What Makes a Protest Movement Successful?
Summary
Social Change Lab’s previous work on protest outcomes found that Social Movement Organisations (SMOs) that use protest as a main tactic can significantly impact public opinion, voting behaviour, public discourse and to a lesser degree, policy outcomes. Despite this, it’s not clear what strategies and tactics these movements should pursue in order to maximise their impact on these outcomes.
​
As a result, we conducted a 6 month-long research project looking into the factors which make some protest movements more successful than others, focusing primarily on Western democracies, such as those in North America and Western Europe. In this report, we provide a synthesis of the research that we conducted using a variety of research methods. These methods include a literature review, public opinion polling, expert interviews, policymaker interviews, and a case study. This report specifically focuses on protest-based movements, movements that use protest as a key tactic, as opposed to social movements more broadly. For example, this report is more aimed at how the climate movement or animal advocacy movement could improve, rather than an intellectual social movement such as effective altruism. That said, we believe some factors will overlap for both protest-focused movements and non-protest focused social movements.
​
We believe this report could be useful for grantmakers and advocates who want to pursue the most effective ways to bring about change for a given issue, particularly those working on climate change and animal welfare. We specifically highlight these two causes, as we believe they are currently well-suited to grassroots social movement efforts.
​
The report is structured so we only present the key evidence from our various research methods for the success factors we believe are the most important, such as numbers, nonviolence and diversity. Full reports for each research method are also linked throughout for readers to gain additional insight.
Summary Table
Below, we scored success factors on two different scales, one to highlight the causal importance of the success factors to achieving desired outcomes, and another to highlight the strength of evidence behind our claim. Our magnitudes of expected effects are given relative to other factors. However, there are often many different outcomes that movements or specific social movement organisations (SMOs) are optimising for, such as changing public opinion, policy, public discourse, and so on. We present our expected effect sizes assuming that one could translate them all into a unified metric of success.
The ‘strength of evidence base’ column also expresses the relative strengths of the evidence bases. For example, ‘Strong’ indicates high levels of agreement between all relevant academic studies on this topic, as well as supporting evidence from expert interviews and other research methods. ‘Weak’ might reflect substantial disagreement in the academic literature, there being few empirical studies on this topic or that the evidence was derived solely through expert interviews. The evidence related to each outcome, and our estimations regarding its importance are explained further in the sections below.
​
​Table 1: Our current estimates, and evidence bases behind these estimates, for the relative causal importance of different success factors.
Contents​
1. Executive Summary
1.1 Illustrative findings
1.2 How to read this report
1.3 Scope of this report
2. Introduction
2.1 Why we think this work is valuable
2. 2 Research Questions & Objectives
2.3 Audience
3. Methodology
4. Violence
4.1 Literature Review
4.2 Expert interviews
4.3 Policymaker interviews
4.4 SHAC Case Study
5. Numbers (size of protest or protest movement)
5.1 Literature Review
5.2 Expert interviews
5.3 Policymaker interviews
6. External Factors
6.1 Elite Allies
6.2 Public Opinion
6.3 Luck
7. Diversity
7.1 Literature Review
7.2 Expert interviews
7.3 Policymaker interviews
8. Other potentially important factors
8.1 Commitment and Unity
8.2 Radical Flank Effect & Disruption
8.3 Trigger Events and Protest Timing
8.4 Intra-organisational factors (e.g. governance, systems and team experience)
8.5 Protest frequency and protest length
9. Future work that could be promising
10. Limitations
11. Conclusion
Acknowledgements
Appendix A - Methodology
Illustrative findings
​
Below are some of our notable findings, that highlight some of the summary claims in the table above:
-
Teeselink and Melios (2021), when analysing the impact of Black Lives Matter protests in 2020 on voting behaviour, estimate that a percentage point increase in the fraction of the population that goes out to protest (i.e. larger protest size) raises the share of Democratic votes by 5.6 percentage points.
-
Wasow (2020) finds that during the US Civil Rights Movement, non-violent protest was more successful in increasing Democratic vote share than was violent protest, which had negative consequences. He found that nonviolent protest resulted in a 1.6 percentage point higher Democratic vote share relative to a ‘control’ county with no protest, whereas violent protests increased votes for Republicans by 2.2 - 5.4 percentage points (in opposition to the aims of the Civil Rights movement).
