Willingness-to-pay for reduced carbon footprint and other sustainability concerns relating to pork production A

,


A B S T R A C T
There is growing societal concern about the impact of animal production on climate and other dimensions of sustainability.Continuing development of production in response to this will require the industry to know how much consumers are willing to pay for more sustainable pork, and how they prioritize different sustainability dimensions.
In this 2022 questionnaire-based study, we investigated willingness-to-pay for more sustainable pork among consumers in Denmark, Germany, the UK and Shanghai, China.We examined how respondents prioritised lower climate impact in comparison with four other dimensions of sustainability: improved animal welfare, decreased use of antibiotics, freedom from Salmonella and other harmful bacteria, and the avoidance of rain forest depletion in the production of pig-feed.
While many respondents were found to be willing to pay a price premium for more sustainable pork, only 10% (approx.)were prepared to pay more than a 20% premium.In all four countries, lower climate impact was among the least important reasons for paying a price premium.In the Western countries animal welfare was the most important reason while food safety was the most important reason in China.The most frequently stated reason for prioritizing animal welfare over climate impact was the belief that paying a price premium is necessary to make a difference for the animals while climate impacts can be handled by other means.An important insight gained is that the current focus on reducing climate impacts of pig production should not blind the relevant stakeholders to the importance of better animal welfare and food safety.If they lose sight of the latter, stakeholders will fall out of step with the priorities that many consumers currently have.

Introduction
During the past five decades there has been a growing focus on various challenges in animal production falling under the now familiar umbrella of 'sustainability'.A relatively recent study of sustainability in cattle production is Segerkvist et al. (2020).The latest studies which, like the present paper, focus on pig production include Andretta et al. (2021) and Ruckli et al. (2022).
Historically, the first issue of public concern to affect pig production was animal welfare.This issue surfaced in the mid-1960s (Sandøe and Christensen, 2023) and is still very much with us, also from a scientific point of view (Pedersen, 2018).The importance of animal welfare, as a sustainability issue (typically, viewed as part of social sustainability) is now recognised (Bonneau et al., 2014;Torpman and Röcklinsberg, 2021).The second issue to emerge concerned the environmental effects of pig production.Here the focus was on the pollution of the local environments by excessive nutrient loads.However, although this issue has had huge policy implications in some parts of the world, it has been either underestimated or not fully comprehended by consumers (Hartmann and Siegrist, 2017;Vanhonacker et al., 2013).A third issue, revolving around food safety, both in terms of zoonotic risks (Fosse et al., 2009) and anti-microbial resistance (Monger et al., 2021), has also been growing in prominence for some time now, with a special focus on intensive pig production (Gilbert et al., 2021).
More recently, concerns about climate change have been linked to animal production in the wider public debate.Agricultural production is estimated to be responsible for between 19 and 29% of anthropogenic climate change (Nielsen et al., 2022;Vermeulen et al., 2012), the largest factor in which is livestock production (Steinfeld et al., 2006;Crippa et al., 2021;Xu et al., 2021).Lately, the protection of the natural world from the risk to biodiversity degradation presented by feed production, and notably soy production linked to the clearing of rain forest and land use in general, has become a matter of escalating concern (Crenna et al., 2019;Ortiz et al., 2021).Other aspects of sustainability include economic (Malak-Rawlikowska et al., 2021) and social issues, with a special emphasis, where the latter is concerned, on working conditions for employees and conditions more generally for local communities (Zira et al., 2020).
Concerns about the sustainability of pig production have so far been an issue mainly in the richer Western countries, but they are growing (at least, as far as some of the aspects are concerned) in developed Asian countries like Japan and South Korea (Shimokawa, 2015) and in emerging economies such as China and Brazil (Vilas-Boas et al., 2022;Zhuo and Ji, 2019).
Often, particularly in affluent Western countries, legislation is in place to ensure at least minimum standards around the environmental and animal welfare issues raised by pork production (Sandøe and Christensen, 2023).More recently, China has introduced minimum requirements for food safety (Guo et al., 2019) and it has also introduced a growing body of legislation on environmental impacts (Chen et al., 2022a).
However, legislation can at best ensure a 'decent' minimum.It will always lag behind the ambitions and demands of the most concerned, or conscientious (and often better-off) pork consumers (Sandøe and Christensen, 2023).Hence, various market-driven initiatives have been developed in which either products with a special focus on various aspects of sustainability are sold with a label, and typically also at a price premium, or specific retailers or food service companies guarantee a certain standard of sustainability for all products sold.An overview of initiatives used to improve animal welfare which could well be relevant for bringing about improvements in other sustainability-related issues is provided in Christensen et al. (2019).
For many years there have been market-driven initiatives designed to improve animal welfare in pig production, and these make a significant difference in some European countries (Sandøe et al., 2020).Also, some pork is produced under an organic label, with a focus on food safety (in terms of pesticide-free production and longer time to slaughter after an animal has been treated with antibiotics) and nature protection as well as animal welfare (van Wagenberg et al., 2017).Labeling of climate friendly food is in its infancy, and there is limited consumer knowledge of carbon footprint labeling (Rondoni and Grasso, 2021), but the idea behind such labeling has been evaluated positively by consumers (Edenbrandt and Lagerkvist, 2021).
Generally, each initiative to improve sustainability involves added production costs, and the willingness of consumers to pay for these has its limits, of course (Vanhonacker and Verbeke, 2014;Denver et al., 2022).Some traits are conflictingnotably, animal welfare and lower climate impact might present consumers who are motivated by sustainability with a dilemma (Sonntag et al., 2018;Olsen et al., 2023).A recent review by Bussel et al. (2022) highlighted the importance of understanding consumers' perceptions of sustainability to support a transition towards a more sustainable food system.Given this, there is a need for specific knowledge about how consumers value and prioritize between different dimensions of sustainability.
So far, little empirical work has been done in this area (Goddard et al., 2019;Sato et al., 2017).We conducted the present study, examining the potentials of market-driven sustainability in pig production and pork consumption, to help fill this research gap.To support continued development of pork production for sustainability we aimed to find out how much consumers are willing to pay for more sustainable pork in multiple sustainability dimensions.
Animal production involves substantial emissions of greenhouse gasses (GHG).Production of red meat, notably from cattle and sheep, are the worst offenders here, while pork production is much less of a problem.In pork production, about half the climate impact is from feed production, so there seems to be a potential to develop production in a more climate-friendly direction by optimizing feeding strategies (Dorca-Preda et al., 2021).To this end, Mackenzie et al. (2016) found a 17% reduction in climate impact of pork production could be achieved by optimizing feed ration with respect to climate impact instead of lowest possible cost.Another climate mitigation strategy was investigated in Pexas et al. (2020), who found that using manure for biogas could reduce climate impact by 9%.
Political focus on the climate impact of food consumption is increasing.Studies have indicated that the consumers' interest in food with low climate impact is not increasing in the same way, however, although younger people do seem to be more concerned about climate change (McCright et al., 2016).Skeiryt ė et al. ( 2022) recently conducted a study using almost 28,000 respondents spread across the EU.The results revealed that younger people had a higher level of knowledge about challenges created by climate change than older people did.They were also more likely to take personal responsibility by adopting more climate friendly behavior.
Our overarching research question concerns the extent to which low climate impact, specifically, is a driver of willingness-to-pay (WTP) a price premium for improved pork in comparison with other dimensions of sustainability.To address this question, we surveyed consumers in four countries (Denmark, Germany, the UK, and Shanghai, China).Some of these are affluent Western countries, two of which, Denmark and Germany, are major pork exporters, and one of which, the UK, is net importer of pork.China is a large, emerging economy with rising per capita disposable income and a growing middle class, and the city of Shanghai is the commercial and financial center of China.We included the following dimensions of sustainability-enhanced pork (hereafter referred to simply as 'dimensions') in the analyses: production with lower climate impact, improved animal welfare, meat with guaranteed freedom from zoonotic bacteria, decreased use of antibiotics and no rain forest depletion related to the production of the feed.We aimed to answer the following research questions: RQ1: How much are consumers in each of the four countries willing to pay for more sustainable pork?RQ2.How important is climate impact in comparison with the four other sustainability dimensions of pork and how do the consumers' priorities in relation to these dimensions depend on their WTP for more sustainable pork, their age and their interest in climate impact?RQ3.Faced with the dilemma between animal welfare and climate impact, which dimension would consumers in the four countries prioritize when purchasing pork?
The data and the statistical methods used to investigate the research questions are described in Section 2. The results of the analysis are presented in Section 3. In Section 4 we discuss the findings of the study, and in Section 5 we set out our conclusions.

