Irish farmers’ engagement with dairy calf welfare: An exploratory analysis^†

Abstract

Dairycalf welfare is receiving increasing public attention. To ensure optimal practices, farmers need to engage and be mindful of unethical behavior. However, avoiding information on animal welfare is common and often driven by willful ignorance. We conduct an exploratory analysis on survey data from 546 Irish dairy farmers. We investigate farmers’ choice to view a picture of transported dairy calves and find that over 20 per cent of farmers prefer to remain in a state of ignorance. Higher self-reported calf mortality and education increase the odds of viewing the picture while being a female decreases them. Farmers’ reasons for avoiding include the lack of new information in the picture, anticipated negative feelings, and biased information expectations. Additional explorations suggest that the farmers’ decision is not a survey artifact, but the results are not robust and further research is needed to confirm. Possible suggestions to improve calf welfare are to foster farmers’ trust, package information as new evidence on best calf-rearing practices, and better understand perceptions of animal welfare.

1. Introduction

Farm animal welfare is receiving considerable attention from the public. About 84 per cent of European citizens believe that farm animals should be treated better and 83 per cent support limited transportation time.¹ Consumers increasingly value animal welfare by displaying greater willingness to pay for products that are more respectful of animal welfare (Norwood and Lusk 2011; Lagerkvist and Hess 2011; Mulder and Zomer 2017) and showing higher support for meat taxation when it is justified by animal welfare (Perino and Schwickert 2023). The increased valuation of animal welfare as a product attribute suggests that sub-optimal agricultural practices may draw negative attention with knock-on effects for the animal-based sector. While most of the animal welfare concern has been related to intensive meat production (Faucitano et al. 2022), the attention is shifting to dairy production (Ritter et al. 2022).

As consumers are becoming informed on the living conditions of dairy farm animals, they appear increasingly concerned by the cows’ transportation (Waldrop and Roosen 2021), lack of pasture access (Schuppli et al. 2014), early calf cow separation (Busch et al. 2017; Sirovica et al. 2022), and slaughtering of calves at young age (Ritter et al. 2022). In particular, perceived sub-optimal treatment of calves evokes negative emotions (Busch et al. 2017) and is rejected by consumers who associate such practice with elevated stress levels for the animals (Placzek et al. 2021). The fate of surplus dairy calves² is a topic that has received considerable attention from the public. For example, reports about dairy calves treatment in Ireland have been in mainstream national (i.e., RTE News³) and international media (i.e., the Guardian⁴). In addition, animal welfare NGOs regularly post triggering videos online showing the mistreatment of animals on farms.⁵ While such reports are informative and place farm animal welfare into the public debate, they draw a negative picture of farmers. In turn, this may polarize farmers and lead them to distrust or reject any information from outside of the farming world. As a consequence, farmers might associate any new piece of information as either negative or biased. In this study, we explore Irish dairy farmers’ willingness to engage with calf welfare during transportation and assess reasons for engagement (or otherwise).

Societal concerns about animal transport are also reflected by the European Commission’s proposal (published in December 2023) to update animal welfare legislation for the first time since 2004. The proposal highlights the need to ensure the welfare of animals during long-distance transportation and unweaned animals receive particular attention. For example, results from the consultation activities that were conducted to revise the guidelines report that citizens and animal welfare NGOs prefer to ban the transport of vulnerable animals, especially unweaned animals. While national authorities support the introduction of stricter measures for unweaned animals, similar to producers’ views who support specific requirements, these groups do not support a ban on transport (European Commission 2023). This underlines that the transportation of unweaned calves is a topic of broad interest with competing interests and increasing societal disapproval.

