Sample description

Of the 747 participants in our final data set, 410 were Canadian, 381 were female, 441 had children, 416 had completed less than four years of post-high school education, 492 were at least 45 years old, and 675 were meat-eaters. Approximately equal numbers of participants were allocated to each of the three scenarios (Table 2).

Table 2. The number of participants in relation to key demographic categories: country of residence, participant age, gender, education, and dietary preference. Numbers are shown separately for participants allocated to each of the three scenarios (blind chickens, insentient animals, and laboratory meat)

¹ 361 of these participants identified as male, one identified as gender variant/non-conforming, one as transgender, and one as ‘not listed’

² This category includes all responses other than ‘I eat meat’

Quantitative results

Participants expressed negative attitudes toward the blind chicken (mean [± SEM] acceptability of 2.5 [± 0.1]; where 1 = very wrong to do, and 7 = very right to do) and insentient animal scenarios (2.6 [± 0.1]), and ambivalence towards the laboratory meat (3.7 [± 0.1]). To some extent, the insentient animal and cultured meat scenarios can be considered similar, as they both result in the production of animal protein without risk of animal suffering. From this perspective, it seems counter-intuitive that the respondents in this study would perceive the insentient animal scenario more negatively. Further work is required to understand this result, but it is possible that respondents were perceiving morally relevant distinctions between these cases. For example, participants may have been concerned that the insentient animals were also unable to experience pleasure or other positive emotions, and this inability to experience positive states may be eliciting moral concern. Research on robot ethics suggests that physicality is also important. For example, Darling (Reference Darling2016) explored attachments to virtual life forms with robots that can be physically engaged with, concluding that people are more likely to relate to, and form emotional attachments with, those who inhabit physical space and have characteristics we recognise as belonging to a sentient life form. Thus, physicality can affect our assessment of whom we owe moral consideration (Darling Reference Darling2016; Heller Reference Heller2016), and in the current study the physical nature of the insentient animals may have provoked greater moral concern than the cell culture.

We found univariable associations between participant support for scenario and gender, if they had children, support for animal agriculture, and if they ate meat, so our final model included all four covariates. However, in the full model, the effects of gender and children (and interactions between these variables and scenario) were not significant, so these are not reported further. We did find an interaction between scenario and support for animal agriculture (F _2,732 = 9.66; P < 0.0001), and some evidence of an interaction between scenario and whether participants ate meat (F _2,732 = 2.84; P = 0.0591), so these covariates were explored separately by scenario. For participants assigned to the blind chicken and insentient animal scenarios, those more supportive of animal agriculture, were less supportive of the technology (slope = –0.43, SE = 0.09, t₂₅₄ = –4.58; P < 0.0001; and slope = –0.38, SE = 0.09, t₂₄₅ = –4.03; P < 0.0001, respectively). For participants assigned to the laboratory meat scenario, we found no evidence of an association between support for animal agriculture and support for the technology (slope = 0.15, SE = 0.11, t₂₃₆ = 1.40; P = 0.1641) (Figure 1[a]). We suggest that participants most supportive of animal agriculture were more engaged with the questions this study raised, and thus more likely to express a firm opinion, than those less supportive of agriculture or those generally ambivalent toward agricultural practices. These results also suggest that people most supportive of animal agriculture are least willing to accept changes to these animals but are perhaps more ambivalent towards the creation of new products such as lab meat. Additionally, people most supportive of animal agriculture may be least willing to accept the use of genetic technologies on these animals. Some people supportive of agriculture may still recognise the welfare and environmental consequences of certain practices and thus view the BC and IA as “the wrong kind of solution to the problem” (Devolder Reference Devolder2021; p 15); for these participants, lab meat may be perceived as a way of reducing harms and a viable alternative for the market.

Figure 1. Mean (± SEM) participant responses to the question, “In your opinion, do you think it’s right or wrong to [use this technology] to deal with these problems?”, varying from 1 = ’Very wrong’ to 7 = ’Very right.’ Responses are shown separately by scenario (i.e. each of three proposed technologies: blind chickens, insentient animals and creating laboratory meat), and in relation to whether participants (a) expressed support for animal agriculture, (b) felt that the technology respected the dignity of life, and (c) believed that the technology was currently available on the market.