Data and materials
The questionnaire was initially developed in Danish, and then translated into German and English by professional translators with a very good knowledge of the source language (Danish), and of German and English, respectively.The translations were subsequently reviewed by the authors to ensure there was similar meaning across versions.Chinese contractors CRT carried out the translation from English to Mandarin for the Chinese questionnaire.Through personal networks we were also able to find two native Chinese persons to validate the translations.CRT was also responsible for distributing the questionnaire in their consumer panel.In the three European countries, data were collected by the company Norstat, which hosts and maintains panels of consumers in all three countries.The study was approved by the Research Ethics Committee of Science and Health at the University of Copenhagen.
The questionnaire was tested in a pilot study involving around 100 respondents in each of the four countries in November 2021.Results from the pilot study were used to adjust the questionnaire's comprehensibility and length.The main survey was carried out from mid-January to start of February 2022 and included around 1500 respondents in each country.In the European countries, respondents were recruited so that the country-specific samples were aimed to be representative of the adult population in terms of gender, age and region.In China, respondents were recruited from among residents of Shanghai, and the sample only aimed to be representative in terms of gender and age in Shanghai.There were no other inclusion criteria in relation to the data collection.Hence, respondents who did not eat meat were included in the samples.To ensure that the results were representative for the target population regarding gender, age and region in the three European countries, weighted results were reported in all univariate and bivariate analyses, and in inferential statistics.The weight variables adjust the data so that they reflect the population census in the three European countries regarding age, gender, and regions (NUTS2).Sociodemographic characteristics of the respondents in the three European countries are shown as unweighted shares of the population in Appendix 1 and compared with the population census.The sample weights are shown as table notes in Appendix 1. Unweighted data were used for Shanghai, China, as weights were not available.Informed consent according with national requirements was collected and handled by Norstat/CRT, and the data were received in pseudonymised form.All data management and statistical estimations were carried out using the software SAS 9.4.
The respondents were given the following preliminary information: 'When answering the questions, think about what you usually consider and do in your everyday life when you buy or eat meat.Try to imagine a new everyday life where all Covid-19 restrictions are lifted.The study is about meat from all animals with two or four legs (poultry or mammals) -animals like fish or shellfish are not part of the study and are thus not to be considered when answering the questions.' In the survey, the five dimensions of sustainability of relevance to pork were chosen with reference to a literature review, by consulting a wide range of Danish stakeholders (Christensen et al., 2021), and with a view to formulating questions and response options that were communicable to consumers across the four countries.One part of the questionnaire focused on the dimensions in general terms.Other parts focused on specific improvements within individual dimensions.The specific levels of improvement of the five dimensions were set on the basis of realistic but still somewhat ambitious targets for pig producers.Descriptions of the five dimensions and the specific levels of improvement used in some of the questions are listed below: • Improved animal welfare: As the vast majority of pigs worldwide are produced in indoor production systems, and since the intention of the study was to analyze large-scale market-based changes, the animal welfare improvements were chosen to be in line with existing welfare labels on pork from indoor pig production in Denmark (Sandøe et al., 2020).Specifically, the improvements analysed in the questionnaire consisted of: indoor conditions in which the pigs have 30% more space and extra straw in comparison with Danish legislative requirements, and in which sows are always loose.• Lower climate impact: Based on the range of existing practices, a relatively ambitious but easily communicated improvement was selected for the questionnaire: GHG emissions from the pigs are reduced by 50% relative to average emissions in Danish pig production.
• Decreased use of antibiotics: Use of antibiotics in animal production and the related risk of antimicrobial resistance is one of the big sustainability challenges worldwide (Llor and Bjerrum, 2014), and it has been a matter of debate in Denmark, particularly in relation to pig production.Based on the range of existing practices, a realistic and easily communicated, but still rather ambitious, improvement was selected.Existing niche production of 'pigs raised without the use of antibiotics' together with the strict requirements on the use of antibiotics in organic production (Olsen et al., 2023) indicate that significant reductions in the use of antibiotics are feasible.The level of improvement analysed in the questionnaire was: 50% decrease in the antibiotics used in production relative to average use in Denmark.
• Guaranteed freedom from harmful bacteria: Food risks from zoonotic bacteria are known to have large societal costs in terms of hospitalization, lost productivity and reduced quality of life (Bernstein et al., 2022).The level of improvement analysed in the questionnaire here was: pork is guaranteed to be free of Salmonella and other potentially harmful zoonotic bacteria.• Protection of rainforests: Animal production, and in particular feed production, requires a lot of land.There are increasing concerns about the protection of rainforests linked to the safe-guarding of biodiversity, the reduction of GHG emissions and the protection of local cultures (Phelps et al., 2017).National legislation and market-based initiatives aim to bring about increased use of feed sourced in ways that do not involve the cutting down of rainforests.They also target reduced use of soy in pig feed.The formulation of the land use dimension for the questionnaire was: no rainforest is cut down for the cultivation of soy to be used in feed for the pigs.