-
Wouters (2019), using an experimental approach, finds that the diversity of a protest is one of three factors that influence the public to increase support for an issue. Worthiness (or nonviolence) and unity also led to increases in public support for a cause after being exposed to protests, whereas changes in commitment of the protestors or size of protest did not.
-
Simpson et al. (2022) uses an experiment to find that radical tactics, and not radical goals, can increase support for more moderate groups. This study, examining the climate and animal advocacy movements, shows that radical tactics can boost support for more moderate groups without harming overall support for a movement’s policy goals.
1.3 How to read this report
In this report, we summarise the key findings from various research methods (e.g. our literature review), rather than replicating them in full. For example, in the violence section, we reference the papers with the most valuable and relevant evidence from our literature review, rather than all the papers that examine violence as a strategy. For those who want to read more into a particular success factor or methodology, we encourage you to read the full, linked reports. All sections are intended to be standalone, which means that some papers, which support multiple factors, are repeated in several sections.
1.4 Scope of this report
Rather than this report being an exhaustive handbook of what protest-focused movements should do to be successful, it is a summary of the current available evidence. Due to some factors being easier to measure relative to others, such as the importance of protest size relative to internal culture, we believe there is a bias in this report for factors that are more measurable.
Additionally, as identified in our protest outcomes report, there are often many different outcomes that movements or specific social movement organisations (SMOs) are optimising for, such as changing public opinion, policy, public discourse, or more. Due to this, it can be hard to compare some factors to one another, as they might be trying to optimise for different outcomes. Throughout the report, we note where specific factors might be more successful in achieving some outcomes relative to others.
This report covers high-level movement strategy, such as decisions around how much relative effort movement leaders should put into size, diversity, and internal governance. We don’t cover more on-the-ground details on how to run effective campaigns, as we think this has been covered fairly well in other pieces of work. For this, we refer those interested in the nuts and bolts of how to campaign effectively to the following resources: Activist Handbook, The Commons, Campaign Bootcamp and Effective Activist to name just a few.
Additional limitations can be seen in our Limitations section.
Introduction
2.1 Why we think this work is valuable
In our report on protest outcomes, we found that Social Movement Organisations (SMOs) that use protest as a main tactic can significantly impact public opinion, voting behaviour, public discourse and to a lesser degree, policy outcomes. We think that if our findings on protest outcomes are accurate – that protests can significantly alter public opinion and/or affect policymaker’s beliefs – then there is a strong case that philanthropists and social change advocates should be considering this type of advocacy alongside other methods.
Additionally, we believe this is an especially neglected area of research, given that protests are an extremely commonly used tactic for achieving social change. For instance, we believe that we are the first to conduct a (preliminary and uncertain) cost-effectiveness analysis of a social movement organisation, and the second to commission bespoke public opinion polling to understand the impact of protest on public opinion. We see this as an opportunity to add value by providing research that addresses some unanswered questions about the role of social movements in improving the world.
2.2 Research Questions & Objectives
Primary research questions we have been investigating for this report:
-
Which tactics, strategies and factors of protest movements (e.g. size, frequency, diversity, etc.) most affect their chances of success of achieving their desired outcomes? Such as changing public opinion, influencing policymakers, shifting public discourse, and so on.
-
How important are factors within a movement or organisation’s control, such as their strategy or structure, relative to external societal conditions, such as pre-existing public opinion or friendly elites?
Other questions we are tackling or will tackle in the near future can be seen in our section on future promising work.
2.3 Audience
We think this work is valuable to two audiences in particular:
-
Philanthropists seeking to fund the most effective SMOs - For example, if we discover that SMOs are generally more successful if they incorporate good governance and strict nonviolence, then grantmakers should fund these organisations (all else being equal) relative to SMOs who don’t.
-
Existing social movements - who can employ and integrate best practices from social science literature to make their campaigns more effective. This includes Effective Altruist community-builders, who thus far seem to have made relatively little effort to learn from previous social movements and examine why and how they succeeded or failed.