Participants who ate meat and those who did not were similarly supportive of the blind chicken and insentient animal scenarios (t₂₅₄ = 1.85; P = 0.0649; and t₂₄₅ = 0.017; P = 0.7863), but non-meat-eaters were more supportive of laboratory meat than were meat-eaters (averaging 4.8 [± 0.34] versus 3.5 [± 0.12] on the seven-point scale; t₂₃₉ = 3.62; P = 0.0004).

People who believed that the technology failed to respect dignity of life were more likely to reject its use; this result held true for the blind chicken (slope = 0.85, SE = 0.06, t₂₅₅ = 13.49; P < 0.0001), and insentient animal (slope = 0.75, SE = 0.07, t₂₄₆ = 10.07; P < 0.0001) scenarios, and especially so for the laboratory meat scenario (slope = 1.02, SE = 0.06, t₂₄₀ = 16.41; P < 0.0001) (Figure 1[b]). The interaction suggests that although few participants in the cultured meat scenario felt that it was disrespectful, those who did feel this way were especially affected by this concern.

Violations of dignity might be interpreted in various ways, including as a reflection of excessive instrumentalisation. In Swiss animal welfare law, the concept of dignity is used as a focal point for understanding the ethical appropriateness of interventions in the lives of non-human animals, including animal experimentation (Bolliger Reference Bolliger2015). Part of the definition of dignity used to inform practices includes violations of dignity through “humiliation, excessive instrumentalisation, and substantial interference with an animal’s appearance or abilities” which includes interacting with animals as objects rather than respecting their own interests and natures (Bolliger Reference Bolliger2015; p 338). Cataldi (Reference Cataldi2002) and Humphreys (Reference Humphreys2016) argue that the moral weight of dignity does not require any psychological states within the animal (i.e. animals need not desire dignity or be aware of any degradation). From this perspective, sentient and non-sentient animals alike may be seen to have their dignity violated.

Humphreys (Reference Humphreys2016) notes that transgressions of dignity take place when an action holds no benefit to the subject; however, the aim of reducing animal suffering was central to the justification for the genetic modifications explored in this study. The concerns in the current study around dignity suggest that some participants did not believe the scenarios were of actual benefit to the animals. We found that participants were less supportive of technologies when they did not perceive these to be beneficial to animals. Overall, 20.9% of participants believed that the treatments benefited the animals; this percentage was lowest for the blind chicken scenario (8.6%) and highest for the lab meat scenario (32.2%), with the insentient scenario perceived as intermediate (22.6%). Participants who perceived the technology to benefit animals were more supportive of its use for all three scenarios (3.0 [± 0.3] vs. 1.4 [± 0.1], F _1,255 = 22.99; P < 0.0001; 2.7 [± 0.2] vs. 1.3 [± 0.1], F _1,246 = 37.46; P < 0.0001; and 3.7 [± 0.2] vs. 2.2 [± 0.1], F _1,240 = 40.57; P < 0.0001, for the blind chicken, insentient animals, and lab meat scenarios, respectively). The purported purpose of genetic technologies (e.g. to benefit human health or food system actors) is known to influence acceptance (Hudson et al. Reference Hudson, Caplanova and Novak2015; Ritter et al. Reference Ritter, Shriver, McConnachie, Robbins, von Keyserlingk and Weary2019), such that technology framed as benefiting some aspect of social good (including the welfare of the animals) is generally more supported. Further work is required to better understand the type of benefits required, and what type of evidence of benefits people find compelling.

Participants who believed that the technology was available on the market showed more support for the creation of blind chickens (slope = 0.39, SE = 0.12, t₂₅₅ = 3.21; P = 0.0015) and for the creation of insentient animals (slope = 0.28, SE = 0.10, t₂₄₆ = 2.81; P = 0.0053), but there was no evidence of a relationship for laboratory meat (slope = 0.11, SE = 0.12, t₂₄₀ = 0.93; P = 0.3525) (Figure 1[c]). That people were more likely to support technology when they believed it was already on the market, could reflect a sense of complacency; believing that a technology was already prevalent may have resulted in participants believing that there was little to be done to change the situation.