Method for RQ1
Consumers' WTP for more sustainable pork (RQ1) was investigated by asking the respondents to state their WTP for a bundle of all five sustainability dimensions.The payment card method was used.This method asks respondents to pick a WTP level, or an interval, from a list of given values (see e.g.Yu et al., 2014).
To reduce difficulties in comparing WTP across the four countries due to the substantial differences in price levels, we followed the approach used in Denver et al. (2022) and presented price premiums as percentage increases above the local price of standard pork.To keep things simple, the pork product was not specified further (e.g. as regards cut).Thus, we implicitly assumed that the respondents were willing to pay the same percentage price premium for expensive as they were willing to pay for cheaper cuts.As we focused on percentage price premiums, we avoided the potential bias that would arise from informing respondents about a reference price of standard pork.The following price premium intervals were presented to the respondents: 0%, 1-2%, 3-5%, 6-10%, 11-20%, 21-50%, 51-100%, and more than 100%, plus a 'don't know' option.In order to facilitate presentation of the results, we subsequently grouped the WTP categories into four categories: No WTP (0%), Low WTP (1-5%), Medium WTP (6-20%) and High WTP (above 20%).For each country, we present descriptive statistics of the shares of respondents who ate pork and how much they stated they were willing to pay for more sustainable pork.

Method for RQ2
In RQ2, we compared the importance of reduced climate impact as driver of positive WTP for more sustainable pork with the other four dimensions.We used a scoring system in which respondents with a positive WTP (i.e.> 0%) distributed 100 points between the dimensions according to the stated subjective importance of the dimensions.The precise wording of the information provided to the respondents was as follows.
'You have stated that you are willing to pay more for the improved pork.Now, please state how important the individual improvements are to you.You have a total of 100 points to divide between the improvements (the more points, the more important an improvement is to you).For instance, you can give all 100 points to one improvement and 0 points to the others.Or, you can give 2 improvements 50 points each and 0 points to the rest.Or, if you think that all improvements are equally important then you can give them 20 points each, etc. Please record your points in the empty column.' The scores given by the respondents were investigated for four different groups of respondents: Overall priority of dimensions: For each of the five dimensions, mean scores were calculated for all respondents, in each country, who stated a positive WTP for more sustainable pork.
Priority of dimensions among respondents with high WTP: For each of the five dimensions, mean scores were calculated for respondents, in each country, with high WTP (as opposed to those in one of the two other groups: low or medium WTP).This enabled us to analyze whether respondents with high WTP and those with medium or low WTP prioritised the dimensions differently.The argument for separating out respondents with high WTP was that the additional costs of producing the five dimensions are expected to be at least of the magnitude that those with high WTP are willing to pay (e.g.see Olsen et al. (2023) for Danish reference figures).
Priority of dimensions among young respondents: For each of the five dimensions, mean scores were calculated for respondents, in each country, in the youngest age group of 18-34 years (as opposed to those in other age groups: 35-49, 50-64, 65 and older).The reason for analysing preferences among different age groups was that, as mentioned in the introduction, studies have found younger people to be more concerned about the climate than the rest of the population.By separating out the group of younger respondents, we were able to investigate the extent to which this group would also prioritize lower climate impact to a larger extent than older age groups in the context of pork consumption.
Priority of dimensions among respondents with concerns about climate change: For each of the five dimensions, mean scores were calculated for respondents, in each country, who had stated that climate is essential to them when they buy meat.This enabled us to identify to what extent low scores assigned to climate impact as one of the drivers of a WTP for more sustainable pork would simply reflect a lack of interest in climate impact when buying meat.To identify concerns about climate impacts we used one question from a questionnaire battery where respondents were presented with 18 production-related or product-related characteristics, and asked to evaluate the importance of them: What is important to you when you choose to buy one kind of meat instead of another?One of these characteristics was 'The meat is produced with little impact on global climate' (see Appendix 2 for an overview of all 18 characteristics).Altogether six response categories were available: Not important at all for my choice; Somewhat important; Important; Very important; Absolutely essential for my choice; Don't know.Those who chose 'Absolutely essential for my choice' were categorised as 'climate essential', and those who chose any of the other response categories were categorised as 'climate not essential'.
The results addressing RQ2 are presented as mean scores assigned to each dimension together with 95% confidence intervals among all pork consumers with a positive WTP, and within the groups defined above (i.e. size of WTP, age groups and interest assigned to global climate).Moreover, as the relative importance of lower climate impact in comparison with other sustainability dimensions is central in this study, we tested pairwise differences between scores that each group of respondents assigned to lower climate impact relative to each of the other four dimensions.Plots indicated that these pairwise differences were approximately normally distributed and we therefore used paired t-tests to identify potential statistical differences between scores.Note that in order to limit the number of tests and increase readability, we limited the analyses by only testing the differences within a group of respondents and thereby we did not test e.g. for differences in scores assigned to the dimensions across different groups of respondents.Equally, as we focused on the relative importance of climate impact, we only tested for pairwise differences involving climate impact which implied that we did not test for potential differences between other dimensions e.g. between the mean scores assigned to animal welfare and to reduced use of antibiotics.

