It’s Easy to Maintain When the Changes Are Small: Exploring Environmentally Motivated Dietary Changes From a Self-control Perspective

Current levels of meat and dairy consumption in Western societies are unsustainable and need to be rapidly reduced to curb climate change (Clark et al., 2020; Committee on Climate Change, 2018; Masson-Delmotte et al., 2019). Modern meat and dairy farming not only contributes to greenhouse gas emissions and environmental degradation (Federici et al., 2015; Vermeulen et al., 2012; Willett et al., 2019) but also raises ethical concerns (Cornish et al., 2016). Further, the excessive consumption of these foods can negatively impact human health (Hansen et al., 2018). Consumers are increasingly aware of the environmental impacts, and many are open to adapting their diets (Sanchez-Sabate & Sabaté, 2019).

Research has generated a large body of literature on vegans and vegetarians (Hoffman et al., 2013; Judge & Wilson, 2019; Rosenfeld & Burrow, 2017), but meat and dairy reducers remain an underexplored group (Graça et al., 2019; Taufik et al., 2019). Especially dairy reduction is under-researched (Sandberg, 2021). While vegan diets may be the most sustainable, it is unlikely that most Western populations would adopt them. Therefore, encouraging smaller-scale reductions among mainstream consumers may be more realistic (Graça, Oliveira, et al., 2015). Consequently, it is important to understand the daily-life experiences of consumers reducing meat and dairy consumption, in order to understand what could be done to best support their efforts. This paper aims to explore the experiences of meat and dairy reducers who are driven by environmental motives.

We define meat and dairy reducers as individuals who are actively trying to reduce their meat and dairy intake, even though we acknowledge that they may not always be successful. Meat and dairy reducers may vary widely in their consumption frequencies of certain foods and their dietary identification (Malek & Umberger, 2021). The reduction process often follows a specific order based on food status hierarchy (Grassian, 2020), potentially based on perceived human-animal similarities (Rothgerber, 2014). Reducers often begin avoiding red meat first, followed by white meat, dairy, eggs, then seafood. However, there is variability in this pattern (Calton et al., 2014). Additionally, there is variability in dietary group identification: meat and dairy reducers may identify as omnivores or as flexitarians, vegetarians, or semi-vegans. At face value, these dietary groups may seem categorically different in how frequently people consume animal foods, but consumers understand and identify with these groups in variable ways. For example, individuals who eat fish, yet do not consider fish as meat, may identify themselves as vegetarians (Rosenfeld & Tomiyama, 2021), and flexitarians who do not consume meat may not identify with vegetarians to avoid stigmatisation (Rosenfeld et al., 2020). We will explore this issue in the current study to better understand the role that such identity processes may play for meat and dairy reducers.

Meat and dairy reducers may reduce consumption of these foods for various reasons, which might translate to different processes and outcomes. Numerous researchers have explored the primary motives that meat and dairy reducers hold, such as health, animal welfare, and the environment (De Backer & Hudders, 2014; Grassian, 2020; Hielkema & Lund, 2021; Lacroix & Gifford, 2019; Rothgerber, 2015). While one systematic review suggested that individuals reducing meat intake for ecological reasons are a minority (Sanchez-Sabate & Sabaté, 2019), another more recent review highlights that both sustainability and health are currently important motives for reducing meat and dairy consumption (see Dagevos, 2021). Given the climate emergency, we predict that in the future, the group of consumers trying to reduce their meat and dairy consumption for environmental reasons will increase (Bastian & Loughnan, 2017; Mathur et al., 2020). Therefore, the current article focuses on exploring the experiences of environmentally motivated meat and/or dairy reducers.

The literature on shifting from meat-based to plant-based diets is rapidly growing (Graça et al., 2019; Stoll-Kleemann & Schmidt, 2017; Taufik et al., 2019), and has identified a broad array of barriers and enablers to reducing meat consumption. Meat disgust (Rothgerber, 2014), awareness of climate impacts (Kirsten et al., 2020), as well as supportiveness from others (Haverstock & Forgays, 2012; Hielkema & Lund, 2021), are some of the factors that enable consumers to reduce their meat and dairy consumption. In contrast, meat attachment, hedonic enjoyment of meat (Graça, Calheiros, et al., 2015), and the belief that consuming meat is natural, normal, necessary, and nice (Piazza et al., 2015) can be barriers to reducing the consumption of meat. Meat reducers also need to confront norms and attitudes that favour meat-eating (Grassian, 2020), especially for men, and may struggle to overcome their own meat consumption habits (Hoek et al., 2017). Barriers may also include the lack of information and cooking skills, social prejudices and lack of support, as well as moral disengagement, liking of meat, and frequent meat-eating (Graça et al., 2019). Meat reducers may further face barriers such as food neophobia and identity-incongruence (Hielkema & Lund, 2021). For young adults, their sense of control over their food choices, cravings, conflicting eating motives, and compromises at social gatherings may be additional barriers (Kemper & White, 2021).

Similarly, environmental factors such as visibility, proximity, and availability of foods, can strongly shape choices. A review of 15 articles on nudging consumers’ food choices shows the promising effect of making changes to the food environment on choice without depriving consumers of choices (Bucher et al., 2016). Other research showed similar effects of nudging on increasing fruit and vegetable intake (Broers et al., 2017) and healthier diets and nutritional choices (Arno & Thomas, 2016).

Thus, there are many barriers and facilitators to reducing meat and dairy intake, and meat and dairy reducers need to navigate often challenging food environments with varying availability of sustainable food choices. This suggests that self-control might play an important role in maintaining reduction behaviour. To our knowledge, no research has attempted to explore the experiences of environmentally motivated meat and dairy reducers from a self-control perspective. Therefore, that is the focus of the current article.

Meat and dairy reducers often need to overcome pre-existing meat-eating and dairy-eating habits, temptations to consume meat and dairy, or social norms promoting the consumption of meat (Zur & Klöckner, 2014). Encountering such barriers may lead to the need for self-control. While the evidence on barriers to reduction is increasing in the literature, little is known about the psychological experiences of meat and dairy reducers and the social and environmental factors influencing them (Graça et al., 2019). In this paper, we address these gaps by exploring whether reducing meat and/or dairy requires self-control, the situations that necessitate the exertion of self-control, and how individuals manage these challenges.

