Over the last few decades, animal farming supply chains have progressively modified their rearing systems to meet the increasing demand for cheaper food products of animal origin (FPAO) (1). From the sixties to the eighties, to increase production while reducing food prices, agri-food production systems were driven by technological innovation and European agricultural policies¹ toward farming practices that enhance production performance. In the case of farmed animals, the practices have turned toward genetic selection for productivity, adoption of feeding regimes that are far from natural feeding behavior (chemical composition and physical form of the feeds offered), cage housing systems, high stocking density, automation of husbandry procedures, and use of antibiotics within prophylactic approaches (2). However, such practices have been widely questioned by the public, especially in Europe, regarding environmental sustainability and farmed animal welfare (FAW) (3, 4). That is why European Institutions have actively responded to such requests through legislation and study commissions (5)². Together with complying with the evolving EU legislation on animal welfare (AW), sectors of the animal farming industry, at the same wavelength as the related policymaking authorities and food supply chains, have been trying to steer their farming practices (e.g., feeding, housing, husbandry, transportation) toward welfare-oriented and environmentally sustainable production processes, adopting mostly voluntary certification schemes (6, 7)³.

Within this context of evolving individual attitudes (both as citizens who express their concerns overtly, thus affecting public opinion and urging political decision-making, and as consumers), recognizing and investigating the interactions between demographic and socioeconomic traits and individual convictions could offer valuable insights into the various nuances characterizing the opinions of a multifaceted public (8). In fact, over the past 20 years, EU authorities have sought to assess and monitor EU citizens’ opinions and perspectives on AW through the Eurobarometer (Eb). Since 1974, this polling instrument has been used to monitor the social and political attitudes of citizens in EU Member States on a wide range of topics, including EU policies, health, social issues, and the environment, and more⁴. In particular, the 2023 Eb survey provides, by far, the most recent and largest pool of data, offering valuable insight into current EU citizens’ declared attitudes toward AW⁵. However, such resources appear to be neglected by the scientific community, together with survey investigations of the same nature and methods worldwide (8), which have not been using them to their full potential to tailor research according to citizens’ requirements, thus depriving policymakers and other stakeholders of helpful inputs to set the direction of their decisions and carefully design their strategies.

Based on the data collected through the 2023 Eb survey, the present study aimed to:

This second analysis intended to suggest new approaches for policymakers and the food supply chain to build interactions with citizens to support and improve animal welfare.

The cluster analysis implemented in the present study identified seven clusters, more than those of another recent work by de Boer and Aiking (12), which was also conducted at the European level and based on a 2022 Eb survey on food and safety, where five clusters were obtained. However, if described according to the percentages of positive answers (>80%, Table 2), our seven clusters can be grouped into three larger macro-clusters characterized by two opposing attitudes for the extreme groups (clusters 1 and 4 vs. clusters 5 and 7) and, in the middle, a group whose concerns focused especially on specific farming practices (clusters 2, 3, and 6). Nonetheless, considering the weight of such clusters (Figure 2), the ones that were highly concerned about farmed animal welfare (FAW) overall include the vast majority of respondents (74%), while the poorly concerned ones represented only a small minority (6%).

The total number of 2023 Eb respondents accounted for 0.007% of the total European population over 15 years old in the 27 Member States, which was higher than the population sampled in other investigations carried out in the last few decades on European citizens’ opinions on animal welfare (AW), conducted only in a selection of representative European countries (13–16). This confirms that the Eurobarometer provides by far the most comprehensive datasets within the European context. However, very few other peer-reviewed articles resulted when searched through scientific literature databases, such as Scopus, and using the following string: “Eurobarometer AND animal AND welfare.” As of 2005, when the first Eb survey on FAW was published, nine results were found. Of these, only two papers were relevant and comparable in terms of the investigated topics, data, and work methodologies.