Method for RQ3
RQ3 concerns the trade-off between two of the five dimensions, and in particular how likely respondents were to choose pork produced with lower climate impact rather than pork produced with improved animal welfare.We focused on lower climate impact and better animal welfare for two reasons: there is a potential tension, or dilemma, between the two dimensions, and animal welfare, as we mentioned in the introduction, is known to be an important issue for consumers in three of the four countries.The formulation of the question used in the analysis was as follows: 'Imagine that you must choose between two types of pork.The price is the same and the product is slightly more expensive than the cheapest standard pork.The first type of pork is produced with improved animal welfare whilst the second type of pork is produced with lower impact on the global climate.Imagine that you must choose between improving the welfare of the pigs and reducing the climate impact.What would you choose?' The response categories offered were: Definitely lower climate impact; Probably lower climate impact; Probably improved animal welfare; Definitely improved animal welfare; Neither -I choose the cheaper standard pork; I am not able to chooseboth aspects are very important; Don't know.
In a follow-up question, the respondents who had probably or definitely prioritised improved animal welfare over lower climate impact were asked why they did so.The following response categories were provided: Pork production has, as far as I know, little impact on the global climate; I think that I can make a real difference to the welfare of pigs through my choice of pork; I do not think that I can affect climate change very much through my choice of pork; I prefer to affect the global climate by other means than through my choice of pork; Other; Don't know.
The data set does not include information about respondents' reasons for choosing lower climate impact over improved animal welfare.Moreover, only respondents with a positive WTP for more sustainable pork were included in the analyses.Analyses of data used to address RQ3 were based on descriptive statistics.

Results on RQ1
Table 1 shows that approximately 10% of the respondents in Denmark and China, and around twice that in Germany and the UK, do not eat or buy pork.The questionnaire also contained information about whether the respondents were eating or buying any type of meatnot just pork.We found a similar pattern for eating or buying meat in general as we did for pork.More specifically, in comparison with China and Denmark, twice as many respondents in the UK and Germany did not eat or buy any type of meat (around 5% in the UK and Germany as against 3% in China and 2% in Denmark).
Respondents with a positive WTP for more sustainable pork comprised around 75% of the respondents in Denmark, Germany and China, and around 60% of the British respondents.In all four countries, the groups of respondents who were willing to pay a medium price premium of 6-20% tended to be the largest.They made up around 45% of the respondents in Denmark, Germany and China, and 28% in the UK.The second largest group of respondents consisted of those willing to pay between 1% and 5%.These made up a quarter of the respondents in Denmark and Germany and one fifth of the respondents in the UK and China.The group stating the highest WTP (more than 20% price premium) included between 8% and 13% of respondents.The remaining 10-20% of the respondents either had a WTP=0 or answered don't know to the question about WTP.Information on the socio-demographic characteristics of the respondents in the three groups (eating no pork, eating pork and having a WTP=0 for improved pork, and eating pork and having a WTP>0), in each country, is presented in Appendix 1.