Self-control has been defined as “the ability to restrain impulses in the service of greater goals and priorities” (Milyavskaya & Inzlicht, 2017, p. 1). While this has often been understood as requiring effort, it has also been suggested that self-control can be effortless, for example when adaptive habits are developed and automatically inhibit desires (Adriaanse et al., 2014). The relationship between self-control and behaviour change maintenance has been previously examined mainly in health domains. For instance, self-control predicted adherence to weight-loss programs (Baker & Kirschenbaum, 1993; VanEpps et al., 2016). In the domain of meat reduction, one study found no effects of self-control on adherence to vegan, vegetarian, and health-related diets (Cruwys et al., 2020). Others suggested that reducers who identified as omnivores more strongly reported a need for self-control than vegetarians (Allen et al., 2000), perhaps because vegetarians rely more on their social identity and motivation to adhere to their diets rather than their psychological capacities. To the best of our knowledge, there has been no research with the primary aim of exploring the role of self-control in the process of reducing meat and dairy intake, and we aim to fill this gap in the literature.

Meat and dairy reducers may experience conflict between their long-term reduction goals, and their habits and desires to consume these foods. Like most eating behaviour, the consumption of meat and dairy is influenced by habits and liking, as most people have frequently consumed these foods from an early age (Papies, Johannes, et al., 2020; van’t Riet et al., 2011). Inhibiting habits and food temptations typically requires self-control (Adriaanse et al., 2014; van’t Riet et al., 2011; Wood & Neal, 2007), as also suggested by dual-process theories (Hofmann et al., 2008). Indeed, self-control processes are often activated when conflict between desires and higher-order goals is detected (Inzlicht & Schmeichel, 2012; Kotabe & Hofmann, 2015). Some of these processes may be deliberate and effortful, but dual-process models of behavioural regulation suggest that effortful processes are more likely to guide behaviour when sufficient self-control resources are available (Baumeister et al., 1994; de Ridder et al., 2012). Thus, both self-control processes and self-control resources may play a key role in enabling the reduction meat and dairy intake and in developing strategies to maintain this behavioural change.

The mechanisms of behavioural maintenance that have been identified for successful health behaviour change may also play a role in reducing meat and dairy intake for environmental reasons. Health behaviour maintenance models suggest that self-control is essential for dealing with sources of tension, or struggles of maintenance (Greaves et al., 2017; Kwasnicka et al., 2016). Tension may arise from external influences (e.g., social pressure) and individual capacities (e.g., motivation). In the context of our research, we propose that managing tension, such as inhibiting desires and old habits, may be required to transition to and maintain the reduction of meat and dairy.

Another self-control challenge that may influence the reduction process is the management of conflicting social norms. Social norms and the social environment influence eating behaviour, such that people are motivated to eat the foods that are normative in their society or the social groups they identify with (Demarque et al., 2015; Higgs, 2015). Consuming meat and dairy is the norm for most Western societies (Willett et al., 2019), and deviating from this norm may challenges to one’s social identity. Vegans and vegetarians have often managed this challenge by developing a strong sense of identification with their dietary group, a key factor in the maintenance of their diet (Cruwys et al., 2020). On the other hand, flexitarians view their diet as less central to their identity, which might make their transition more challenging (Rosenfeld et al., 2020). Meat and dairy reducers may not necessarily strongly identify with any dietary group. Little is known about meat and dairy reducers’ social identities and how they shape their experiences and behaviours. The current research addresses this gap.

Understanding what facilitates and hinders meat and dairy reduction is important, because it could inform ways to better support individuals transitioning to reduced meat and dairy diets (Graça et al., 2019). Here, we explored the experiences of individuals in this transition, to ultimately identify ways to better support them toward more sustainable diets. We conducted qualitative research to address the following research questions:

1. How do people perceive and experience the transition towards more sustainable eating behaviours?

2. Does shifting towards and maintaining sustainable eating behaviours require self-control? If so, what are the situations where people feel they need to exert self-control? And how do they deal with these challenges?

3. What are the roles of habits, identities, and social norms in the transition towards and in the maintenance of sustainable eating behaviours? And how do people experience them?

We developed an online qualitative survey for our comparative case study design (Braun et al., 2020). A key advantage of online survey methods is that it allows for potentially rich data from a broad representation of individuals and experiences (Braun et al., 2020). We asked participants questions related to their current diet and their reduction goals and experiences. All study materials, including the full survey schedule and data, are available on the Open Science Framework (OSF; https://osf.io/vuhsy/).

We held question creation meetings using the research aims, research questions, and relevant theories on behavioural change and self-control (Greaves et al., 2017; Kotabe & Hofmann, 2015; Kwasnicka et al., 2016). We ensured an open-ended format (Korstjens & Moser, 2018) and piloted the survey for comprehension (n = 5). The resulting survey included 14 open-ended questions covering various aspects of the reduction experience (see Table 1).

Then, to yield a deeper understanding of participants’ experiences, we asked four background questions on current meat and dairy intake (e.g., “In a typical week, how often do you eat meat?”, “In a typical week, how many times would you like to eat dairy?”), on current diet (“Which of these describes you best at the moment?” response options: “I am currently trying to reduce meat, not dairy”, “I am currently trying to reduce dairy, not meat”, “I am currently trying to reduce both meat and dairy”), and on the desired future intake of these foods ( “In a typical week, how many times would you like to eat meat?”, “In a typical week, how many times would you like to eat dairy?”). Finally, we asked demographic questions, such as age, gender, occupation, education, perceived social class, and nationality.

Question 13 asks for participants’ suggestions about what they, or others, could do to make their dietary transition easier and more enjoyable. The responses to this question were gathered into one subtheme (3.3.3) where we included counts of participants mentioning each contextual strategy. Although generating meaning from the data partly depends on more comprehensive patterns (Sandelowski, 2001), we believe that displaying the counts for this subtheme will provide a comprehensive overview for the reader of what lay people think could help them in their reduction process.

The study received ethical approval from the University of Glasgow Research Ethics Committee. To enhance transparency, we pre-registered the study and documented the complete research process (see OSF; https://osf.io/bhvyw). Pre-registration can help ensure that the a-priori decisions to observing the data are maintained and encourage new intentional decisions in case of changes (see Haven & Van Grootel, 2019; Kern & Gleditsch, 2017).