The welfare of dairy calves has also become a growing concern for the Irish dairy industry over the last few years due to a major expansion of the dairy sector. This expansion was initiated by the EU milk quota abolition in 2015, which allowed farmers to expand production unconstrained for the first time in over a generation. In Ireland, dairy cow numbers have increased by almost 50 per cent over the last decade, and there are over 1.5m dairy cows in Ireland (Office 2022). The major expansion in dairy cows, coupled with an increasing breeding focus on milk characteristics (Kelly et al. 2020), led to more dairy calves’ birth and increased the number of surplus dairy calves. In addition, the spring calving system in Ireland adds to the challenge of calf welfare, as the vast majority of dairy calves arrive in a 6–8-week calving interval early in the year. This leads to a high workload for farmers during this time and a marketing challenge due to heightened calf supply in the spring. Both of these factors might imply an increased risk for animal welfare.

In relation to calf marketing in Ireland, in 2019, over one quarter (26 per cent) of calves were reared on dairy farms. Approximately one-third of calves were sold directly to other farmers, over a quarter of calves were sold through the mart, 12 per cent were sent for live export, and 4 per cent were culled shortly after birth or disposed of by animal by-product collection services (Läpple and Ramsbottom 2020). The increasing number of calves sent for live export and calves that are culled shortly after birth (or disposed of) has caused animal welfare concerns and disapproval by the public. This is because live export of unweaned calves and premature culling create stress factors for animals due to transport (often over long distances and different means), food withdrawal, and movement through different markets (Pardon et al. 2014; Haskell 2020; Kuo and von Keyserlingk 2023). As such, the negative impact on animal welfare in relation to surplus dairy calves is of concern, and options for how to improve the situation are needed. One possibility is to engage farmers with animal welfare to inform them about potential malpractices that happen during calf transports, which they may not be aware of. For example, studies have found that farmers are not always fully aware of the level of animal welfare standards provided on their farms (Cutler et al. 2017) or might not engage with information about animal health (Bell et al. 2006).

If engaging with new information could lead to positive outcomes, why would some dairy farmers refuse to do so? Informational biases alter information acquisition. Numerous studies in economics identify situations where people avoid information, even when it is free and could improve decision-making (Golman et al. 2017; Ho et al. 2021).⁶ Several reasons are provided in the literature as to why people strategically avoid information. Information avoidance can arise to escape the arousal of unpleasant emotions (Sweeny et al. 2010). For example, individuals might be guilt or regret-averse and decide not to acquire new information if they expect it to trigger such emotions (Golman et al. 2017; Gabillon 2020). Individuals might also try to downsize the moral discomfort generated by cognitive dissonance, where their actions are inconsistent with their beliefs (Festinger 1962). Willful ignorance, or strategic ignorance, is particularly predominant in meat consumption and animal suffering, for which individuals would rather stay uninformed on the living conditions of farmed animals to keep eating meat without experiencing moral tension (Onwezen and van der Weele 2016; Bell et al. 2017; Leach et al. 2022).

Information avoidance can also happen due to optimism bias (Huck et al. 2017). Individuals expect that new information will lower their optimism and prefer to avoid it to maintain positive expectations. In the case of dairy farmers, engaging with calf welfare transportation might highlight some wrongful practices and generate unpleasant emotions, moral discomfort, and alter optimism (e.g., information on how their calves are treated once they leave the farm). Individuals also avoid information because they expect it to be in contradiction with their own beliefs, out of distrust, or because they anticipate that the information is biased. Dairy farmers, who are professionals in daily contact with calves, most likely have their own belief system about welfare and optimal practices. For example, Bell et al. (2006), based on a study of 61 dairy farmers in the UK, report that over 10 per cent of farmers incorrectly assess their herd health status in relation to lameness, and the majority of farmers had not reviewed any health records. This suggests that some farmers avoid external information and rather rely on their expertise, thus perpetuating sub-optimal practices.

In this paper, we explore whether Irish dairy farmers are willing to engage with animal welfare in the form of an image of surplus dairy calves on a lorry. Overall, assessing farmers’ attitudes in relation to calf transportation helps in understanding the perception of current practices on the production side. This is important as it can help design optimal information provision to improve the sector in terms of animal welfare and public opinion.