Results on RQ2
RQ2 investigates the relative importance of the five dimensions as drivers of WTP for more sustainable pork.Table 2 shows the mean scores allocated to each of the five dimensions (among respondents with a positive WTP).With mean scores of around 26-28 points (out of 100), respondents in the three European countries considered improved animal welfare to be most important.Respondents in China, on the other hand, assigned the highest scores to pork produced with decreased use of antibiotics and pork that is guaranteed to be free of potentially harmful bacteria.Hence, we found that improved animal welfare ranks highest on average in the European countries, and that the two food-safety dimensions rank highest on average in China.
The mean scores for lower climate impact were between 16.1 for Germany (95% CI 15.4;16.8)and 18.3 for Denmark (95% CI 17.5;19.1).We found differences in the relative scores for the dimensions across the European countries.In Denmark, lower climate impact received lower scores than improved animal welfare, but the climate impact scores were similar to those given to the two food-safety dimensions and protection of rainforests.In Germany, lower climate impact received lower scores than the other four dimensions, and in the UK both improved animal welfare and guaranteed freedom from harmful bacteria were more important than lower climate impact.Chinese respondents considered both improved animal welfare and the two food-safety dimensions to be more important than lower climate impact.
Table 3 indicates the extent to which priorities regarding the five dimensions depend on WTP for more sustainable pork.All climate impact scores were in the range of 15.8 to 19.3.Again, we found interesting trends in and across the countries.First, food safety tended to be ranked as less important by respondents with higher WTP than it was by respondents with a low or medium WTP.Taking Denmark as an example, the score for guaranteed freedom from bacteria was 19.5 (95% CI 18.5;20.4)among respondents with a low or medium WTP and 15.5 (95% CI 13.5;17.6)among respondents with high WTP.Again, taking Denmark as an example, the t-tests did not reveal differences between the scores allocated to lower climate impact and guaranteed freedom from bacteria by respondents with a low or medium WTP, but they suggested the opposite among the respondents with high WTP.Here, lower climate impact was assigned a higher score than guaranteed freedom from bacteria.Second, in Germany and China, improved animal welfare was more important to respondents with high WTP than it was to the other respondents.
Differences in scores for the five dimensions categorised according to age groups are shown in Table 4.In the European countries, all age groups gave improved animal welfare higher scores than lower climate impact.Unlike those in the corresponding older age groups, younger respondents in Denmark and Germany considered lower climate impact more important than, or as important as, decreased use of antibiotics, guaranteed freedom from bacteria and protection of rainforests.In the UK, we found very little differences across age groups: improved animal welfare and guaranteed freedom from bacteria were more important than lower climate impact for all age groups.In China, all age groups considered decreased use of antibiotics and guaranteed freedom from bacteria to be more important than lower climate impact.However, the respondents in the youngest age group were distinctive in also considering improved animal welfare as more important than lower climate impact.
The relationships between stating that climate impact is an essential factor when buying meat and scores assigned to the five sustainability dimensions for pork are shown in Table 5.In all four countries respondents who did not indicate that climate impact was essential when buying meat gave higher scores to improved animal welfare, and often also to food safety, than they did to lower climate impact.For respondents who considered climate impact essential when buying meat, we did not find any significant differences between scores assigned to lower climate impact and those assigned to the two food-safety dimensions.Moreover, for this group of respondents in Denmark and China there were no differences between improved animal welfare and lower climate impact, while animal welfare in fact received higher scores than lower climate impact in Germany and the UK.

Results on RQ3
Table 6 shows respondents' choices when confronted with the choice between pork with improved animal welfare and pork with lower climate impact (among respondents with a positive WTP).The first thing to notice is that Table 6 suggests around 20% of the respondents with a positive WTP in all countries considered both dimensions to be important and thus were not able to choose between them.This group would

Table 1
Categorization of respondents in the four countries according to whether they eat pork and the price premium they were willing to pay for more sustainable pork (shown as column% within countries).clearly face a real dilemma in a food shopping situation where they cannot have both improved animal welfare and lower climate impact.Second, in the European countries, the vast majority of the respondents with a positive WTP preferred pork with improved animal welfare over pork with reduced climate impact.The Chinese respondents appeared not to have a similarly strong preference for animal welfare and tended to be more evenly distributed than respondents in the other countries as regards whether they found animal welfare or climate impact most important.
Fig. 1 provides details of the respondents' reasoning when they prioritised improved animal welfare over lower climate impact.In all four countries, the reason most often given by the respondents for Note: Based on weighted data.Only respondents with a positive WTP (willingness-to-pay) are included in the analysis.Only statistical differences between low climate impact scores and the other four dimensions are reported, where a +/-sign indicates that the score differs positively/negatively (at 0.05 significance level) from the score assigned to lower climate impact.Number of respondents with WTP>0 using weighted data: Denmark = 1166; Germany = 1100; the UK = 927; China = 1163.

Table 3
Mean scores of the five sustainability-related dimensions by country categorised according to WTP (95% confidence interval in parenthesis).choosing animal welfare over climate impact was that it made a difference to the welfare of the pigs.In each country, only small shares of the respondents chose improved animal welfare over lower climate impact for the reason that 'pork production has, as far as I know, little impact on global climate'.

Discussion
This study investigated WTP for more sustainable pork, breaking the concept of sustainability down into five dimensions.The discussion is structured around the three research questions and concludes with reflections on the study's limitations.

Discussion of RQ1
Regarding RQ1, on consumers' WTP for more sustainable pork, we found that in the UK and Germany 20% and 16%, respectively, of the respondents stated that they did not eat or buy pork.These shares are approximately twice as high as the corresponding shares in Denmark, where 7% did not consume pork, and in China, where 10% did not consume pork.Similar results were found for eating or buying meat in general, in that, compared with Denmark and China, twice as many respondents in the UK and Germany stated that they did not eat or buy meat.Hence, the surveyed British and German consumers did not seem to avoid pork specifically to a greater extent than consumers in the other two countries.
The study showed that only around 10% of the respondents in each country stated that they were willing to pay a price premium of more than 20% extra for more sustainable pork.This is in line with what we know from the existing markets for pork and other meat produced with an added focus on animal welfare in different European countries (Benchmark for pigs, 2022) and it suggests that although demand for improved animal welfare (and sustainability) exists it may be limited.Thus, in Denmark, the market shares for pork from indoor production with improved welfare were around 7% in 2021.Shares for outdoor production, likewise, were around 7%.In Germany, around 40% of consumption is from indoor production systems with welfare requirements slightly above those set out in national legislation, and around 7% is from outdoor production systems.In the UK, around 2% of pork consumption is from pigs with improved indoor conditions above the Red Tractor certification scheme and approximately 10% is from outdoor production systems.According to the RSPCA (2021), 23% of pigs in the UK were RSPCA certified in 2021, the vast majority of these being from outdoor production systems.However, the UK imports around 50% of its pork, and this comes from animals in indoor production systems: hence, the consumption share of outdoor bred pigs is smaller than the production share (Sandøe and Christensen, 2023).
Our findings are in line with a meta-analysis reported in Li and Kallas (2021).This examined 80 quantitative studies and found an overall average WTP a premium for sustainable food of 29.5%, with a slightly lower WTP for meat and dairy.The meta-analysis also found that the WTP for organic food products was higher than the WTP for other foods labelled as sustainable.

