'price-oriented' for short) were not very food involved when it came to cooking and to deriving self-fulfillment from the task. Compared to the other segments, they were especially characterized by attention to price as a criterion when selecting food. They were balanced in terms of gender, of medium age, and of relatively lower education (although insignificantly). They reported low levels of food waste in their own household, in particular for prepared dishes.

Segment 3 -the 'Very Involved and Cooking-engaged'
Consumers in segment 3 (the 'involved' for short) was characterized by high engagement and involvement with food overall, as can be seen in the interest in cooking and culinary experience, and a highly favorable response across the majority of dimensions, including planning purchase and meals in advance and considering the price criterion in purchase. They tended to disagree with the 'security' dimension -indicating an interest in culinary experiences and tasting new foods. Segment 3 had a high share of females, a higher medium age, and the lowest share of single-households, thus indicating that persons with families were overrepresented. The involved were most likely to choose the 'suboptimal' food in the supermarket situation compared to the other segments. Involved consumers made relatively low estimates across all food categories for food waste in their households, similar to segment 2; segment 3, however, made the lowest estimation of all segments for the category of bread.

Segment 4 -the 'Good Food-involved and Pricedismissive'
Segment 4 (the 'price-dismissive' for short) consumers were particularly characterized by being uninterested in using the price criterion during shopping. They consisted of moderately food-involved consumers that indicated importance of taste, food safety, and optimal choice in purchase. The segment also had a relatively higher interest in credence attributes of food. Segment 4 was especially characterized by a greater share of respondents with above average income, and had a high share of females and respondents of higher age. Concerning food waste measures (see Table 5), the price-dismissive estimated their own food waste at a medium (compared to the other segments, and similar to segment 1).

Segment 5 -the 'Least Concerned, Normative and Social'
Segment 5 (the 'unconcerned' for short), comprised of not very food-involved consumers, with low levels of interest in, and self-fulfillment derived from, cooking. They appeared to be the least focused on social relations around meals, and relatively least concerned about food safety and the price-quality relation. The un-concerned were the youngest segment, composed largely of males, and the segment was characterized by the highest share of single households. Of all segments, the unconcerned found the topic of food waste least important, reported the highest estimates of own food waste, and were most likely to indicate to choose the optimal in the store and at home.

Discussion
Our findings reveal that consumer segments can be developed along the continuum of 'involvement' with the food issue at hand. Previous studies have indicated the existence of so-called very involved versus very un-involved groups, or 'indifferent' versus 'responsible' groups in the area of food (e.g. Burke, Eckert, and Davis 2014;Delley and Brunner 2017;Mallinson, Russell, and Barker 2016;Verain et al. 2012;Verain, Sijtsema, and Antonides 2016). Various dimensions in our survey, such as food selection, cooking, and eating, can be regarded as an involvement with food (Marshall and Bell 2004).
We also found dimensions of food-related lifestyles, such as enjoying cooking, price-orientation, social relations via meals, or food quality, to be related to the core of characterization in other food-related lifestyle studies (e.g. Grunert et al. 2011;Mallinson, Russell, and Barker 2016;Ryan et al. 2004). The explorative segment characterizations presented in the findings provides a good basis for developing ideas on effective policy and marketing actions (see Table 6 for an overview). Such segment-specific actions may increase the effectiveness of policies in their fight against food waste (WRAP 2016).

Macromarketing and Policy Actions Targeting 'Cookinginvolved and Spontaneous'
The food-related aspects of the 'spontaneous' (segment 1) demonstrate that these consumers can further improve their avoidance of food waste, which might be related to a lack of planning. They are likely to be motivated or proactive when the action relates to food experiences and food enjoyment. If they already act against food waste, the spontaneous are motivated to further reduce waste given they think the topic is important, and may accept new tools to help them achieve this (e.g., phone apps on food planning and handling, suggestions of recipes, credence logos highlighting food waste avoidance contributions). Moreover, as a younger and experience-seeking group, they could be pioneers or trendsetters who spread their knowledge and concern by talking to friends and relatives, through their social networks, or by volunteer membership in NGOs. They might be front-runners in shaping new social norms that favor sufficiency, enjoyment, and experience, instead of oversupply. In sum, the cluster characterization suggests that the "spontaneous" are already in a stage where they possess information (Beverland 2014) and think food waste is important to tackle, and could be leading sustainable behavior via enacting new social norms (Nyborg et al. 2016).

Macromarketing and Policy Actions Targeting 'Least Concerned, Normative and Social'
On the contrary, the un-concerned (segment 5) are unlikely to be motivated by ethical or economic motives, by normative appeals to avoid food waste, and not be put-off by constraints on their food assortment, as they are not particularly interested in food. This segment may respond best to strategies that alter the standard situation, such as nudging (Thaler and Sunstein 2009). For example, segment 5 will not be disappointed by food variety restrictions if stores sell fewer varieties of fruits to avoid food waste of unsold products. Any upstream alteration in the supply chain that does not involve consumers (e.g., packaging improvements to keep products fresh longer) will also work well for this group. Thus, for this consumer segment, social marketing campaigns should be focused on supply chain stakeholders, technological innovations, and alterations of food choice situations, rather than on targeting the consumers. Also, un-concerned consumers need institutional change first (Prothero et al. 2011), which might call for policy makers (Thøgersen 2014) to institute macro-environmental changes (e.g. laws on aesthetic standards, or funding research on technological innovation). Moreover, this segment consists of relatively younger consumers who are more likely to be male, which provides the opportunity for lifestyle changes as they age and move into another life stage (Devine 2005).