Participants were recruited through the online research platform Prolific (prolific.co). Inclusion criteria were: living in the UK, between 18 to 65 years of age, and being fluent in English. One thousand participants answered three screening questions (average duration 1 min; payment £0.1); “Are you currently trying to reduce your meat and/or dairy intake?”, “What is the most important motive for you to reduce meat and/or dairy right now?”, and “Which of the following dietary categories do you see yourself belonging to?”. We selected participants who were trying to reduce meat and/or dairy and indicated that their most important motive was environmental. In the rest of this article, we refer to these participants interchangeably as meat and dairy reducers or meat and/or dairy reducers. We excluded vegans and self-identified omnivores who did not want to reduce meat intake (n = 192). Then, 239 participants who satisfied the inclusion criteria (female: 192/ male: 42/ other: 5) were invited to the main survey. Participants completed the survey on Qualtrics (average duration 30 min; payment £3.75). In line with research on reducers (Dagevos, 2021), most participants were female.

We stopped data collection once 80 participants had completed the main survey. This predetermined sample size was based on recommendations for sample size with online qualitative surveys (Braun et al., 2017, 2020; Malterud et al., 2016) and our budget limits (see OSF). We also used the definition of data saturation as “the point where no new and meaningful information is being generated” (O’Reilly & Parker, 2013). Saturation of new information started at the 65^th case. However, we were more interested in the quality of the generated data described as data ‘sufficiency’ or ‘quality’ (Braun & Clarke, 2019). All authors discussed data saturation and quality thoroughly in weekly analysis meetings.

A summary of demographic information about our sample is presented in Table 2. Detailed demographics can be found in the Supplemental Materials on the OSF.

Self-identified omnivores in our sample reported consuming meat meals moderately often (range: 4 – 10 meat meals per week) and dairy more frequently (range: 5 – 21 dairy-containing meals per week). Self-identified flexitarians reported consuming meat and dairy less frequently (1-10 and 0-14 meals with meat/dairy, resp.). Some self-reported vegetarians reported eating meat (including fish) once or twice a week and dairy between 0 to 14 meals per week, while others described themselves as “strict vegetarians” and reported eating no meat and eating dairy at most a few times a week.

In our analysis, we developed thematically organised patterns throughout the dataset, supported by quotes. We adopted a reflexive thematic analysis (Braun & Clarke, 2006, 2013, 2014, 2019) and followed the six stages of reflexive thematic analysis using NVivo Software (Windows Version 12) as a qualitative analysis management tool (Castleberry & Nolen, 2018; Maher et al., 2018; Silver & Lewins, 2014). We used a hybrid approach of inductive and deductive coding (Fereday & Muir-Cochrane, 2006), generating broad data-driven conclusions and using pre-existing theories to guide our observations with an a-priori list of codes (see OSF). In other words, we explored and analysed the data separately from the relevant theories highlighted in the above section, and later, we discussed how the data links to the theories. The generated codes followed the process described in Table 3.

Since the flexible theoretical framework of this method can lack a grounding orientation (Braun & Clarke, 2013), we adopted critical realism as a methodological framework compatible with thematic analysis (Braun & Clarke, 2019; Fletcher, 2017). Critical realism’s approach consists of causal or generative mechanisms (Bhaskar, 2013; Sayer, 2010). It begins with identifying the social problem and infers backwards to conceptualise from empirical data about the phenomenon whilst drawing on understandings from previously established knowledge in a different context. This perspective aligns with our aim to understand the subjective meaning and experiences and identify the mechanisms that underpin the maintenance of reducing meat and dairy for environmental reasons.

We conducted an additional exploratory analysis to explore how our 80 participants reported their dietary group membership in the pre-screening study versus the main study. In the pre-screening study, participants responded by choosing one of the following categories: ‘omnivore’,’flexitarian’, ‘vegetarian’,’vegan’, or ‘other’. We compared these responses to participants’ answers to the open-ended question in the main survey ’Do you currently see yourself as a person who eats meat, a flexitarian, a vegetarian, or a vegan? What do you think about these groups? Please explain’.

Credibility was ensured through persistent observation of the data (Korstjens & Moser, 2018). Detailed descriptions of the participants’ experiences and demographic context enhance transferability, making connections to help provide a comprehensive understanding of the participants’ setting. While there are no set rules as to how many quotes should be used and from how many participants, the focus is on detecting themes that are reflected across the data (Eldh et al., 2020; Sandelowski, 1994). We held meetings periodically during the analysis phase to bring in different perspectives and build consensus. We also discussed and integrated the deviant cases from the patterns that emerged from the data analysis, to support a reflexive approach to research (see OSF). And finally, as the authors’ positionality is pivotal in qualitative research, we declare that LW and EP do not consume meat and are currently trying to reduce dairy intake, while KB seldomly consumes meat and dairy and identifies as a flexitarian. LW kept a reflexive diary (Langdridge, 2007) throughout the data analysis process (see OSF).

Participants’ responses varied across questions and ranged between 14 to 236 words per response, with an average of 73 words per response and no missing data across responses. We generated three themes from the data (see Table 4). The full supporting quotes for all themes can be found in the NVivo file uploaded on the OSF. Typographical errors were corrected to safeguard the flow of the quotes while conserving the meaning (see OSF). In each subtheme, tables include extracts from participants’ experiences alongside a higher-order descriptive pattern that, together, provide a comprehensive overview of the theme at hand. Following each quote, we provide participants’ self-reported gender to add depth to the illustrative purpose of our chosen quotes. Specific symbols in participants’ extract include: […] indicating where text has been removed to avoid redundancy and [Text] indicating where text has been replaced for clarifications and descriptions.

Participants expressed many motives and triggers that prompted their intention to reduce meat and/or dairy consumption. Environmental motives together with other motives such as financial or health concerns, or animal ethics made participants want to limit their consumption of these foods (e.g., P40, M). Most participants described that their decision to change emerged gradually due to a cumulative awareness of the negative impacts of the meat and dairy industry on the environment. For instance, they continuously and increasingly sought knowledge from the media or the news (e.g., P20, F). Some changes occurred due to their worry and anticipation of future events (e.g., P50, F).