We run an exploratory analysis using data from a two-wave survey experiment on Irish dairy farmers conducted in 2020 and 2021. We focus here on a subsection of the survey, in which 499 dairy farmers are given the choice to visualize an image of Friesian bull calves on a lorry or a blank screen for five seconds. We collect dairy farmers’ choices alongside their reasons for viewing (or not) the picture, socio-economic characteristics, and farms’ characteristics. We explore the reason for viewing (or not) the picture by running a logit model to assess the correlations between the role of farmer and farm characteristics. Of particular interest are calf housing (i.e., the proportion of calves that can be housed on the farm) and self-reported calf mortality. We interpret higher calf mortality as lower calf welfare and higher calf housing as higher calf welfare. In addition, we explore whether the decision to view calves on a lorry is associated with the nature of the picture by using a control/treatment procedure. A separate group of 47 farmers is randomly allocated to a control condition, in which they are asked to choose between viewing the image of grazing dairy cows or a blank screen for 5 seconds. We explore possible treatment differences due to the nature of the picture but draw no causal conclusion due to a lack of statistical power.

The results show that a large proportion of farmers (i.e., 21.24 per cent) play with the moral wiggle room by willfully avoiding the picture and staying in a state of ignorance. The econometric analysis reveals that the likelihood of viewing the image is correlated with farm and farmers’ characteristics. Over 70 per cent of the farmers who decide to view the picture do so because they care about their animals once they have left the farm. Interestingly, some farmers feel that the picture is biased. For farmers who decide to look at the blank screen instead, the main reason is that they know how calves on a lorry look. However, some farmers expect it to be biased information, suggesting that the source and nature of the information might matter when attempting to inform farmers. Lastly, while sending calves on a lorry is part of dairy farmers’ tasks, many avoid viewing the picture in anticipation of feeling guilty or sad. In turn, this reveals that some farmers stay in a state of ignorance to avoid the negative cognitive load associated with the picture.

Furthermore, the nature of the image is correlated with farmers’ choices. Exploring the treatment effect reveals that dairy farmers are more likely to avoid engaging (i.e., deciding to see a blank screen) when the picture depicts calves on a lorry, in comparison with grazing dairy cows. This suggests that calves on a lorry could trigger more negative emotions for farmers than grazing cows, for which they could be more likely to stay in a state of ignorance. However, we are cautious when analyzing this result due to the small number of observations in the control group (i.e., underpowered design) and failure to pass a robustness check. Further research on this result is necessary to confirm the impact of the nature of the image.

The remainder of the paper is organized as follows. Section 2 details the experimental procedure and design. Section 3 presents the main results. Last, Section 4 concludes the paper and discusses potential follow-ups.

2. Survey Experiment Design

2.1. Sample

The sample of Irish dairy farmers was recruited using a two-wave survey. After removing incomplete responses, observations with implausible values (e.g., herd sizes greater than existing farms in Ireland or mismatch between herd size and farm size in hectares), and inattentive responses (i.e., no variation in Likert scale answers for attitudes), our final data set comprises 546 observations.

2.2. Survey procedure

The two-wave online survey experiment was conducted in Qualtrics with Irish dairy farmers at the beginning of 2020 and in September 2021. The survey was distributed to farmers through a survey link in a popular Irish farming press and through farm advisors, who sent the link directly to their clients.⁷ The survey was financially incentivized and all farmers who completed it received a gift voucher. The farmers were made aware that their answers were fully anonymous. They were also informed that the goal of the survey was to gather details on breeding, calf markets, and facilities in Ireland, as well as collect their attitudes and opinions in relation to the treatment of animals and potential solutions to improve markets for dairy calves.

Figure 1 displays the survey procedure. The survey comprised the following sections: Farm characteristics and breeding strategy; willingness to pay for a policy to improve animal welfare; animal welfare attitudes; social value orientation; information use; calf housing; and socio-economic characteristics. The order of the sections was not randomized. The survey procedure and analyses were not pre-registered.