Macromarketing and Policy Actions Targeting 'Price versus Quality-oriented and Disliking Cooking'
The price-oriented (segment 2) are aware of food prices, and will probably respond favorably to price incentives to reduce food waste. This can entail price changes in retail or canteen settings (e.g., price reductions of suboptimal foods close to the best-before date, see Aschemann-Witzel 2018) or when selecting fewer side dishes in restaurants. They may also be motivated by monetary bonuses for environmentally friendly behavior (e.g., bonuses for households that have no food waste), or social marketing campaigns focused on economic 'thriftiness' arguments (e.g. for food waste avoidance in households). Given that the price-oriented show less concern about food credence attributes and do not have an overly heightened food safety concern and no particular ambition for cooking out Apply nudging strategies to sell food otherwise wasted in-store Change products and offers to reduce food wastage in store (e.g. reduce assortment) and in households (smart packaging, longer expiration date) Support supply chain stakeholders collaboration in reducing food wastage in the operations and in store Fund research on technology and processes for innovation in the supply chain or on nudging strategies in store Revise food-related laws which affect food waste of the ordinary, they especially might be the target group for suboptimal foods offered in alternative retail stores, including food banks, and they may adapt to foods that do not conform to current market standards if such foods are significantly cheaper. Care needs to be taken to ensure that such actions do not move food waste from the supermarket to households (Setti et al. 2016). However, the low self-reported food waste level of this group suggests that greater problems should not be expected here. Price-oriented consumers can improve their planning, and may therefore be sensitive to practical advice on this aspect, though this advice should be easy to implement and justified by economic savings, as they do not seem to enjoy cooking and food provision.

Macromarketing and Policy Actions Targeting 'Very Involved and Cooking-engaged'
That there is a certain price-orientation during shopping of the very involved (segment 3) indicates that similar considerations can be made for segment 3 as for segment 2. However, focus needs to be given to the fact that the food is safe and of good quality and that the food waste avoidance is a sign of ethical actions of the supply chain actors. The low self-reported food waste and high likelihood of choosing 'suboptimal' items suggests that involved consumers might already be 'best' in avoiding food waste and already possess high perceived behavioral control over food waste (which is necessary to reduce food waste, Visschers, Wickli, and Siegrist 2016). Thus, they may be less of a priority for targeted policies, as their behavior may hardly improve. As involved consumers gain self-fulfillment from cooking, are 'thrifty' given they plan well, choose suboptimal at times, and waste less, they may nevertheless be receptive to further household management advice. This entails storage adapted to each type of fruit and vegetable, recipes for leftovers, and composting inedible parts. It may be particularly effective to include this segment in public social marketing neighborhood campaigns as ambassadors or role-models, where this group may share their advice and experience with others.
The cluster characterization suggests that involved consumers are closest to avoiding unnecessary food purchases, making use of the suboptimal, and enacting sufficiency in their behavior (De Coverly et al. 2008), as they seem to manage their food stock well. It is a segment enacting a social norm that seems to have disappeared through the impact of consumerism (Assadourian 2010). Interestingly, their behavior might be seeing a new recognition, given that new social norms (Nyborg et al. 2016) of simplicity and thriftiness are currently formed in societal groups, as for example by groups that are regarded as 'voluntary simplifiers' (Huneke 2005). Given that segment 3 consists of primarily women and the actions might appear 'housewifely', care should be taken to ensure that their actions can appeal to both genders. The proliferation of food handling that avoids food waste as enacted by segment 4 could be further supported by macromarketing and policy. For example, local governments could support social initiatives that share food surplus (e.g. installing cupboards or fridges to place the food) in communities, or that 'glean' overproduction left on the field or on fruit trees in public spaces (allowing or organizing access).

Macromarketing and Policy Actions Targeting 'Good Food-involved and Price-dismissive'
Finally, the relatively price-dismissive consumer segment 4, characterized by higher income, will probably not respond to price reductions on suboptimal food or to economic incentives. Instead, the price-dismissive consumers might become interested in food waste avoidance through communication that reminds them of their feelings of guilt and anxiety when wasting food. Suggestions that food waste avoidance will increase their well-being and that food waste avoidance is an ethical action conducted by the supply chain ought to be effective. Other possibilities are communications that highlight the selffulfillment aspect of thriftiness and that frame food waste avoidance as an element of good cooking or as a sign of higher status in society. Moreover, price-dismissive consumers have the economic resources to purchase services that support them in planning meals, as for example food box delivery schemes with recipes and the respective measured ingredients. Pricedismissive consumers may also respond well to interesting food product innovations that result in less food waste earlier in the supply chain, such as premium foods or restaurant meals made from surplus. These consumers might act sustainably through 'positive buying' (Harrison, Newholm, and Shaw 2005) and might be receptive to changing social norms in joint meal contexts (Beverland 2014), such as accepting that 'less is better' when inviting guests.