A few participants reported being triggered by turning point events, specific occurrences such as the Australian wildfires or watching slaughterhouse footage (e.g., P68, M), watching documentaries made by activists (e.g., P71, F), and their limited spending capacity due to the COVID-19 lockdown (e.g., P79, F). Nonetheless, environmental impacts and animal welfare provoked awareness of conflicting attitudes, beliefs, and behaviours that many participants held at the time of change initiation. As a result, participants’ intentions (e.g., P50, F; P71, F; P21, F) and actions (e.g., P68, M) to reduce were primed by the discomfort they experienced. Most participants who did not want to eliminate meat entirely also strongly identified with meat consumers and viewed meat as central to their identity (e.g., P8, M). For supporting quotes, see Table 5.

In summary, the data depicts various trajectories to changing behaviour. Sometimes, change was sudden, while other times, it gradually emerged through actively seeking information and increasing awareness about environmental, health, and animal welfare factors. These changes were initiated by past events or in anticipation of future events.

Participants described their liking of meat and dairy, and their desires to eat these foods. These desires varied in strength (e.g., urges, cravings, or temptations). Some participants struggled with the idea of completely cutting out meat, because they would miss the enjoyment from eating it (e.g., P55, F). They expressed how effortful they found it to eliminate these foods from their diets, especially when the sensory features of the foods triggered desires for meat and dairy (e.g., P66, F; P13, M), such as their smell, taste, and texture. Some participants described hunger as a factor that hindered their meat and dairy reduction, and some shared experiences of not feeling satiated after vegetarian dishes (e.g., P6, F). They also mentioned that it was easier to reduce meat consumption during the summer months, as opposed to winter, due to their desire for comforting foods in winter (e.g., P59, F).

Some participants justified their strong cravings by tying them to their felt bodily changes or perceptions of nutrition deficiencies such as iron and Vit. B12 (e.g., P38, F; P18, F). Many described that moods, such as feeling forgetful, distracted, lazy, stressed, or bored, triggered desires to eat more meat or dairy (e.g., P74, M; P26, F). A few participants shared experiences with feelings of idleness and tiredness towards the end of the week or day. In these situations, participants chose comfort, familiarity, and convenience, over the harder task of resisting meat. For supporting quotes, see Table 6.

In sum, participants described liking and desires for meat and dairy foods. Internal situations such as hunger, health concerns, and various mood states often prompted the desire to eat meat or dairy.

Most participants experienced a range of conflicting motivations in many situations prior to and after consuming meat and dairy. The uncomfortable experiences of cognitive dissonance, where one’s actions are not in line with one’s beliefs, left some participants feeling guilty. Participants varied in their detection of conflict and motivation to control these difficulties and efforts. Some reported not feeling conflicted, felt comfortable bypassing the dissonance and justified why they do not want to fully eliminate meat or dairy (e.g., P53, M).

Participants described the conflicts between their desires to consume meat and dairy and their longer-term reduction goals. For instance, some felt conflicted between the desire to follow through with their new habits and the desire to consume comforting foods. They also mentioned the desire to eat a plant-based diet and to eat a nutritious diet, which requires additional awareness and research (e.g., P5, F). Others experienced strong desires to consume meat and dairy despite their awareness of the environmental and moral impacts of where their food was sourced from (e.g., P70, F).

At times, falling back into old habits and consuming meat and dairy also led to other emotional experiences of conflict, such as feeling disheartened (e.g., P57, F), worried, less determined, or even questioning own beliefs and reasons for wanting to reduce (e.g., P18, F). Finally, participants described conflict arising from having to choose from many options that vary on many dimensions (e.g., ethics, nutritional qualities, or taste) and from having to deal with conflicting information (e.g., P52, F; P32, F). For supporting quotes, see Table 7.

Overall, participants mentioned the many internal conflicts between their desires for meat and/or dairy and their longer-term reduction goals. As a result, they described cognitive dissonance, negative affect, and the struggle to balance sustainability, health, and taste.

Participants noted that the food culture in the UK normalises meat and dairy consumption. Participants found it effortful to maintain resistance against the mainstream environment that normalised eating these foods since childhood (e.g., P25, F; childhood conditioning: P49, F). The availability of meat and dairy alternatives and the proximity and distribution of food shops affected participants’ dietary choices. Participants who had started reducing meat and dairy intake several years ago reported that the availability of alternatives was much better now than in the past (e.g., P36, F). Nonetheless, most participants described that local food environments (e.g., shops, take-away food outlets, restaurants) were not particularly encouraging of vegan and vegetarian eating. However, this was better in summer (e.g., P17, F).

Most participants found vegan and vegetarian eating more difficult when eating out than at home. Although our survey was conducted during the COVID-19 lockdown, when restaurants were closed, participants described the lack of availability of meat-free or dairy-free options when eating out in the past (e.g., P40, M; P38, F). In contrast, most participants found eating at home was effortless, especially if they were in control of the food (e.g., P38, F). However, special occasions, such as Christmas, were challenging, mainly because they encouraged old habits of eating traditional foods and reduced participants’ sense of control (e.g., P41, F).

Affordability was another factor influencing participants’ food choices. Many mentioned their willingness to try meat and dairy alternatives, but found those foods more expensive than the foods they were trying to avoid (e.g., P38, F). Meat promotions enhanced participants’ temptations to consume meat in restaurants when they wanted value for money (e.g., P51, F). Health-motivated participants avoided meat alternatives that they considered ‘highly processed’ and sought to source more expensive but higher quality meat from local or organic stores (e.g., P49, F). Participants reported feeling conflicted or uneasy (e.g., P30, F) when the food environment prompted them to make choices incompatible with their reduction goals (e.g., P5, F), for instance, when they wanted to order take-out foods and there were few or no meat-free options. For supporting quotes, see Table 8.

Overall, participants’ responses evidence the influences of the food environment on their choices. The availability of foods, environmental food cues in different contexts (e.g., shops, in restaurants, at home generally, or at home during special occasions), and affordability impacted participants’ purchase and consumption decisions. This, in turn, often led to experiences of conflict.