Methodology
As a preliminary step, we sought to understand the academic literature pertaining to protest movement success factors. We conducted a comprehensive literature review to identify key papers and analyse their robustness. Our literature review found several success factors that were validated by multiple compelling academic studies. However, the overall evidence base within the academic literature was relatively small, and we had to rely on other methodologies to supplement our conclusions. For instance, we only found three relevant studies that examined the importance of diversity and unity, compared with 11 seemingly robust studies on the question of whether nonviolence or violence is generally a more effective strategy.
We decided to tackle subsequent research by approaching our research questions from many different angles and using several different methods. We believe this is a more robust way of tackling our research questions, as no single methodology provides an empirically foolproof approach, given the limited evidence base. Instead, we attempted to evaluate the evidence base for social movement success factors using a variety of different methods, to understand where this broad evidence base converges or diverges. We explain the methodologies in-depth in Appendix A and in the full reports, but in summary, the research methods we used are shown in the table below. The limitations of our overall methodology are addressed in more detail in the Limitations section, with specific research-method focused limitations in our individual reports.
Throughout this report, we focus on social movements that use protest as a main tactic within their activities. We define protest fairly broadly, using the classification provided by Beer (2021), which finds 346 different methods of civil resistance.
Table 2: This table highlights the various methodologies used to produce the findings in this report.
4. Violence
Whether violent protests are more or less likely to succeed than nonviolent protests is a hotly debated topic among those who study social movements and protest activity. The question is an important one - it is useful to know whether there is some downside to activists who only carry out nonviolent activities. The evidence suggests that nonviolent protests are more likely to be successful than violent protests.
To clarify, we are referring only to protests that are clearly violent, such as attacks against individuals, protests where people are threatened with physical harm, rioting, etc. Additionally, most of the literature we examine treats property destruction or damage to objects as violent, so this is reflected in our work (even though this is a contentious distinction). We are not referring to protests that are disruptive but nonviolent, although the literature occasionally conflates violent protest with disruptive protest. We think evidence against violent protest should not necessarily be taken as evidence against nonviolent yet disruptive protest.
Most of the available literature suggests that violence reduces the chance that a protest movement will succeed, and our conversations with academics generally corroborate this view. That being said, we are not claiming that this is a settled debate, nor that all of the literature is consistent on this, but we are claiming that the majority of high-quality studies indicate that nonviolent protests are more likely to be successful than violent protests. We should add that the question of what is considered violent vs nonviolent is a highly ethical one, and there is widespread disagreement on the topic. As a result, both researchers and advocates may have biases that influence both research on this topic, as well as how to apply research on this topic to social movement activities.
4.1 Literature Review
In our Literature Review on factors that make protests more likely to succeed, we found that the consensus in the literature was generally that nonviolent protests are more likely to succeed than violent protests. Studies that showed that nonviolent tactics are more likely to lead to positive outcomes were:
-
Wasow (2020) demonstrates with an instrumental variable approach that during the US Civil Rights Movement, non-violent protest was more successful in increasing Democratic vote share than violent protest. He found that nonviolent protest resulted in a 1.6 percentage point higher Democratic vote share relative to a ‘control’ county, whereas violent protests increased votes for Republicans by 2.2 - 5.4 percentage points (in opposition to the aims of the Civil Rights movement).
-
Research from Feinberg et al (2017) also showed that extreme activism (which was usually violent action), resulted in reduced public support for the issue being protested and fewer people identifying with the protest movement.
-
Work by Muñoz and Anduiza (2019) supports the hypothesis that nonviolent protests are better at building public support for a social movement. They examine the 15-M anti-austerity protesters in Barcelona, exploiting a sudden shift from nonviolent activity to violent riots and protests in 2016. Surveys were conducted both before and after the sudden shift from nonviolent to violent activity. The level of support for 15-M decreased by 12 percentage points after the shift, from 65% support to 53% support. It should be noted that the fall in support was concentrated among voters of parties that did not have sympathy for the movement to begin with, whereas the fall in support among ‘core supporters’ was smaller. The figure below shows the interaction effect between an individual’s past vote and change in support for the movement.
Figure 1: The interaction effect between past vote and change in support for the protest movement, after the public being exposed to riots. Source: Muñoz and Anduiza (2019)
-
Wouters and Walgrave (2017) conducted an experiment which exposed Belgian legislators to different news stories showing protest activity, manipulating whether the stories depicted protesters as ‘worthy’ (serene atmosphere, calm and peaceful demonstrators, etc.) or ‘unworthy’ (disruption initiated by protesters, broken shop windows shown, etc.). They found that legislators shown ‘worthy’ protesters were more likely to take the position of the protesters, although they were no more likely to say:
-
that they would actually take action in support of the protesters or
-
that they found the issue they were highlighting to be particularly salient (see Figure 1 below).