2.3. Dairy calf welfare engagement task

In this paper, our focus is on the section “information use”, highlighted in blue in Figure 1. We ask participants to either avoid or engage with a picture. Similar to Bell et al. (2017), farmers were told that they could either view a picture of Friesian bull calves on a lorry or see a blank screen for 5 seconds. The question from the survey is presented in Figure 2. We used a picture as a visual communication as photographs have been found to be the most effective tool to influence viewers’ moods, attitudes, and reactions in general (Scultz 2002; Berenguer 2007; Whitley et al. 2021). In fact, Whitley et al. (2021) label photography as likely to be the most influential and significant part of contemporary visual culture. The question assumes that a photograph is more appealing than looking at a blank screen, which is rooted in the ubiquity of pictures in human culture (Bell et al. 2017). Therefore, by choosing to see a blank page, farmers deliberately avoid engaging with calf transportation.

Additionally, we collected the farmers’ motivations for viewing the picture (or not) by asking them to indicate reasons for their choice and feelings. We included reasons relating to feeling guilty or sad as it is a popular reason for not engaging with information (Thunström et al. 2016). We also anticipated that some farmers could believe that the information was biased or simply express a lack of interest (due to sufficient knowledge) and included questions to capture such attitudes (see Appendix  A).

Lastly, we recorded data on farmers, practices, and farm heterogeneity. In relation to farm characteristics, we recorded herd size as the number of cows that were milked in the previous year, and farm size as the area farmed, for which respondents had the choice to report in hectares or acres. The number was then converted into hectares by dividing acres by 2.471. We collected data on the marketing of surplus calves, where we defined “surplus calves” as all calves that are not for replacement for the dairy herd. Of particular relevance for this study is the proportion of calves sold for live export (export surplus calves). We asked farmers to self-report calf mortality, which we use as an indicator for animal welfare (Uetake 2013; Osawe et al. 2021), and recorded the proportion of calves that can be housed on the farm at any one time (calf housing). Finally, we collected some socio-economic characteristics, such as the age of the farmer, the level of education, and gender.

2.4. Engagement task with treatment effect

In addition, we collected some data for a “control group”,⁸ where farmers were asked to either view a picture of grazing dairy cows or a blank screen for five seconds. The difference between the control and treatment groups lies in the nature of the animals’ welfare conditions. In the control group, grazing dairy cows are associated with a state of happiness and a stress-free environment whereas, in the treatment, Friesian bull calves on a lorry are associated with a high-stress environment and possibly lower animal welfare. Therefore, if the proportion of farmers who choose to avoid seeing the picture in the treatment group is greater than in the control group, then farmers actively avoid engaging with calf transportation. The question design with the treatment allocation is presented in Figure 3. The control group is limited to 47 observations, which greatly impacts the statistical power. We report an effect size and a direction but complementary data is needed to fully confirm the effect.

3. Results and Discussion

3.1. Summary statistics

The summary statistics for the total sample and according to the decision to view (or not) the picture of calves in a lorry are reported in Table 1.⁹ The farmers who participated in the survey have an average farm size of 82.9 hectares with 137 dairy cows, which is higher than the 64 hectares and 92 cows average dairy farm in Ireland in 2019 (Dillon et al. 2020). In Ireland, farming income generally increases with farm size, but Dillon et al. (2023) highlight a wide variation in farm income of large dairy farms. In general, farm size is a key factor in explaining productivity in developed countries, as has been shown for the United States by Sumner (2014). Regarding the activities with calves, on average, farmers self-report that they can house 78.5 per cent of calves on their farms at any one time, have a 3.85 per cent calf mortality,¹⁰ and export 11.33 per cent of surplus calves. Regarding socioeconomic characteristics, the average farmer in the sample is a male (over 91 per cent of male respondents) who received third-level education or higher (over 63 per cent) and is younger than 46 years of age (over 61 per cent). Here, we note that the farmers in the sample are younger than the average Irish dairy farmer (54 years old) (Dillon et al. 2020). Overall, when interpreting the results, it should be noted that survey respondents are from larger farms managed by younger farmers. While this is generally associated with the adoption of new technologies and farm practices (Feder et al. 1985; Doss, 2006), in relation to animal welfare however, there is no clear relationship between farm size and animal welfare provision (Grethe 2017).