As a result of a stratified multi-stage, random (probability) sample design in terms of overall socioeconomic variables, and if compared with Eurostat data, the final respondent sample was representative of the socioeconomic composition of the actual 15-and-over population in 2023. In fact, a stratified multistage sampling appears to be the most suitable choice for large-scale monitoring surveys that aim at assessing status, change and trends of one or more parameters within a highly geographically diverse group, such as the 15-and-over European population; being both practical and effective and cost and time saving (17, 18). As a multistage random approach, where nested or hierarchical structure of the members within the population is taken into account, and then arranged in clusters that will be randomly sampled at each stage, it ensured a representative final sample of the original population (16, 19). Finally, as a stratified sampling designed according to the EUROSTAT NUTS II (or equivalent) and the DEGURBA Urban Rural classification, it ensured representative coverage based on the entire national territory (10, 17, 19). In light of the great representativeness of the final sample of respondents to the 2023 Eurobarometer survey, the clusters produced through our statistical approach are a representative clusterization of both the attitudes toward FAW of the actual EU population and of its socio-economic characteristics.

The majority (74%) of the respondents, represented by clusters 5 and 7, who were highly concerned about (almost) all the investigated question items about FAW, could also be profiled as being mainly composed of females, younger individuals, with a tertiary education level, likely from Southern Europe, and in regular contact with companion animals. Indeed, such clusters included most respondents that, in their daily lives, were in regular contact with animals overall, especially companion animals, as opposed to clusters 1 and 4 that included most of those who were in contact with farmed animals (Table 3). It should be noted that a source of bias influencing respondents’ answers could have arisen from the formulation of QC5 (Table 1). Since it explicitly specified that the practices listed in the question could have been applied both to “farmed animals and the breeding of cats and dogs for commercial purposes,” it might have instilled a deeper feeling of sympathy in the respondents, coming from the involvement of companion animals (20). From our findings, the answers to both “QC1 More information” and “QC3 Better protection” also confirmed the trend described above, with a majority of respondents calling for a better protection of FAW in their own countries (88%; i.e., clusters 2, 5, and 7) and 74% of individuals declaring a desire for more information on FAW (i.e., clusters 5 and 7). This shows once more that this portion of respondents is more skeptical about the fact that animals are kept in good condition, and therefore, ask for further information and improvements in FAW.

Such higher levels of concern about FAW topics among women, younger citizens, and those who have spent more time in education, both in Europe and across the world, have already been reported on many occasions (3, 21–24). The importance placed on AW appears to decrease with age, as older individuals are more readily accepting current welfare standards and have more production-related views of animals (25), which is in line with the moderate attitude held by older respondents in clusters 2 and 6. However, although there have been studies reporting lower levels of knowledge about actual farming, transportation, and slaughter conditions and practices (26, 27), from our findings we cannot establish neither what is the factual knowledge that Eb respondents had on AW or how such knowledge was acquired by the investigated sample of EU citizens. In contrast to previous Eb surveys (2007 and 2016), in 2023, no questions were asked to the interviewed citizens about their knowledge of the topic or the source of information they relied on (28, 29). In fact, although without implying that attitudes must only be based on factual knowledge nor that they would be necessarily influenced by it, such data would provide valuable inputs for designing information strategies by policy makers, in the first place, and by the food supply chain. Despite a lack of concrete knowledge, consumers were reported to share the impression that the living conditions of farmed animals are far from optimal, especially in conventional production systems (30, 31). This impression, leading to the inevitable conclusion that animals suffer despite all the advances made in animal farming, once again appeared to be typical of women, younger, and well-educated individuals in professional positions, who, therefore, tend to hold on to their negative views of modern farming. Studies have highlighted that these categories of people are more exposed to vague and general information that do not come from thoughtful, accurate, and contextual sources. As a consequence, it is harder for them to realize that farming can be completed also via systems that are both sustainable and animal friendly, and that such systems may differ significantly from the narrative they are exposed to (31, 32). The results of the present study showed that even if it was a small group, likely mirroring an equally small share of farmers and agricultural workers⁷, those in regular contact with farmed animals held a less critical position toward modern farming systems, echoing other studies that have shown divergent perceptions between ordinary citizens and livestock producers or other animal scientists (e.g., veterinarians) (33–35). It must be noted that the attitudes of such a smaller share of respondents (clusters 1 and 4) may be driven by being too accustomed to the reality of breeding and farming practices, or by their likely more pragmatic view of animals’ roles, which makes them readily accept current welfare standards. Furthermore, as clusters 1 and 4 are mainly made up of manual workers and individuals with a secondary level of education, they might have less disposable income allowing them to purchase food products that may be perceived as more expensive (i.e., animal welfare-friendly food products) (1). Indeed, by looking at their answers to QC11, it seems that they were less likely to search for such products, especially when compared to cluster 7, where 82% of the respondents declared that they were looking for animal welfare-friendly products.