Discussion of RQ2
Regarding RQ2, on how consumers in the four countries prioritize the five dimensions, we identified some differences between the scores allocated to the five dimensions, but the average scores in the various groups investigated were all between 15 and 30 (out of 100).This relative lack of variability may reflect the fact that the respondents found it difficult to interpret some of the dimensions, or hard to prioritize and allocate points among them.It is also possible of course that the respondents actually considered all of the dimensions to be more or less equally important.That would imply that consumers would be willing to pay a price premium for sustainability improvements but are indifferent about the specific kinds of improvement.In future studies it would be interesting to apply a survey design that does not allow respondents to give all dimensions an equal score, thereby forcing them to prioritize between the dimensions.Alternatively, it would be helpful to dig into reasons for why some of the respondents gave all dimensions equal scores.
Our choice of method was inspired by Bussel et al. (2022), who conclude that consumers find it difficult to assess sustainable food production.In order to reduce the cognitive burden for the respondents, we divided the valuation process into two sub-tasks: first, stating a WTP for sustainability, and second, ranking the sustainability-related attributes.We have not found other studies using this simplified approach.Consequently, we see our work as a first step in the difficult field of eliciting consumer priorities among sustainability-related improvements in pork and in other similar products, and we hope that future studies will further develop the approach we have introduced.
The results showed that, on average, respondents in the three European countries considered improved animal welfare to be the most important of the five proposed reasons for paying a price premium.Even among the groups of respondents that we had expected to be particularly concerned about climate issuesi.e.young respondents and respondents who had stated that they think about the climate when buying meatwe often found that lower climate impact was less important than improved animal welfare in pork purchasing decisions.
We found that lower climate impact was more important for the younger respondents than for the older respondents in Denmark and Germany.A study by Jürkenbeck et al. (2021) investigated climate change awareness among young consumers (15-29 years of age) in Germany, asking how this awareness was reflected in dietary habits.Half the respondents sampled were highly aware of climate change, and the authors found a positive relation between climate awareness and having a climate-friendly diet.In our study, we did not examine variation in climate awareness within different age-groups of respondents.This issue would be an interesting topic for a follow-up study.The review article by Corner et al. (2015) focused specifically on younger people's engagement with climate change although not in a food consumption context.They found that in many ways, the views of the adult population on climate change were also held by the younger generation.For example, they found relatively high levels of climate change concern among the younger age groups (in some cases higher than those in older age groups), but that climate was rarely the top priority among younger people or the general population.
Among the respondents from China in our study the picture was different, since food safety, for them, was the main reason, of the five presented dimensions, for paying a price premium for improved pork.The high priority given to food safety among Chinese pork-eating respondents is in line with Lai et al. (2018), who compared preferences among 480 consumers in Shanghai and Beijing.They found that the consumers generally had higher WTP for food safety than they had for the other attributes, and they linked the strong focus on food safety with a number of food safety scandals in China.Along similar lines, Xu et al. (2019) used a sample of 316 Chinese pork-eating consumers in a large city in the Jiangsu province to elicit WTP for several pork-characteristics (traceability, lean meat essence testing, animal welfare, appearance, and price).They found that the respondents attached most importance to price, followed by traceability and then animal welfare.Finally, Lin--Schilstra et al. ( 2022) compared consumer views on pork in ten EU countries and four non-EU countries (USA, Russia, South Korea and China) based on survey data from 2013.They found that Chinese consumers saw food safety as the most important reason (together with health) guiding meat choices, and that food safety was also more important for Italian and South Korean consumers than it was for consumers from the other countries.
Generally speaking, cultural differences influence practices around, and attitudes towards, foods (Jeong and Lee, 2021).Chinese citizens have a historical and cultural background in Confucian religion and are S. Denver et al. more traditionally and hierarchically oriented (Schwartz, 2006).They are also, in general, more concerned with proximate, material values than citizens in other countries, especially Western democracies, where longer-term, post-material values are more widespread (Nový et al., 2017).The kinds of value orientation that prevail in China seem to leave less room for concerns about animal welfare and the environment, and this may explain the relatively higher levels of concern about food safety among the surveyed Shanghainese.Still, the younger Chinese generations are increasingly prioritizing post-materialist values (Chen et al., 2022b), so a change could be on the way also in China.