We now separate the total sample into two subgroups, according to the farmers’ decision to view the picture of calves on a lorry or a blank screen. In total, 21.24 per cent or close to one in five farmers decide not to view the picture (two-sided proportion test, P-value <0.01). We also explore the heterogeneity, both in terms of practices with calves and socioeconomic characteristics, between the farmers who choose to view the picture and those who do not. Regarding calf management practices, farmers who choose to view the picture have a higher self-reported calf mortality rate (two-sided t-test, P-value <0.1) and higher self-reported calf housing (two-sided t-test, P-value <0.1). For the socioeconomic characteristics, farmers who are males (two-sided Fisher exact test, P-value <0.01) and are more educated (two-sided Fisher exact test, P-value <0.01) are more likely to view the picture. We further explore these relationships in Section 3.2.

3.2. Dairy calf transportation engagement

In this section, we explore the relationship between engagement with dairy calf transportation (i.e., choosing to view the picture or not) and farm and farmers’ characteristics. We estimate a logit regression following equation (1):

where Pr is the probability of farmer i viewing the picture, X_i is a matrix of k of predictors containing farm and farmers’ characteristics, year is a dummy variable equal to 1 if the survey was completed in the second wave (i.e., 2021) and 0 otherwise, β are coefficients to be estimated and ϵ_i is the stochastic error term. For a more intuitive interpretation, we discuss the results in terms of the odds ratio, calculated by taking the exponential value of the logit coefficient. The results of the estimation are reported as odd ratios in Figure 4.¹¹

First, regarding the farm characteristics, a 1 per cent increase in calf mortality is associated with an increase in the odds of choosing to view the picture by 12.9 per cent (two-sided coefficient testing, P-value < 0.05). This may be because farmers with higher calf mortality are aware of the problem and are thus willing to engage, or alternatively farmers with higher calf mortality may be less sensitive to animal suffering and thus do not expect to be affected emotionally by calves on a lorry. The latter explanation is in line with findings by Andreoni and Payne (2013) that people who do not care about the subject gain nothing from avoiding it, as they do not feel guilty. Yet, when exploring our sample farmers’ reasons behind viewing the picture using two-sided t-tests, we do not find statistically significant differences for any reasons with respect to self-reported calf mortality (see Section 3.3).

Regarding calf housing, a 1 per cent increase is associated with a 1 per cent increase in the odds of viewing the picture (two-sided coefficient testing, P-value < 0.05). Due to the small economic effect, this finding is not a key result. Nevertheless, in line with our interpretation that farmers with higher calf mortality are aware of the problem and willing to engage, this may also be true for farmers with more calf housing. For example, we find that farmers who decide to view the image and select the statement “I care about my animals after they left my farm” have higher calf housing (two-sided t-test, P-value < 0.05). Conversely, farmers who select the statement “I felt guilty when looking at this picture” have lower calf housing (two-sided t-test, P-value < 0.01). This suggests that farmers are aware of potential calf welfare issues. Herd size, utilized agricultural area, and the percentage of exported surplus calves are not significantly correlated with the odds of choosing to view the picture.