Figure 2: Predicted Values of Salience, Position, and Intended Action Effects by Worthiness, measured in Belgian legislators. Source: Wouters and Walgrave (2017).
Whereas the following studies showed that violent tactics can have positive outcomes, at least in some cases:
-
Shuman et al. (2022) analysed the impact of violence during the Black Lives Matter protests in 2020, using three groups: people living in areas with no violent protest, people in areas with only nonviolent protests, and people living in areas with both. They found that protests involving both violent and nonviolent activity were the most likely to be successful in persuading conservatives in relatively liberal areas (a group who is relatively resistant to BLM’s policy goals).
Figure 3: Support for the policy goals of Black Lives Matters, by political ideology and the combination of violent and nonviolent protests in a particular area. Source: Shuman et al. (2022)
The regression analysis comparing all treatment conditions to the control condition indicated that most of the conditions were associated with significantly lower willingness to act than the control condition (see Figure 6): horse racing disruption with solution-led messaging (estimate=-0.18, 95% CrI [-0.33, -0.02]), KFC blockade with values/norms-led messaging (estimate=-0.19, 95% CrI [-0.35, -0.03]), KFC blockade with problem-led messaging (estimate=-0.27, 95% CrI [-0.43, -0.10]), KFC blockade with solution-led messaging (estimate=-0.30, 95% CrI [-0.46, -0.15]), open rescue with norms/values-led messaging (estimate=-0.28, 95% CrI [-0.43, -0.13]), open rescue with solution-led messaging (estimate=-0.23, 95% CrI [-0.39, -0.08]).
Figure 6. Forest plot showing the estimated effects of each experimental condition relative to the control condition, along with 95% confidence intervals around the estimates.
​Thus, similar to the results above regarding Animals Solidarity scores and support for AR’s policy goals, willingness to act was negatively affected by nearly all experimental vignettes relative to the control condition. Again, the horse racing campaign was associated with milder negative effects.
​
Discussion
Overall, the results suggest that people’s attitudes are negatively affected by exposure to disruptive animal rights protests. Most of the experimental conditions either trended towards or had a significantly negative effect compared to the control condition. This pattern was most pronounced for people’s willingness to act, followed by their support for Animal Rising’s goals. Only when looking at Animal Solidarity scores were there two conditions (horse race disruption with values/norms-led messaging and problem-led messaging) with a significant positive effect.
​
Animal Rising disrupted the 2023 Grand National horse race, triggering a considerable media reaction and a debate on the use of animals for entertainment. We previously reported on a poll before and after the protest to assess its public opinion impacts. The evidence suggested that people’s attitudes toward animals worsened as a function of how much they had heard about the protest, suggesting that the protest negatively impacted public opinion. Here, we sought to investigate in a controlled experimental setting how different protest campaign types and messaging strategies affect people’s attitudes towards animals, their support for Animal Rising’s goals, and their willingness to take action for animals. We looked specifically at people whose attitudes towards animals are already quite positive (score 5 or higher on the Animal Solidarity scale) but who are neither vegans nor engaged in animal advocacy. These people constitute the primary target group for animal rights groups such as AR, because they are the most likely to develop more pro-animal views and behaviours if nudged in that direction.
​
Our results revealed interesting differences between the different campaign types and messaging strategies. Overall, the data suggest that horse racing campaigns have less of a negative effect than KFC blockades and open rescues, and that norms/values-led messaging is more effective than problem-led and solution-led messaging (even though it too was generally associated with negative effects relative to the control condition). Solution-led messaging was associated with the most negative effects across measures. In more detail, the results indicated that regarding Animal Solidarity scores, norms/values-led messaging and the horse race disruption campaign had more positive effects than the other conditions. By contrast, KFC blockades were associated with the lowest Animal Solidarity scores, especially in the context of solution-led messaging. Regarding support for AR’s goals, horse racing campaigns again did relatively better, with open rescue campaigns doing particularly poorly, and solution-led messaging again having the strongest negative impact. Regarding willingness to act, most conditions had robust negative effects relative to the control condition. Again, horse racing campaigns had more positive effects compared to the other campaign types.