Many participants mentioned negative social perceptions of specific dietary groups. Self-reported vegetarians and self-reported omnivores both negatively perceived the flexitarian label (e.g., P69, M; P21, F). Although some participants identified as flexitarian, they found the term flexitarian vague and unclear, did not necessarily want to be identified as a flexitarian in their social context (P23, F), and indicated that this unclear dietary category may not be taken seriously by others (e.g., P29, F; A, P18, F). In addition, many participants and their social circles perceived vegans or vegetarians as a clear out-group (e.g., P14, F). Regardless of identification, participants reported perceiving negative attitudes from others that led them to doubt their own beliefs, negatively impacting their behaviour. While most of these impacts came from close family members (e.g., B, P18, F), a few participants reported being hindered by friends (e.g., P40, M).

Participants’ sense of control diminished around others who consumed meat and dairy. Some reported feeling conflicted in these situations and found giving in was the easier choice (e.g., P34, F). Another significant barrier to participants’ reduction was the traditional mindset of older generations around food (e.g., P41, F). When participants were offered foods containing meat or dairy, they did not refuse these foods, because they feared causing inconvenience to others (e.g., P4, NB) or wanted to avoid the confrontation (e.g., P71, F).

For some participants, having different diets within the household was a major barrier, causing additional expenses and food waste (e.g., P5, F). Others reported having full support from members of their household who were also reducing meat or dairy intake, or were reducing for longer periods of time. The variability in support impacted participants’ enjoyment and effort. For many participants, social support facilitated their reduction maintenance. Participants felt validated and encouraged by social support and others’ reduction efforts (e.g., P5, F), which increased their self-efficacy (e.g., P37, F). For others, however, this was more challenging: for example, a male participant reported facing stereotypes around men for trying to reduce the consumption of foods commonly associated with masculinity (e.g., P40, M). For supporting quotes, see Table 9.

In sum, attitudes towards dietary groups affected participants’ desire to identify with them. Many participants’ responses highlight the social challenges they encountered when changing their diet, while some also reported supportive social influences, especially from others with similar dietary goals.

Low-effort strategies, such as behavioural substitution, were most helpful, because they provided comfortable and small changes that fit participants’ reduction and environmental goals, but also their liking and taste. These strategies entailed, for example, increasing the consumption of vegetables, other types of animal-based proteins, plant-based alternatives, or meat of better quality (e.g., P78, M; P22, F). Participants also emphasised positive feedback strategies or rewards from experiencing health benefits from their dietary change, which helped maintain their motivation (e.g., P36, F). Additionally, participants often re-evaluated their goals through this feedback process. They assessed their new habits, successes and failures, or enjoyment of their reduction experience, encouraging them to pursue their reduction goals further (e.g., P12, F).

Some participants acknowledged the effort and perseverance required for forming new habits, mentioning willpower, self-determination, and willingness. Once they experienced achievements and new habits, their confidence and self-efficacy increased, while the effort needed to manage barriers decreased (e.g., P43, F; P35, F). Participants noted that the process of changing their behaviour was often messy. They employed reframing strategies to accept the efforts needed to deal with challenging situations and reduce the dissonance when they had not acted according to their reduction goals. Such reframing was often effortful, but participants reported rewarding themselves for their efforts and progress when abstinence from meat and dairy was successful (e.g., P27, F; P26, F). They adopted a flexible mindset to help resolve and balance conflicting motivations (e.g., P60, F).

To stop themselves from consuming meat or dairy when feeling tempted, participants used cues or reminders of the health effects of eating meat and of the negative impacts of meat and dairy on the environment, future generations, and animal welfare. These reminders increased their sense of agency and impact on the world (e.g., P51, F; P18, F). Some participants described cooking new recipes as a strategy to deal with desires to consume meat and dairy, starting with small, achievable cooking tasks, while increasing the variety of dishes they could prepare to fuel their sense of enjoyment (e.g., P30, F; P68, M). Finally, some participants reported simply liking meat and dairy alternatives more as time passed (e.g., P30, F; P24, M). This shift in liking may result from associative learning from their social context or the increased exposure to these alternative foods. For supporting quotes, see Table 10.

In sum, participants evidenced both low-effort substitution and effortful reframing and reminder strategies to manage their internal struggles and conflicts throughout their reduction experience. Experiencing tangible health benefits helped participants maintain their efforts.

Participants resolved conflict from food environments with various low-effort strategies, such as avoiding the exposure to tempting foods (e.g., P48, F) and tempting choice situations (e.g., P1, F). Participants also used communication strategies, for example, when they feared causing inconvenience to others or when others had ignored or forgotten their diet (e.g., P30, F). Others hid their dietary identity from others to minimise negative perceptions (e.g., P36, F), and sometimes gave in to temptations in social situations, compensating later by restricting their meat consumption (e.g., 71, F).

Action planning and meal planning were described as very valuable, and helped with time management, effort, and mindfulness. Participants used planning to increase the availability of non-meat or non-dairy meals at home, including freezing plant-based and prepared foods (e.g., P15, F; P16, F). Some participants rejected negative mainstream perceptions of vegans and vegetarians. Those surrounded by others with previous experience in reducing their meat and dairy intake, for instance, vegans and vegetarians, found their support helpful (e.g., P4, NB), for example, for discovering and exchanging recipes and for holding each other accountable. For supporting quotes, see Table 11.

Overall, participants used strategies such as avoidance, communication, action planning, and recruiting social support to resolve the conflicts between their dietary goals and the food and social environment.

Participants described what they or others could do to make their reduction experience easier and more enjoyable.

Participants mentioned individual-level strategies (n = 10), stating that these mainly relied on self-control (n = 7). Participants also mentioned increasing their knowledge through recipes, videos (n = 4), and following social media accounts (e.g., P38, F) to maintain their motivation or continue with their improvements (n =2). One participant mentioned the need to communicate more effectively with others to reduce temptations when offered meat (e.g., P20, F). As a result of contentment with their dietary reduction, some mentioned contemplating further reduction after receiving a new cookbook (e.g., P49, F).