Second, farmers’ characteristics are correlated to the decision to view the picture. For farmers, having a third-level education or higher is associated with a 127.6 per cent increase in the odds of viewing the picture (two-sided coefficient testing, P-value <0.01), while being a female farmer decreases the odds by 56.7 per cent (two-sided coefficient testing, P-value <0.01). These findings suggest that a higher level of education is associated with better engagement and that female farmers might be more sensitive to animal welfare concerns. This last result is in line with the literature suggesting that females are generally more sensitive to animal welfare concerns (Amiot and Bastian 2017), even though this phenomenon can be more complex (Boaitey and Minegishi 2020). We further explore this relationship in the next Section by assessing reasons for viewing the picture.

3.3. Motivations for engagement

Next, we explore the farmers’ reasons for looking at the picture and their feelings. Farmers who chose to look at the picture are presented with three statements for which they could tick a box if the statement accurately describes how they felt when looking at the picture, plus an additional open-ended question option labeled “other”. The statements and choice frequencies are reported in Figure 5. About 72.27 per cent of farmers who looked at the picture ticked the statement that they care about their animals after they leave their farm. This is especially true for male farmers (two-sided Fisher Exact test, P-value <0.05) and those who have a higher calf housing (two-sided t-test, P-value <0.05). The results show that male farmers self-declare caring about their animals more than female farmers, confirming that the relationship between farmers’ gender and animal welfare requires further assessment. Overall, it seems that the farmers care about the well-being of their calves but that the information processing and feelings differ. On the one hand, 10.69 per cent of the farmers indicate that they felt guilty when looking at the picture, suggesting that they process information with minimal bias and are aware of the negative impact of transportation on calves. Interestingly, farmers who select this statement have a lower calf housing (two-sided t-test, P-value <0.01). On the other hand, 22.65 per cent of the farmers feel that the picture is biased information, among which female farmers are more likely to select this statement (two-sided Fisher Exact test, P-value <0.05). It appears that for this group of farmers, the picture does not paint a realistic view of calf transports. As the picture is neutral and does not depict any animal suffering, it may be the case that this group of farmers is already sensitive towards information about animal welfare coming from outside the farming community.

Finally, 9 per cent of respondents state other reasons. For example, one farmer writes: “I feel I’ve little choice on the outcomes of calves that leave my farm”, while another one states: “It would be false or disingenuous of me to feel anything by looking at this picture I am a farmer I drink milk I eat meat I am aware of the implications of those decisions”. Other farmers comment that the calves look “happy enough” or compare the calves to people on a tube at rush hour. Another farmer writes: “I looked after them well. It’s the next owner’s responsibility to do the same”. The responses suggest a great amount of heterogeneity among the participants’ views and different ways of processing the information. Further explorations on the farmers’ attitudes and beliefs on information could bring some insights on better informing this specific population, especially on sensitive topics such as animal welfare.

We also explore farmers’ reasons for choosing to look at the blank screen instead of the picture as well as their feelings. The responses are reported in Figure 6. The main reason for not looking at the picture is that farmers know what calves on a lorry look like, with over 50 per cent of farmers choosing this option. However, expectations of feeling guilty or sad are also reasons selected by a significant proportion of farmers. Specifically, 27.36 per cent of farmers select that they expect to feel sad when looking at calves on a lorry, while 15 per cent indicate that they expect to feel guilty. It is also interesting to note that 23.59 per cent expect the picture to be biased information. Only 8.49 per cent of farmers select that they are not interested in calf transports and again 8.49 per cent indicate that it is not their business anymore once they sell their calves. Farmers who select this last statement have a lower calf housing (two-sided t-test, P-value <0.01) and have lower education levels (two-sided Fisher Exact test, P-value <0.05).