Qualitative results

Qualitative analysis of the open-ended text responses was intended to help understand participant rationale regarding their responses to the three scenarios. For each scenario, participants were first told that animals typically experience certain harms as part of conventional agriculture. For example, in the blind chicken scenario condition participants were told that, “Chickens housed in groups can peck one another, sometimes resulting in injury or even death. To deal with this problem, farmers generally remove part of the beak of young chicks without the use of anaesthetic (i.e. pain-control).” They were then told that the scenario was intended to mitigate this harm. In the case of the blind chicken scenario, participants were told, “Alternatively, it is possible to use genetic modification to create chickens that are born blind. Blind chickens live similarly to sighted chickens but are less likely to engage in harmful pecking, eliminating the need to partially remove their beaks.” Thus, our study design was intended to encourage participants to balance the perceived harms associated with current practices against any new harms associated with the proposed technology. Given this framing, we expected participants to use a type of trade-off logic in their responses; expressions of or trade-off thinking was included as our first a priori theme. Our second a priori theme was intended to capture the rejection of this framing, specifically by participants rejecting both the new technology and the status quo. In the sections that follow we illustrate both themes with participant quotes (citation also includes the country of the participant (Canada = CA, United States = US), participant number, and the scenario this participant was assigned to (Blind chickens = BC, Insentient animals = IA, Cultured meat = CM).

Trade-off thinking

The trade-off thinking theme included comments that displayed a cost-benefit analysis of scenarios, where perceived harms and benefits were weighed against one another, for example, in their considerations of the different options offered, or other issues of importance for participants. In all scenarios, participants who supported the proposed solution (i.e. blind chickens, creating insentient animals, and growing cultured meat) mentioned reducing harms associated with the initial problem they were presented with (e.g. feather-pecking, and stress related to handling, transport, and slaughter). For example, one participant [CA 217 IA] stated: “I don’t think humans will stop eating animals and animal products anytime soon, at least not in most countries, so it’s important to do what we can to prevent harm to those animals.”

Responses to the laboratory meat scenario often mentioned that this option was preferable to harming animals. For example, one participant [CA 36 CM] said, “I think [cultured meat] could be the right option because that way animals are not being harmed.” Another [US29 CM] stated, “I’m an animal advocate and vegetarian. I’d rather people eat artificial stuff than torment animals.” Other participants mentioned that cultured meat products were desirable alternatives because they were better for the environment, or for feeding a growing population. In all these cases participants seemed comfortable within the dichotomy used to frame the issue and felt that one of the response options was acceptable. For example, CA 44 CM stated, “Perhaps lab processed [meat] would be better for environment.”

Some participants applied trade-off thinking but expressed some unease in doing so. For example, one participant [US 4 IA] explained: “I can completely see why this would be beneficial, but at the same time, it feels very ethically wrong.” Similarly, US 266 IA stated that, “It would be better to create something that doesn’t know or feel what is about to happen”, but still questioned whether it was right “to mess with living creatures’ lives?” Another participant [US 317 BC] commented that [creating blind chickens] “Seems bad but I guess it’s for a good reason. Rather have blind chickens than them being pecked to death by each other.” These responses appear to illustrate unease in accepting seemingly logical proposals to alter animals to reduce animal suffering in animal agriculture, when the moral intuition is that adopting such proposals is wrong (Thompson Reference Thompson2008; Palmer Reference Palmer2011).

A number of participants expressed discomfort with the premise and the limited range of response options. For example, one individual [CA 427 BC] questioned the morality of systems that would lead to such choices, stating “This is what mass-production does. I’m sure there would be a fight in any circumstance that 100 creatures were packed into a tight cage”, but concluded that, given the seemingly dichotomous response options of either accepting the status quo or the new technology, that they would “go with the [new technology]”. This participant recognised that the genetic application could provide immediate relief but their response points to the inability of this choice to engage with systemic issues (Murphy & Kabasenche Reference Murphy and Kabasenche2018). Some participants called for something to be done for animals in agricultural systems (e.g. stating that it is “…a travesty that animals suffer” [CA 20 IA]), that there are likely other options available to address suffering (e.g. “we could also just outlaw the small, confined quarters” [CA 78 BC]).