Finally, the 2023 Eb survey did not contain any questions investigating the respondents’ lifestyles in terms of diet and religious beliefs, which could greatly affect their approach to FAW (8). Among the majority of highly concerned respondents just described and the smallest group (6%) of citizens that appeared to be poorly involved in the topic, there was a share of respondents (21%) who were concerned mainly about specific FAW aspects regarding farm facilities and rearing systems, namely qualified personnel, enough space and food, adapted environments, mutilations, and the use of individual cages. In fact, these specific farming practices appeared to be a concern for almost all surveyed individuals (97%; all but cluster 1; Table 2 and Figure 1). This is consistent with the findings of Pejman et al. (15) who, still in 2019, reported that AW was most frequently put in relation with the general concepts of “natural outdoor conditions,” “clean and healthy housing environments,” and “good feeding.” According to recent studies, the public seems to be more aware of and sensitive to aspects and practices that have been put under the spotlight of animal protection associations through information campaigns⁸ (1, 24, 36). Furthermore, European studies have reported that animal farming systems are perceived as less animal friendly when there is a lack of free movement, too little space per animal, and non-transparent locked systems, especially for pigs and poultry (30, 37–39). These are the more common farming conditions in Southern Europe, which could partly explain the significantly higher proportion of respondents in clusters 5 and 7. However, it should be noted that this is in contrast to the biosecurity protocols required in highly specialized rearing systems and the necessary adaptations of the rearing techniques, such as grazing practices or outdoor living, to the specific adverse climate conditions (heat stress, rain, and wind) and breed characteristics that make it necessary to protect the animals (40). On the other hand, transport and slaughter conditions still appear to be underrated AW issues; perhaps because they are perceived as short-term issues compared to on-farm living conditions that persist during the entire animal’s lifetime or matter with few obvious alternatives. Furthermore, the most recent information campaigns run by animal protection associations have focused more on animals’ on-farm life stages, rather than on transport and slaughter, to raise awareness of topics such as individual cages and mutilation⁹.

If looking at the present results from a geographical perspective, it emerged that respondents from Eastern Europe were less concerned about FAW overall, as opposed to other European countries, many of which have been exposed and sensitized to FAW issues since the seventies (5). This might be of valuable support to European policymakers and authorities, as it further suggests that considerations and decisions should be tailored to both the targeted socioeconomic groups and specific geographical zone (8).