While all respondents who decide to avoid looking at the picture actively avoid engaging with calf transports, the motives seem to vary. First, for those who expect to feel guilty or sad, avoiding the picture might stem from a desire to avoid a negative cognitive load, such as cognitive dissonance or unpleasant feelings. Second, for those who expect the image to be biased, are not interested in calf transportation, and declare that it is none of their business once the calves are sold, the willful ignorance could be an attempt to play with the moral wiggle room. Indeed, by willfully maintaining a state of ignorance and moral disengagement about calf transportation welfare, these farmers might maintain negative practices while keeping a positive self-image. In particular, this is confirmed by the fact that farmers who declare that it is none of their business once calves are sold have a lower calf housing proportion.¹²

3.4 Treatment effect

As a final part of our analysis, we explore the effect of the random treatment allocation on the decision to view (or not) the picture.¹³ The number of farmers who chose to view the image over a blank screen, per condition, is presented in Table 2. Out of the 47 farmers in the control group, only one decided to look at the blank screen over an image of grazing dairy cows. In the treatment group, out of the 499 farmers, 106 decided to view the blank screen over a picture of Friesian bull calves on a lorry.

First, we explore the treatment effect by looking at differences in proportions between the control and treatment groups, which yields a statistically significant difference (two-sided Fisher Exact test, P-value < 0.01). We then estimate a logit regression to assess the extent to which the exposure to the treatment lowered or heightened the likelihood of viewing the image. Formally, we first estimate equation (2a) to recover the pure treatment effect:

where T represents the treatment (T_i equals 0 if farmer i is in the control group and 1 if farmer i is in the treatment group). We then estimate equation (2b) with added farm and farmers’ controls:¹⁴

We report the results from the logit regression as odd ratios in Table 3. The odds ratio is 0.081 (P-value = 0.013, CI = [0.011; 0.591]). Due to the sample size differences, we focus solely on the direction of the treatment effect and do not comment on the magnitude of the effect nor explore any heterogeneous effects. Overall, the results show a significant treatment effect on the likelihood of avoiding looking at the picture. In other words, farmers in the treatment group avoid seeing the picture more than in the control group, suggesting that the nature of the image correlates with the decision to view the image or not. The result is robust to the inclusion of control variables.

We run additional robustness checks to account for biases induced by the sample size differences between the control (n = 47) and treatment group (n = 499). We verify the sensitivity of our results to a bootstrap resampling procedure. We resample the data 10,000 times and estimate the odds ratio for each subsample. The results are reported in Table C.1 and Figure C.1 in the Appendix. The bootstrap procedure suggests that the result is not robust to the resampling, with a mean estimated odds ratio of 0.131 (P-value = 0.269, CI = [−0.062; 0.223]). Additionally, and following advice from Gelman and Carlin (2014),¹⁵ we calculate the ex-post power of the experimental design. We report it in Figure D.1, as well as power simulations for different levels of the control group in proportion to the treatment group. As anticipated, the ex-post power is 62.8 per cent, which is below the recommended 80 per cent β level. This compromises our confidence in the treatment effect estimated above, both in terms of magnitude and direction. The two robustness checks confirm the need for additional data to support the conjecture that the nature of the picture affects farmers’ decisions.

4 Conclusion

Public interest in animal welfare is rising, with a particular focus on the well-being of farm animals. While intensive meat production has traditionally been the center of attention (Faucitano et al. 2022), recent concerns are increasingly directed towards the dairy industry (Ritter et al. 2022), notably in relation to the treatment of dairy calves. Specifically, long-distance transportation of unweaned calves is increasingly rejected by the public (European Commission, 2023). An essential element of ensuring optimal animal welfare implies farmers having a comprehensive understanding of the topic and the implications of their decisions. Yet, there are instances where farmers either inaccurately assess the welfare conditions of their animals or eschew available information (Cutler et al. 2017; Amiot and Bastian, 2017).

In this paper, we explored Irish farmers’ engagement with dairy calf welfare during transportation. We presented farmers with a choice of looking at a picture of Friesian bull calves on a lorry or a blank screen for 5 seconds. Engagement with the topic is required if animal welfare is a concern. However, we found that one in five farmers deliberately avoided looking at a picture of dairy calves on a lorry. This finding suggests that there is a significant portion of farmers who play with the moral wiggle room by willfully ignoring calf welfare during transportation.