Conclusions and Implications
Based on the results and characterization of the consumer segments, we conclude that consumer groups specifically differ along the following dimensions: first, in their involvement with food in varied dimensions, most notably, their interest in and self-fulfillment through cooking; second, through their orientation towards price as a criterion in shopping; third, in the degree to which they are planning their meals and respective purchases; fourth, based on whether they are concerned with issues such as social relations via meals, food safety and quality. We conclude that comparing the consumer segments on food waste indicators shows that both involved and uninvolved consumers report relatively higher levels of food waste, but they differ in the perceived importance of the food waste issue and the tendency to choose the 'suboptimal' products in the store or in their homes first. Thus, food involvement, planning, price orientation, and food-related concerns are consumer lifestyle characterizations of specific relevance for differences in self-reported food waste and food wasterelated behaviors.
As a practical implication, the study suggests that food market actors and policy makers should take the identified similarities and differences between consumer groups into account when devising actions to encourage consumer food waste avoidance. Both social marketing and commercial marketing that contribute to corporate social responsibility of the company can make use of the recommendations in targeted ways depending on the consumer segment and activity in question. In addition, this study indicates where micro-level approaches reach their limit. To reduce food waste, policy actions (Vittuari et al. 2015) are needed which aim at changing the macro-level context and create more systemic change.
Our findings advance the literature on food waste. By applying the lifestyle concept, we have been able to identify segments of relevance with distinctive patterns that help to understand differences in food waste related behaviors. Our consumer segments reveal why some consumers waste suboptimal foods at their homes, while others do not. The findings can provide some indications for why consumers differ in their self-reported household food waste.
The present study extends previous conceptual research on lifestyle by considering food waste. It does so by adapting and applying the food-related lifestyle to the issue of food waste, and showing that the core dimensions of the measure are applicable. In addition, the findings demonstrate that the newly emerged dimensions of 'food safety', 'purchase planning', 'norms and control' and 'optimal choice during shopping' play a role in consumer food-related behaviors related to the issue at hand. These new dimensions theoretically expand the concept of food-related lifestyle and should be considered in future research.
By looking further into the relation between consumer lifestyles, food, and sustainability, the present research underpins the importance of understanding a sustainable consumption issue in its context, and the value of approaching it from a macromarketing perspective. The findings show that consumers might appear similar, but nevertheless could behave unsustainable in certain aspects but for different reasons. Thus, it is useful to apply multiple approaches both on a micro-and on a marketing-level to tackle unsustainable behavior.
In sum, the present research presents the successful identification of five food (waste)-related lifestyle consumer segments. It also reveals how these can be used to derive further research and suggests recommendations for food waste mitigation targeted to specific consumer groups. Ultimately, the findings may contribute to a better fine-tuning of social marketing actions and supply chain policies addressing the sustainability issue of food waste from a macromarketing perspective.

Limitations and Future Research
As a limitation, we caution that cultural and socio-economic differences among countries can be expected to influence the results and transferability of this research to other country contexts. At the time of the survey, the topic had already received media coverage in Denmark for multiple years. Furthermore, results of the survey may have limitations due to the potential method bias inherent in the self-report element. This means that our findings need to be interpreted with awareness of the type of data, especially because food waste is difficult to assess for consumers (Fusions 2014). A survey about the topic of food waste practices can only partly capture the complexity of the causes of food waste and of the practices and behaviors related to food waste in consumer households, and is prone to potential deviations between what consumers do and what they express. However, the research goal was to quantify lifestyle segments in relation to indicators of consumers' knowledge, perceived importance, and their own behavior regarding food waste.
A second potential weakness of our study is the omission of statements based on differences in meanings. This method has led to multiple dimensions that are based on two statements, which might be considered sub-optimal. Relatedly, because we retained the five dimensions of the food-related lifestyle and tried to cover a rather diverse set of consumer interaction with food, we conducted separate factor analysis and accepted poorer measures of internal consistency (a threshold of 0.5 for Cronbach alpha values). Some researchers might consider such an internal consistency threshold too low. Future research is therefore poised to examine the consistence of our developed food-related lifestyle measure across countries and setting.
Further research might establish whether the consumer groups identified in the present study are indeed more or less receptive towards the discussed segment-unique marketing actions and policies. Future studies should explore the promising characteristics of segments with low food waste and choice of suboptimal food by more in-depth qualitative methods. Conducting surveys of policy acceptance or experiments on the latter issue would make it possible to identify whether the segments are responding favorably to the policy measures that appear best fitting to their profile. The findings contribute to the exploration of the concept of lifestyle within the domain of food. Further research might provide greater insight into how food behavior and waste behavior is conceptually linked, and how certain food waste avoidance behaviors can be explained by tendencies such as 'thriftiness.' Research into food consumer behavior needs to move further into experimenting and testing with suggested recommendations, to test the effectiveness of our findings in real-life context. All in all, it seems that with our results in hand, there are many interesting lines of research and potentials to move the fight against food waste at the consumer level forward.