The answers to QC11 and QC12 highlighted whether the respondents looked for labels identifying welfare-friendly food products, and whether they thought there was a sufficient choice of such foods in shops and supermarkets. The analysis of the answers showed a certain level of disagreement among clusters and within previously unanimous socioeconomic groups. A total of 43% of the 2023 Eb respondents looked for welfare-friendly labels, mainly female, younger, and well-educated individuals, likely from Southern Europe, and in regular contact with companion animals. A decreasing percentage, if compared to 2016, when the same question was paused and 52% of the respondents answered positively (28). It is possible that, as consumers, EU citizens do not want any further labels on FPAO, and it is likely that such products respect animal welfare as a necessary prerequisite. Otherwise, a dissociation might be observed between what respondents declared through their answers and what they actually looked for when shopping. Alonso et al. (3) found that although FAW has been viewed as an increasingly important attribute of food quality and safety, EU consumers still prioritize their own human safety and individual benefits over societal or animal benefits per se, using AW more as an indirect indicator for human health. Moreover, while in our previous results, the young, female, and well-educated socioeconomic group represented by clusters 5 and 7 was always in agreement on every question item, it split when it came to QC11 and QC12, with half of them stating that they look for welfare-friendly labels and think there are enough welfare-friendly foods (40%; cluster 7) and the other half not searching for nor thinking there are enough of such products (34%; cluster 5). Furthermore, the evenly distributed overall positive and negative answers (49% vs. 41%, respectively) and the highest prevalence of “Do not know” (10%) to “QC12 Sufficient welfare-friendly products” testified to the lack of awareness surrounding these products, which seemed to be highly disregarded since 20% of the respondents openly declared they do not search for labels identifying them (clusters 2, 3, and 4). For this reason, it is crucial to understand the drivers of the respondents’ shopping habits and have more information about their dietary regimes (omnivores, vegetarians, vegans), whether they place importance on food price and are responsible for shopping, and who they trust for certifying products’ origins. In general, even if perhaps ambitious, knowing the respondents’ ethics in relation to animals and nature would be extremely helpful in driving conclusions, especially given the increasing divergence between science-based welfare assessments (i.e., ethological issues, physiological parameters, and health status) and ethical assessments involving ethical, socioeconomic, cultural, and religious aspects. Unfortunately, these data were not collected through the 2023 Eurobarometer survey. Furthermore, a discrepancy between self-reported public concerns about FAW and shopping habits has been widely reported and investigated, along with consumer dissociation between live animals and the food they produce (3, 41, 42). For example, although EU citizens appeared increasingly worried about FAW, they were not willing to cover the costs of FAW improvement. This phenomenon was confirmed by the results of the Eb surveys conducted over the years, where a slightly increasing share of more than a third of the respondents declared that they were not willing to pay more for welfare-friendly products (34, 35, and 37% in 2005, 2016, and 2023, respectively) (10, 28, 43). Finally, recent literature has discussed that although consumers welcome additional labeling on production quality overall, they also accept simple labeling better (44). This calls for accurate but accessible information provision to fill the knowledge gap and distance between actual and perceived reality, thus boosting confidence in food chain stakeholders, as well as for simple and clear labeling schemes to promote informed purchasing decision-making as competent consumers. Implementation strategies for accessible animal welfare labeling are still one of the core topics for agricultural policy at the EU level (45).

Through this study, based on the most recent Eb survey (2023) on EU citizens’ attitudes toward animal welfare, the Eurobarometer appeared to provide a comprehensive dataset capable of capturing the opinions of the composite demographic and socioeconomic categories of the 15-and-over EU population.

As one of the very few research attempts to utilize the Eurobarometer survey, in our reappraisal of the survey data, a cluster analysis allowed the identification, quantification, and further examination of multiple intertwined strands characterizing EU citizens’ opinions in relation to the socioeconomic category to which individuals belong. The seven resulting respondent clusters could be further merged into three macro-clusters with two oppositely high (74% of the sample size) and poor (6%) levels of concern for animal welfare, and a third macro-cluster in between, concerned especially about specific farming practices (20% of the sample). However, a certain level of disagreement among clusters and within socioeconomic groups emerged in terms of shopping habits, with only a minority of the respondents actively looking for welfare-friendly labels on the FPAO. It would be useful to investigate what the actual EU citizens’ attitudes and desires are on label schemes.

In terms of socioeconomic composition, such aggregated clusters seemed to be driven mainly by gender, level of education, and regular contact with animals, with women and well-educated people in regular contact with companion animals being more concerned about farmed animal welfare overall. Above all, it also appears that the implementation of animal welfare strategies and assessment protocols, tailored to both the socioeconomic group and the specific European zone, needs to be better combined with effective and simple communication to the wider public and aiming at helping citizens orient their shopping actions.

Furthermore, the outcomes of the cluster profiling could contribute to informing policymakers and stakeholders within the animal food supply chain to promote more welfare-friendly scenarios according to the concerns of different socio-economic groups. Detailed knowledge of the different social layers’ opinions and attitudes might help establish better conflictless ways to make the two ends of the food supply chain (i.e., producers and consumers) approach each other.