As discussed by Macnaghten (Reference Macnaghten2004), certain types of framing (e.g. the dichotomous presentation of choices existing within the status quo) “may inadequately accommodate both the range and the potential novelty of ethical concerns raised by animal biotechnology” (p 537), including concerns related to the instrumental use of animals and “wider unease about science, about technological modernity, and about hubris.” (p 533). Our results suggest that some participants engaged with the choice between the new technology and the status quo, but others baulked at the dichotomy, for example, explaining that their acceptance of the new technology was conditional on these two choices being the only ones available. This result suggests that proponents of technology may benefit by framing issues in a dichotomous manner, but this approach can also create a false sense of consensus. In a forced choice scenario, participants may simply be picking the ‘least bad’ option; this does not indicate that either option is considered acceptable or attractive. This line of reasoning would support the use of questions in future studies that explicitly ask participants to rate the acceptability of both options independently, rather than indicate if one is more acceptable than another. Even without this option, some participants in the current study explained in their qualitative response that both options were unacceptable, as we describe in the following section.

Rejecting both response options

Some animal ethicists have argued that, despite the moral unease associated with dis-enhancements like creating individual animals without sight or sentience, these technologies should be adopted given that the status quo results in suffering of “actual, living and breathing animals” (Thompson Reference Thompson2008; p 311). Participants in the current study sometimes appeared unwilling to accept this logic and rejected both options provided, for example, by rejecting production methods that would make these technologies necessary. One participant [US 54 BC] in the blind chicken scenario said that the proposed solution to the problem of feather-pecking still involved “inhumane treatment of the animal” and questioned how this “could… possibly be better for the animal?” Another [CA 148 BC] simply stated that, “Either approach is wrong.” These participants seemed to question the contention that these interventions could be of actual benefit to the animals. Similarly, Murphy and Kabasenche (Reference Murphy and Kabasenche2018) ask “whether the changes being proposed in animal disenhancement are truly for the sake of the animals, or whether they represent an attempt not to change most of the other features of the system” (pp 228–229).

Some participants disagreed that interventions like creating blind chickens and insentient animals were the only or most viable options. For example, CA 278 BC said “…I believe there has to be another way”; a sentiment echoed by CA 289 BC who suggested that the problem “Could be dealt with [using] a different option.” With respect to the issue of feather-pecking, several participants suggested improved rearing conditions. For example, CA 215 BC stated that a “better solution would be not to house chickens in such close confines rather than developing needless genetic modifications.” Similarly, CA 159 BC wrote, “Don’t raise chickens in that sort of environment. The problem can be prevented!” Murphy and Kabasenche (Reference Murphy and Kabasenche2018) argue that if “the starting point for justifying disenhancement is the reduction of suffering, there are many ways we could achieve that” (p 228), including modifying environments. Comments from our participants express what Devolder (Reference Devolder2021) identified as the “wrong kind” (p 3) of solution to address suffering. This author looked to the side-effects of two choices to reduce animal suffering: using genome editing to create pigs resistant to a respiratory disease, or giving them more space to reduce disease transmission, to illustrate why the latter is the preferable option. Increasing space provides a solution to more than one problem; it decreases disease transmission and, by decreasing stressful interactions that arise from dense housing, it increases animals’ well-being. While some participants were not convinced of the proposed dichotomy or felt that there were preferable options available for alleviating suffering, others specifically called for systemic change. For example, CA 364 IA wrote that “Animals shouldn’t have to change; people and their methods should.” Similarly, CA 67 CM asked that “great minds come up with a better way” of doing things. Thus, at least some people were unwilling support these types of interventions do so because they do not see these as acceptable approaches to the problem. Indeed, while there were important differences between the blind chicken and insentient animal scenarios, most notably that the latter scenario described animals with no welfare to be considered, participants responded to these interventions as similarly undesirable perhaps because they considered both to be the wrong type of solution to the underlying ethical issue.

Harfeld et al. (Reference Harfeld, Cornou, Kornum and Gjerris2016) speak to the elements of industrial agriculture that prevent people from seeing both the moral relevancy of animals and the systemic flaws including how animals are spoken about in agricultural systems (i.e. as extensions of production machinery), how they are housed and treated, and the incremental measures that are sought to improve their lives including those tethered to increases in production efficiency. Devolder (Reference Devolder2021) suggested that a desire to reduce a sense of complicity is one reason why some people may reject interventions like the ones our participants encountered, when viable alternative ways to address issues exist. According to this way of thinking, these narrowly framed solutions make people morally complicit in the wrongness that the system is seen to perpetuate.