Discussion of RQ3
RQ3 concerns trade-offs between animal welfare and lower climate impact.Our survey showed that respondents in all four countries considered improved animal welfare to be more important than lower climate impactbut in China, this was only the case for respondents with a high WTP.Lai et al. (2018) found that consumers in Shanghai, which appears to be more influenced by Western values than Beijing, showed more concern about animal welfare and environmental issues than consumers in Beijing.Hence, we cannot directly generalize our results to other parts of Chinanot even to Beijingand the suspicion must be that animal welfare and climate concerns would probably have had an even lower priority if we had surveyed other regions of China.
In light of the substantial focus on climate change in recent years (Carlsson-Kanyama and González, 2009;Filho et al., 2022) the higher priority of animal welfare is interesting.In our study, the large majority of respondents who opted for improved animal welfare explained this priority mainly by saying that they believed they could make a real difference for the animals through their purchasing decisions.Whether this priority will be retained over time remains to be seen.But we speculate that it will endure for quite some time.Partly, because in parallel to the increased focus on mitigating climate change there also tends to be an increased concern for animal welfare in livestock production (Sandøe and Christensen, 2023) and partly because animal welfare is now recognized as a sustainability issue (Torpman and Röcklinsberg, 2021) and has thereby gained visibility on the sustainability agenda.
Another reason that climate impact was considered less important than animal welfare when purchasing pork, and potentially could be so for some time, seems to be that there are many alternative ways to tackle climate change that are not related to food purchases (Moran et al., 2020;Thøgersen, 2021).There are indeed many ways to combat climate change.In the EU, citizens tend to think that larger actors, such as national governments, industrial sectors, and individual businesses, are the main actors responsible for tackling climate change (Eurobarometer, 2019).Relatively few European consumers mention reducing the carbon footprint of their food purchases as a means of taking personal action to reduce climate impact.Much more popular actions include reducing food waste, recycling, cutting down on disposable items, and using environment-friendly transport (Eurobarometer, 2019).The meta-analysis by Bussel et al. (2022) also found that consumers do not find sustainability an important factor in food purchases (at least, not yet).
The relatively high WTP for animal welfare has an interesting twist in that consumers, at least in Europe, value animal welfare not only for the sake of animal well-being, but also because they connect it with potentially healthier meat and better eating quality (Eurobarometer, 2007).This linkage is found especially in the German context (Uehleke and Hüttel, 2019), but it has also been identified in some Chinese studies (Lai et al., 2018;You et al., 2014).
If consumer preferences for welfare-enhanced pork are stronger than those for pork with lower climate impact, the potential for market-based reductions in climate impact is not promising.In line with Tønnesen and Grunert (2021) and Jürkenbeck et al. (2021), our findings suggest that a better way to encourage future demand for more sustainable pork is to bundle animal welfare and food safety with other dimensions involving lower adverse impacts on the climate and wider environmental values.Organic production is an example of a production system combining several dimensions.While organic standards impose requirements on animal welfare, use of fertilizers, etc., they do not contain specific requirements for lower climate impact (European Commission, 2008).Nevertheless, increasing public interest in improving the sustainability of food production (and other human activities) seems to call for action based either on market-driven initiatives or on legislative initiatives at national or international (e.g.EU) level.

Discussion of the study's limitations
The timing of the survey could have had an effect on the respondents' WTP in that it was conducted just after re-opening the international markets after main lock-downs due to COVID-19, just before the Russian invasion of Ukraine, and, for the UK respondents, during an ongoing BREXIT process.The questionnaire did elicit information about reasons for not eating pork that could shed light on this possible limitation.However, we found that only a few respondents stated that they did not eat pork for economic reasons.Hence, the timing of the survey does not seem to have coincided with unusual economic barriers to the consumption of pork for the majority of respondents.Among respondents in the UK and Germany who did not eat pork, taste was reported by almost half as the main reason for avoiding (not eating or buying) pork.Almost half of the German respondents not eating pork mentioned health reasons as a barrier to eating pork, while close to 30% of the British respondents not eating pork mentioned religious reasons.Hence, improving the taste of porkor perceptions of itcould be a way to reduce a barrier to eating pork in the UK and Germany.This ought to be of real interest to those in the pig sector and others with an interest in boosting pork consumption.
Clearly, the present study concentrated on potential market-driven improvements in sustainable pork production.A limitation of this approach is that the consumer preferences we tracked were elicited only from respondents who stated a positive WTP for more sustainable pork.Moreover, the preferences of respondents who stated that they did not eat or buy pork, young people under the age of 18, and Chinese people older than 64, were not included.Whether or not these groups would support the promotion of various dimensions of more sustainable pork production through, for example, stricter legislation or public subsidies, was not investigated.We leave it for future studies to investigate the strength of consumer preferences for more sustainable pork among respondents who did not eat pork in the varieties available at the time of the survey.Also, future studies of whether the unwillingness of consumers to pay a price premium was related to the specified sustainability-improved pork that was offered in our study would indeed be valuable.
Another limitation is that we derive WTP from a hypothetical setting in which respondents face no economic consequences as a result of their choices.This may invite yeah-saying (Lai and Yue, 2020), and therefore an experiment based on observed behavior would be valuable in confirming or disconfirming our results.Furthermore, as highlighted in Lusk and Norwood (2011), a potential lack of knowledge among consumers could affect the effectiveness of market-based initiatives to improve animal welfare.This knowledge issue would also have to be taken into account if retail food stores were moving into marketing sustainability-improved pork.
The present study focused on consumers' preferences for more sustainably produced pork.However, given the alternative (or additional) strategy of limiting animal production and increasing plant-based food consumption to reduce pressure on global resources (Steinfeld et al., 2006;Steffen et al., 2015), an interesting follow-up study would analyze how consumers prioritize between eating more sustainable pork or eating less pork in different price scenarios.
S. Denver et al.