Many participants mentioned interventions they would like others to do, mainly organisations (n = 61), such as increasing availability, accessibility, and variety of foods in supermarkets and accessibility to easy and creative recipes (n = 57), availability of fresh fruits (n = 1), and accessibility and availability of “ethical” meat (n = 2). Participants wanted supermarkets to promote healthy, meat-free, dairy-free food options, rather than only unhealthy ones (n = 25). They also mentioned policy interventions that would increase taxes on the cost of meat (n = 2) (e.g., P9, F) or ban factory farming (n = 1). Others mentioned the need for better advertisement (e.g., P35, F), making vegan alternatives more attractive and meat and dairy foods less attractive (n = 2).

Participants wanted more support from their family, partners, or children (n = 24), and mentioned wishing that their social circle or the general population followed the same diet as them (n = 20). Participants mentioned that their social circle and the general public needed to better understand the negative impacts of the animal industry on the environment (n = 7) (e.g., P13, M) and the impacts of their choices (n = 1), although most emphasised that diet is a personal choice (n = 79) (e.g., P11, F). One participant mentioned that shifting social norms would help.

One participant proposed developing a phone application to increase their sense of community, exchange recipes, and receive feedback on environmental impact (e.g., P71, F). Participants also suggested the need to improve the taste, texture, and quality of sustainable foods (n = 11), especially the taste of cheese alternatives (n = 6). For supporting quotes, see Table 12.

Dietary self-description in the main study was not always consistent with the pre-screening self-identification, particularly for omnivores and flexitarians. For instance, 17 participants who self-identified as omnivores or flexitarian in the pre-screening study later chose not to identify with any dietary group. Similarly, 43 participants who identified as omnivore or flexitarian in the main study reported that they would not choose these labels to identify themselves in social settings, despite choosing an appropriate label (e.g., “To be honest, I’d never heard of Flexitarian until now. But I guess I am flexitarian then.” P68). All participants mentioned the importance of being flexible with their diets. When given the opportunity to express their dietary identity openly through open-ended questions, some of the self-reported omnivores and flexitarians deviated from their previous dietary identification responses. They reported instead that they preferred not using any label, possibly because this better depicts how they present themselves in their daily lives (e.g.,”I just see myself as a person who eats meat. I am unsure what a flexitarian is.”) or because the pre-screening question on diets was multiple choice, whereas the question in the main study allowed for open responses.

These findings could be important to consider in future studies. As lay people’s representations of how they identify themselves may differ from that of researchers, researchers interested in the behaviour of reducing meat and dairy intake should consider not recruiting participants using labels such as “flexitarian”, but rather by focusing on the behaviour of reducing. For further details on this exploratory analysis, see OSF.

This study was designed to explore the experiences of environmentally motivated meat and dairy reducers from a self-control perspective, particularly to understand the role of habit, identity, and social norms in this transition.

Our analysis generated three main themes. The first theme reflects conflicting motivations and the need for self-control in reducing meat and/or dairy intake. The second theme illustrates the influence of food and social environments, such as availability and cost of foods, attractiveness of meat and dairy-based dishes, as well as negative social feedback and social support that impacted behaviour change. The third theme captured the strategies that participants used or said that they could use to help manage the conflicts and challenges resulting from their conflicting motivations and from their food and social environments. The preferred strategies were food substitutions and avoiding temptation. Graça et al. (2019) have suggested that further research is needed to better understand barriers and enablers of the individuals’ capability, as well as aspects of the environmental opportunities, that may hinder or promote sustainable behaviour change. Our work addresses this and shows that self-control resources, social environments, social identity, and affordances of the food environment play key roles in the reduction process, and that dietary identity challenges and the need for self-control vary across situations.

One of the aims of this study was to explore the need for self-control in reducing meat and/or dairy intake. We found that most participants experienced self-control conflicts and, at times, cognitive dissonance and doubts. Self-control conflicts arose when goals to reduce meat and dairy intake were incompatible with desires and habits. We also observed conflicts between two incompatible goals (e.g., reducing meat intake and saving money). For instance, participants found it difficult to choose tasty, sustainable, and healthy foods to replace meat or dairy. Consistent with self-control theories (Inzlicht & Schmeichel, 2012; Kotabe & Hofmann, 2015), we found that detecting conflict prompted self-control processes, and these were often experienced as effortful.

In line with dual-process theories (Hofmann et al., 2008), our results suggest that automatic processes (e.g., desires and habits) often overrode controlled processes (e.g., deliberate pursuit of long-term goals), especially when resources were depleted. In such situations, for example when they felt stressed or tired, participants preferred the convenience of eating familiar foods. Self-control research indicates that working memory capacity and other factors moderate self-regulatory outcomes (Hofmann et al., 2008). Future research could examine which self-control moderators are most important in the transition to sustainable eating.

In line with health behaviour change maintenance research (Greaves et al., 2017; Kwasnicka et al., 2016), our findings suggest that self-regulation of thought and behaviour is essential for dealing with sources of tension during behaviour change maintenance. Participants used various strategies for this. Effortful strategies included persistence, reminders, action planning, meal planning, and effective confrontations with challenging social influences. Low-effort strategies included the avoidance of choice situations, and easy and comfortable meal substitutions. The effort required for reframing strategies varied across participants. Adopting a flexible mindset as well as being open to experiences encouraged persistence and acceptance of lapses. Most participants reported continuously seeking out information from the media. This strategy fuelled the maintenance of the reduction behaviour.

Our findings are consistent with research suggesting that people who identify as vegetarians may be more open to experiences (Milfont et al., 2021), and this personality trait may make sustained dietary change easier. It is possible that, compared to people making health behaviour changes, environmentally motivated meat and dairy reducers are more intrinsically motivated, which might help their behaviour change. At the same time, it is possible that meat and dairy reducers are extrinsically motivated by social norms. Future research could address the role of intrinsic and extrinsic motivation in the maintenance of sustainable eating behaviour.

Our research points to the important role of social identities in the reduction of meat and dairy intake. Consistent with previous findings (Rosenfeld & Tomiyama, 2021), participants’ self-reported meat and dairy consumption did not map perfectly onto dietary identities; for example, some flexitarians reported eating more meat than some omnivores (see section 3.4). Aligned with previous findings (Rosenfeld & Burrow, 2017; Rothgerber, 2014), our findings suggest that participants’ reports of belonging to a dietary group were based on their commitments to their reduction efforts, their social identification, dietary motivation, and adherence.