Schultz-Bergin (Reference Schultz-Bergin2014) makes a distinction between approaches that examine rational choices in a non-ideal world (e.g. where status quo production systems continue to exist because of resistance to change within the industry), versus more ideal-world thinking that includes rejecting status quo conditions and advocating for changes in production methods to avoid creating welfare problems in the first place. Visioning more ideal systems allows us to question the “assumptions that gave rise to the problems in the first place.” (p 204). Thus, it is important to distinguish between approaches that seek to confront the acceptability of a technology more narrowly in the context of a non-ideal, non-changing world, versus more generally “reflecting on what sort of world we want to create and the role such idealisation plays in evaluating our actual world” (p 16). If a person’s ideal is a world without industrial agriculture, for example, then genetic interventions like creating blind chickens and insentient animals are unlikely to be perceived as helpful. From a psychological perspective, reasoning styles differ between people, based on what values they hold, with consequences for decision-making (Billet Reference Billet2021). For people with reasoning styles governed by sacred values, a deontological style of reasoning means that trade-offs offered between the sacred (in this case animals) and the non-sacred (production efficiency) will be viewed as taboo and morally impermissible (Tetlock Reference Tetlock2003). In contrast, those holding trade-off reasoning styles, are more likely to be influenced by instrumental values (Atran Reference Atran2016). In addition, our results suggest that framing genetic technology as the only alternative to the suffering experienced by animals within the status quo system of agriculture is viewed as too narrow by some participants, in part, “because actions occur within a context, and this context bears on our understanding” of actions (Schultz-Bergin Reference Schultz-Bergin2014; p 7).

According to Macnaghten (Reference Macnaghten2004), people contextually develop ethical positions towards animals and how they are treated, related to how technologies symbolise and reflect “specific societal values and assumptions” (p 534), including values and assumptions about what kinds of relationships society ought to expect between humans and animals. Some participants in our study suggested that both the human conduct and the human-animal relationship represented in the scenario were ethically problematic. These participant comments correspond well with discussion in the academic literature; Kramer and Meijboom (Reference Kramer and Meijboom2021) and Sandøe et al. (Reference Sandøe, Hocking, Förkman, Haldane, Kristensen and Palmer2014) discuss animal integrity as a morally relevant source of unease when considering the creation of new strains of animals through genetic interventions, like blind chickens. For example, CA 66 IA suggested that “Just from a moral position, it should be inhumane to change how animals behave just to suit our needs.” Similarly, CA 91 BC stated that creating blind chickens is “ethically wrong and [a] selfish thing to do”, and US 460 BC stated that this is a “self-serving way to raise chickens, depriving them of their sight.” Murphy and Kabasenche (Reference Murphy and Kabasenche2018) suggested that disenhancing animals reinforces principles of domination. Thompson (Reference Thompson2008) described some objections to genetic disenhancements as rooted in concerns about the virtue of those who would do this, saying performing such genetic modifications can be seen as evidence of the “vices of pride, of arrogance, of coldness and of calculating venality” (p 314). Some participants in the current study also referred to such character flaws (e.g. “Genetically altering animals to make them easier to kill is just plain evil.” [US 284 IA]).

Interestingly, the qualitative theme of rejection of both options was observed only twice for the cultured meat scenario; the rarity of this theme for this scenario might be due to participants viewing this technology as a way of rejecting conventional animal agriculture.

Animal welfare implications

Thompson (Reference Thompson2008) noted that social repugnance to proposed interventions for solving animal welfare issues in industrial farming could provide impetus to change circumstances for animals.

Our study shares the perspectives of people who, faced with contentious methods of creating new animals and animal products to address issues within current production systems, sometimes reject these options, and instead call for system-level change. Thus, our study design illustrates a method for understanding when people may be willing to accept changes within a system versus when they will call for more transformative changes in animal agriculture. Work that frames issues as trade-offs within a system may discourage transformational thinking, limiting our understanding of people’s concerns and thus also limiting the development of policy and practices that improve the lives of farmed animals.