Conclusion
In this study, we investigated WTP for more sustainable pork using consumer samples from Denmark, Germany, the UK and China (Shanghai).The vast majority of the respondents (60-75%) were willing to pay a price premium for more sustainable pork.However, only around 10% were prepared to pay more than 20% extra.Hence, our results point towards a positive but limited potential for market-driven increase in the five sustainability-related dimensions of pork in the four investigated countries.
In all four countries, lower climate impact was among the least important dimensions.The marketingand policyimplications of this result are that, despite the increasing focus on the goal of mitigating climate change at the political level, leaving it to consumers to support this goal by offering pork with reduced climate impact at slightly higher prices is not going to be an effective toolat least, for the time being.
In Denmark, Germany, and the UK, improved animal welfare was typically considered the most important dimension.In China, dimensions related to food safety (decreased use of antibiotics and guaranteed freedom from bacteria) were the main reasons for willingness-topay a price premium.Our results therefore indicate that, where sustainably produced pork is concerned, 'one size does not fit all'.Different strategies will be needed if the goal is to increase consumers' interest in buying pork with lower climate impact.In particular, lower climate impact will need to be bundled with other product or production attributes that are more highly prioritised and these attributes differ across countries.
The respondents who were willing to pay a higher price premium for more sustainable pork than other groups of respondents, and also the younger respondents, were both inclined to prioritize lower climate impact more than other respondents were.So, market potential for more sustainable pork is highest among younger consumers and consumers willing to pay a high price premium.This could imply that sustainability-improved pork will be in competition with other high premium products such as organic pork.
In all four countries we studied, the respondents who stated that climate change is an essential consideration when they buy meat were inclined to prioritize lower climate impact of pork more than other respondents.This was as expected, but we also found that even among respondents with a stated concern about climate impact, WTP for more sustainable pork was driven as much, or more, by the desire to improve animal welfare as it was by the desire to lower climate impact.Moreover, we found that, when the respondents had to choose between animal welfare and climate impact in buying pork, then mostfrom the three European countries, at any ratewould prioritize animal welfare.The argument most frequently presented for this attitude was that animal welfare is directly related to the product that is going to be eaten, whereas lower climate impact (by implication) is not and can be achieved by means other than the climate-efficient production of pork.The implication of this result for producers and policy makers is clear: the growing focus on reducing the climate impacts of pig production, and animal production in general, should not blind the relevant stakeholders to the importance of better animal welfare and food safety.If they lose sight of the latter, stakeholders will fall out of step with the priorities that many consumers currently have.Note: WTP = willingness-to-pay.Based on weighted data.The following mean weights (std.dev.) across five regions in Denmark, 16 federal states in Germany and 12 regions in the UK were used: Denmark = 1.21 (0.05) male 18-34 years, 1.09 (0.06) male 35-49 years, 1.06 (0.05) male 50-64 years, 1.04 (0.04) male 65+ years, 0.75 (0.03) female 18-34 years, 0.90 (0.05) female 35-49 years, 1.02 (0.10) female 50-64 years, 1.10 (0.04) female 65+ years.Germany = 1.23 (0.08) male 18-34 years, 0.92 (0.13) male 35-49 years, 0.93 (0.08) male 50-64 years, 0.99 (0.06) male 65+ years, 1.

Fig. 1 .
Fig.1.Distribution of reasons for prioritizing pork with improved animal welfare over pork produced with lower climate impact.Note: Based on weighted data.Only respondents with a positive WTP for more sustainable pork and who chose 'Probably improved animal welfare' or 'Definitely improved animal welfare' instead of pork produced with lower climate impact are included in the analysis.Number of respondents in the shown subcategories using weighted data: Denmark = 1166; Germany = 1100; the UK = 927; China = 1163.

Table 2
Mean scores of five sustainability-related dimensions by country (95% confidence interval shown in parentheses).

Table 4
Mean scores for five sustainability-related dimensions by country categorised according to age groups (95% confidence interval in parenthesis).Based on weighted data.Only respondents with a positive WTP (willingness-to-pay) are included in the analysis.+/-indicates that the score differs positively/ negatively (at 0.05 significance level) from the score assigned by the age group to lower climate impact.a The age group 65 + years is not shown for China as it contained only 18 respondents.In China, 19 respondents in the age group 15-17 years were included in the youngest age group.

Table 5
Mean scores for five sustainability-related dimensions by country categorised according to climate concern (95% confidence interval in parenthesis).Based on weighted data.Only respondents with a positive WTP (willingness-to-pay) are included in the analysis.+/-indicates that the score differs positively/ negatively (at 0.05 significance level) from the score assigned by the group of respondents who did not indicate that climate is important for their choice of meat (Climate not essential) and by the group of respondents who indicated that climate is important for their choice of meat (Climate essential).Number of respondents in the shown subcategories using weighted data: Denmark = 1098 (Climate not essential), 68 (Climate essential); Germany = 1002 (Climate not essential), 98 (Climate essential); the UK = 790 (Climate not essential), 137 (Climate essential); China = 1091 (Climate not essential), 72 (Climate essential).

Table 6
Distribution of respondents' priorities when choosing between pork produced with lower climate impact and improved animal welfare (shown in% of sample with positive WTP).
Note: Based on weighted data.Only respondents with a positive WTP for more sustainable pork are included in the analysis.Fewer than 5% of the respondents in each country answered 'Don't know' (not shown in the table).Number of respondents in the shown subcategories using weighted data: Denmark = 1166; Germany = 1100; the UK = 927; China = 1163.

Table A1
14 (0.07) female 18-34 years, 0.97 (0.08) female 35-49 years, 0.97 (0.06) female 50-64 years, 0.97 (0.06) female 65+ years.UK = 1.04 (0.09) male 18-34 years, 0.98 (0.09) male 35-49 years, 0.99 (0.11) male 50-64 years, 0.98 (0.10) male 65+ years, 1.01 (0.09) female 18-34 years, 0.98 (0.09) female 35-49 years, 1.00 (0.10) female 50-64 years, 1.02 (0.10) female 65+ years.a In China, 19 respondents were 15-17 years.They are included in the youngest age group.b Education was divided into 3 categories: Edu1 (primary school / no formal education), Edu2 (Postgraduate / Masters), Edu3 (Other education).Note that census is based on weights for gender, age and region.c Frequent pork consumption (denoted 'Cons')) includes respondents who consume some kind of pork at home at least 1-3 days per week.Respondents who did not report their pork consumption (answered 'don't know') were excluded from the estimation of the share of respondents showing frequent pork consumption.The following numbers of respondents were excluded from the estimation of frequency of pork consumption: Denmark 12, Germany 5 and UK 8.Note that census is based on weights for gender, age and region.d Respondents who answered 'Don't know' to the WTP question are not shown in the table.They comprise: Denmark 8%, Germany 7%, UK 8% and China 2%.Hence, their shares need to be added to the share of the groups 'No pork', 'WTP=0′ and 'WTP>0′ in order to add up to 100%.e Based on weighted data.f Unweighted shares.g Population census information were based on gender, age and region collected and compiled by the survey company Norstat.h Population census information is not available for the Shanghai, China sample.