When exploring the reasons behind the choice of avoiding the picture, we found that many farmers expect the picture to be biased, to feel sad or guilty, or to feel that they already know what transported calves look like. This suggests that some farmers prefer to remain ignorant, driven by distrust in the information and a desire to avoid the picture’s potential negative cognitive impact. Additionally, we explored a treatment effect. In the control group, 47 farmers were randomly selected to view an image of grazing dairy cows, as opposed to calves on a lorry. The assessment of the treatment effect suggested that the nature of the image matters, as the likelihood of avoiding the picture is lower with grazing dairy cows. Although this result needs further data for confirmation due to its underpowered design, it indicates that avoiding the picture of calves on a lorry is not merely a survey artifact.

This exploratory analysis suggests that getting farmers’ attention may be challenging. Given the reasons outlined previously, we foresee the least challenges with farmers who anticipate feelings of sadness or guilt. This group may be more willing to engage, to overcome negative feelings. Positively framing the information could be beneficial, as individuals often eschew information anticipating negative content (Sharot and Sunstein, 2020).

Literature suggests that farmers often misjudge conditions on their own farms (Cutler et al. 2017); therefore, providing information to those who believe they are well-informed may be helpful, at least for some farmers of this group. Specifically, in relation to farmers who state they know what calves on a lorry look like, information could be presented in a manner that offers new insights to capture their interest. Thus, one possible option may be to package the information as new best practices derived from recent research findings. This approach ensures the message is new and helps mitigate any unintended implications that farmers may feel accused of substandard animal welfare practices on their farms.

Farmers who anticipate biased information are likely the most challenging group to engage. In this case, it appears crucial to improve farmers’ trust in official sources, given that trust is a key facilitator of effective learning (Buck and Alwang 2011). Consequently, information intended for this group should ideally originate from sources that farmers trust and perceive as supportive.

Broadly speaking, additional research is needed to validate the findings of this exploratory study. In particular, we believe that our results capture a lower-bound estimate of the willful ignorance among Irish farmers. First, farmers self-selected in the survey experiment. The survey was advertised as “Survey on dairy farmers’ breeding choices” and was diffused to farmers through an online newspaper ad and farm advisors. This approach likely introduced some bias, as farmers could decide to participate (or not) in the survey knowing the survey’s primary focus. It is possible that only farmers who had an interest in breeding participated in the survey, while others may have abstained to avoid new information, thus preserving their willful ignorance. Second, due to the likely presence of experimenter demand effect (Zizzo 2010), sociability desire bias, and lack of revealed preferences on the task,¹⁶ we cannot rule out that some farmers might have decided to see the picture to conform to the survey procedure or to please the researchers.

Our study provides a good starting point for investigating information avoidance concerning animal welfare and offers multiple directions for extending this research. Future studies could aim to replicate the findings with a representative sample of farmers, sufficient treatment and control group size, and a revised version of the task to reduce the impact of experimenter demand effects and social desirability bias. Future work could also focus on precisely mapping farmers’ behavioral and cognitive biases to maximize the efficiency of information provision and guide farmers toward improved animal welfare practices. This should include more precise reasons for avoiding the picture (e.g., “I found a picture more appealing than a blank screen”) and include symmetrical reasons to better compare across decisions. Alternatively, future research could investigate the conditions under which farmers engage with new information and, upon engagement, which aspects of the information they retain

Nevertheless, the findings in this paper represent an initial step in evaluating dairy farmers’ engagement with calf welfare during transportation. In conclusion, the issue of dairy calf welfare is expected to remain a critical concern for dairy farmers, particularly in light of growing consumer awareness and interest in farm animal-rearing practices.

Data Availability

Here is the link to a GitHub repository: https://github.com/ThibautArpinon/Farmers_calf_welfare.git. The repository is public so that anyone can access it.

Footnotes

References

Appendix A. Survey

B. Logit Regression Analysis Results

C. Bootstrap Robustness Check of Treatment Effect

D. Ex-Post Power Calculations