Additionally, our study adds insight into the highly variable ways in which people conceptualise their dietary identity. A large proportion of our participants who self-reported as omnivores or flexitarians in the pre-screening also reported negative attitudes towards dietary labelling (see section 3.4) in the main study. Possibly, dietary identification varies over time, and those with more variable diets do not find these labels helpful. There was less resistance to dietary labelling among self-reported vegetarians, compared to other participants in our study. The reason could be that vegetarians’ identity is more central to their sense of self (Rosenfeld & Tomiyama, 2019).

Our findings are in line with the identity-based motivation theory (Oyserman, 2015), which suggests that differences in context change people’s self-concept and identity-related motivations. For example, participants reported that eating a vegetarian diet was easier in situations where their reduction goals were salient, and where identity conflict was low. However, when eating out with friends, some participants chose a meat-based diet to conform to the friends’ group norm. In other words, reducers seemed to juggle multiple diet-related identities that differ in salience across situations and prescribe different behavioural norms. This can explain why meat and dairy reducers prefer a flexible identity, although this is not always helpful for their reduction efforts. This is in line with previous research suggesting that identities can be seen as dynamic and fluid (Brubaker & Cooper, 2000), and suggests that this is very much true for food identities at the initial stages of dietary change, where individuals attempt to establish a new identity to categorise and define their eating behaviour.

More generally, our findings suggest that identity plays a key role in behaviour change, and might warrant more central integration in behaviour change models, such as the COM-B model (Atkins et al., 2017), where identity could be considered as affecting subjective social norms (Gkargkavouzi et al., 2019), or the Transtheoretical Model of Behavioural Change (Prochaska et al., 2008), where identity processes could play a role across phases. Specifically, our findings suggest that social interactions polarised the meat and dairy reduction identities. While previous research shows that flexitarians may be less stigmatised than vegans (Rosenfeld et al., 2020), our findings indicate that flexitarians experienced social resistance to their diets as well. The vast majority of participants spontaneously highlighted that their eating behaviour was a personal choice, and participants found that identifying with a dietary group provoked negative stereotypes and social perceptions. At the same time, participants who found the vegan and vegetarian identity helpful mentioned having rewarding support from those groups. Indeed, identification with social groups that engage in meat and dairy reduction could help increase reducers’ motivation and satisfaction with their dietary change.

In contrast, identification with groups for whom eating meat is normative could hinder reduction efforts by increasing the likelihood of social conflict or encouraging desires to consume meat and dairy. Previous research has shown that vegans tend to score lower than omnivores on the personality trait of agreeableness (Milfont et al., 2021), which might make it easier for them to disengage from prevailing social norms, and radically change their diet. Future research could explore how social interaction impacts meat and dairy reducers’ sense of identification with a dietary group, and the role of personality traits in the management of these social influences.

Our findings on identity mechanisms in behavioural change are also relevant to spillover effects. Previous research has explored, for example, the spillover of sustainable eating between home and work settings (Verfuerth et al., 2019). Environmental identity has been previously shown to mediate the effect of spillover on pro-environmental behaviours (Truelove et al., 2016), while more recently, this effect was not found (Xu et al., 2018). Our findings suggest that the settings that are studied in spillover effects should also be considered for their differing social influences, for example, eating with close family vs eating with colleagues, as the social identities activated by these situations matter. Understanding the social and identity dynamics of the different situations could help us understand exactly when spillover occurs, and when it does not.

In line with habit research, our findings show that forming new habits played a key role in participants’ reduction efforts (Gardner & Rebar, 2019; Lally & Gardner, 2013). Participants’ sense of automaticity increased when they repeatedly avoided eating meat and dairy. This was achieved through behavioural substitutions, through the consistent repetition of the new behaviour (e.g., eating plant-based meals), and through reward (e.g., peer support and self-reward).

In line with the Grounded Cognition Theory of Desire (Papies, Barsalou, et al., 2020; Papies et al., 2017; Papies & Barsalou, 2015), we found that habitual situations prompted mental simulations of eating and enjoying meat and dairy foods, which triggered desire, and at times, led participants to consume these foods (see Papies et al., 2021, 2022). For instance, desires were triggered by internal situations, such as hunger or certain mood states, or by external cues such as the sight or smell of liked foods. Especially when participants felt low in self-control resources, the desire to consume habitual foods was stronger. This suggests that participants’ meat and dairy consumption habits were driven by expectancies of enjoying these foods, which is in line with the perspective that habitual behaviour is goal-driven (Kruglanski & Szumowska, 2020).

Our findings suggest that different related behaviours may support each other and interact in the different stages of behavioural change. As an example, engaging in meat reduction and reducing dairy may be interactive and mutually enforcing. Some participants who reported reducing dairy intake only reported having previously successfully reduced their meat consumption. It is possible that their success in one behavioural change informed their willingness and confidence to engage in another.

Additionally, our findings have implications for behaviour change models, for example, the Transtheoretical Model of Behavioural Change (Prochaska et al., 2008). While the Transtheoretical Model conceptualises the various stages through which an individual progresses during behaviour change as relatively distinct, we suggest that, for the complexity of the reduction behaviour, these stages are not mutually exclusive. In other words, individuals can find themselves in two or more stages at once. Most of our participants were reducing their meat and dairy intake and, therefore, engaging in two kinds of behaviour change, each with various strategies. For instance, some individuals reduced their meat portion size by 50%, only ate dairy when eating out, and planned to adopt meat-free weekdays. Future research would benefit from examining how related behaviours interact, and how this informs the progression between different stages of behaviour change.

Our findings have implications for the development of interventions and policy considerations to support the shift to sustainable diets.

Our research suggests that it may be useful to repeatedly expose the public to reliable information about the role of food in climate change. Participants shared that repeatedly seeking information gradually increased their awareness of the climate emergency and propelled them into action. This is in line with research showing that receiving fourteen daily messages on the environmental benefits of reducing meat intake changed participants’ attitudes (Carfora et al., 2019). Our findings suggest that the motivation to reduce meat and dairy intake may develop gradually, and that the public’s awareness may progressively increase their willingness to reduce or engage in further reduction. A large proportion of participants’ suggestions to facilitate this transition was about the increase of awareness and knowledge in their social circles and the need to normalise meat and dairy reduction. Thus, the reinforcement of knowledge may be important to initiate and support the reduction process.

A greater focus of intervention research on the environment and choice architecture (Arno & Thomas, 2016; Broers et al., 2017; Bucher et al., 2016) could help alleviate the decision conflicts observed in our study. Indeed, participants suggested changes in availability and pricing that could help reduce the effort needed for reducing meat and/or dairy intake. Participants also found choosing from many options stressful, task as they held many motives and considerations that needed to be balanced (e.g., healthy, sustainable, and tasty). Again, changes in food policy, for example affecting taxation, subsidies, and food procurement in the public sector, could support environmentally motivated dietary changes by increasing access to tasty, healthy, and sustainable options.

Interventions supporting the transition to sustainable eating should consider the social identity of consumers, and ways to strengthen meat and dairy reducers’ sense of identification with their dietary groups. This can be done by linking social identity to pro-environmental outcomes (van der Werff et al., 2014) and by promoting pro-environmental ingroup norms (Schultz et al., 2007), and could increase well-being and reduce doubt about dietary change. Social identity may also influence people’s taste perception, such that identity-congruent foods are experienced as tastier (Hackel et al., 2018). Strengthening the dietary identity of consumers might further support them in their reduction experiences.

Additionally, the process of reducing meat intake may differ by gender; for example, social expectations around masculinity may deter men in their behaviour change process. Indeed, recent research has suggested that gender conformity is linked to meat consumption frequencies (Rosenfeld & Tomiyama, 2021). Understanding gender differences in how social influences impact people’s meat and dairy consumptions can help strengthen efforts to improve the sustainability of eating patterns. Therefore, future work could consider the individual challenges across genders as to manage these and potentially strengthen their dietary identities with their decisions to reduce meat or dairy.

Policies to encourage sustainable and healthy eating must consider the taste, availability, and affordability of plant-based foods. Our findings suggest that small dietary changes helped participants stay engaged in their reduction efforts, as they led to experiences of success and new habits. Thus, attractive meat replacements continue to be important. Our participants also found that taste was a key factor in their efforts. In line with this, research has shown that labelling plant-based foods by emphasising taste and reward may be an effective and low-cost strategy to increase the appeal of plant-based foods among habitual meat eaters (Papies, Barsalou, et al., 2020; Turnwald & Crum, 2019) .

Finally, our participants reported that the lack of availability and affordability of attractive meat and dairy alternatives hindered their reduction efforts. Increasing the likelihood of people choosing plant-based foods is important and can be achieved by changes to the choice architecture, for example increasing availability (Garnett et al., 2019), using appealing language for plant-based foods (Papies, Johannes, et al., 2020; Turnwald et al., 2019), introducing financial incentives for sustainable alternatives (Willett et al., 2019), or shifting subsidies from animal agriculture to sustainable alternatives (Abadie et al., 2016).

A strength of our study lies in the transparency through pre-registration, and credibility strategies such as bringing in different perspectives, peer debriefing, reflexivity, and negative case analysis, which strengthened the robustness of our analysis process. Additionally, there are benefits to both researchers and participants when using qualitative surveys. Qualitative online surveys offer rich data from a broad representation of individuals and experiences to explorative research. It provided us with diverse perspectives of the reduction experience. This diversity is useful when researching an underexplored area. At the same time, qualitative surveys offer participants full control over their research participation and bypass the traditional the power-dynamics of the researcher and the researched that takes place in qualitative interviews. In addition, participants’ responses often provide more focused and targeted data as opposed to data from interviews (Braun et al., 2020; Braun & Clarke, 2013).

This study is not without limitations. First, online data collection risks excluding individuals from disadvantaged groups in society. Secondly, and as self-control theories guided our question generation, we acknowledge that our findings are limited in that participants were guided by the concepts introduced in the survey questions. In other words, it is possible that participants’ responses referenced constructs such as habits and social norms because we asked about them, and that other aspects of the reduction process were less likely to be shared as a result. Future research may address this issue with either more open questions, or quantitative measures.

Despite the broad representation of experiences in our study, our findings remain contextualised within some boundaries. For example, our sample was predominantly female. It is possible that the processes involved in reducing meat and dairy intake differ between genders, as implied by one participant’s comment on toxic masculinity. Additionally, our UK sample may also limit the transferability of our findings to different Western and non-Western cultures. Finally, only four participants of our sample were reducing only dairy intake. It is possible that the challenges and barriers for this group of reducers differ from the majority of the participants. However, our aim was not to ensure representativeness but to explore the diverse lived experiences of reducing one’s meat and/or dairy intake. Future research would benefit from exploring demographic (e.g., gender) and cultural differences, as well as differences between meat and dairy reducers in self-control, identity, and the social influence processes that affect their experiences of changes in eating behaviour.

This study has developed a rich picture of the experiences of a sample of UK residents reducing their meat and/or dairy intake for environmental reasons. We found that reducers often experienced conflict between different desires, habits, and motives, and needed self-control resources to manage them. Small and comfortable changes were experienced as preferred and effective strategies to maintain the reduction behaviour. However, social challenges and unclear identities hampered dietary change. Interventions should address these processes to support a wide-spread transition to sustainable diets.

All authors have approved the final article.

Contributed to conception and design: LW, EP

Contributed to acquisition of data: LW

Contributed to analysis and interpretation of data: LW, KB, EP

Drafted and/or revised the article: LW, KB, EP

Approved the submitted version for publication: LW, KB, EP

We would like to thank the members of the Healthy Cognition lab at the University of Glasgow for their useful feedback on an earlier version of this manuscript.

This research was funded by the College of Science and Engineering at the University of Glasgow, and by Research Grant ES/T011343/1 from the UK Economic and Social Research Council. For the purpose of open access, the author has applied a Creative Commons Attribution (CC BY) licence (where permitted by UKRI ‘Open Government Licence’ or ‘Creative Commons Attribution No-derivatives (CC BY-ND) licence may be stated instead) to any Author Accepted Manuscript version arising’.

We declare that we have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

All materials can be found on the Open Science Framework (OSF; https://osf.io/vuhsy/).