​
What can account for the differences between messages? It is well established that most people are strongly influenced by what others do or think (Cialdini & Goldstein, 2004; Miller & Prentice, 2016). This is in part because people’s actions and beliefs are cues to social norms (Asch, 1955). Perceived norms are highly dynamic (Sparkman & Walton, 2017), making them an attractive target for framing effects. Indeed, previous studies suggest that messaging strategies that highlight societal norms can be very effective at inciting pro-environmental behaviour (Cialdini et al., 1990; Cialdini & Goldstein, 2004; Goldstein et al., 2008). Moreover, framing effects can be exploited in order to appeal to people’s core values (Dixon et al., 2017). For instance, climate change deniers can be motivated to engage in pro-environmental behaviour if climate action is framed in terms of the care we feel towards one another or the technological (and other) improvements we can make as a society (Bain et al., 2012). It is likely that the values/norms-led messages had less detrimental effects compared to the other messaging types because they combined the strengths of normative and value-based messaging strategies, thereby offsetting the general negative effect of vignettes about animal rights protest.
​
What can account for the differences between campaign types? We see two main alternative explanations: 1) In April 2023, Animal Rising disrupted the Grand National horse race and received so much media attention that an estimated 50% of the UK population heard about it. Those participants who were exposed to reporting about the protest have had time to think about the rationale behind it so might better understand what the activists are doing in the vignettes presented here. 2) Alternatively, horse race disruptions might inherently be viewed more favourably compared to open rescues and KFC blockades. This could be because delaying a horse race is seen as less immoral compared to stealing animals from a farm or physically blocking people from getting the dinner they desire. Moreover, polling suggests that people are more willing to give up horse racing than eating meat (which is what the open rescue and KFC campaigns are more overtly targeting). A follow-up analysis indirectly supports the latter explanation: we ran a model that again predicted Animal Solidarity scores with the ten vignettes and additionally included a factor reflecting people’s awareness of the Grand National horse racing protest and its interaction with the effect of the vignettes. Explanation 1 predicts a positive interaction effect, such that people who have heard more about the Grand National protest would be affected more positively by the horse race disruption conditions. Instead, there was an opposite trend. Our results suggest that horse race disruptions had more positive/less negative effects in all outcome measures we considered, and they indicated that this cannot simply be explained by the fact that similar high-profile protests have occurred this year. Previous experimental work suggests that negative public opinion effects of radical protest are particularly likely if the protests are perceived to be immoral (Feinberg et al., 2020). Open rescues and blockades might be perceived as more immoral and confrontational, as open rescues can be viewed as stealing and KFC drive-thru blockades physically hinder people from obtaining the food they desire. Additionally, horse race disruptions are likely to be perceived as criticisms of the use of animals for entertainment, which people are more ready to give up or change compared to the use of animals for food.
​
In the present study, as well as in a recent one on Animal Rising’s Grand National protest, we observed negative public opinion impacts immediately following disruptive animal rights protests. The largely negative effects reported here were observed despite the fact that the sample tested in the present study was pre-selected to be relatively favourable towards animals. Our previous nationally representative polling work on the short-term effects of the Grand National protest indicated that the more unfavourable somebody’s attitudes are towards animals, the more negative the effects of disruptive animal rights protests (but see the paragraph below for long-term effects). As such, the current study is likely to underestimate the negative effects on the public at large. In the context of climate protests, such negative effects were not seen in recent studies, with some reporting null effects and others reporting positive impacts (Bugden, 2020; Gonzatti et al., n.d.; Kenward & Brick, 2023). There are several reasons why disruptive protest tactics might have a higher chance of producing negative public opinion effects in the animal advocacy domain, as opposed to the climate movement. First, there is more widespread agreement that climate change is a serious problem and that we need to do something about it. Even though animal farming is directly related to the climate crisis, there generally is relatively low agreement that animal farming needs to change urgently. Relatedly, climate change is a much more salient issue that is being discussed in the media on a daily basis. A recent expert survey suggests that disruption is more likely to have positive effects if the issue has high salience and public support.
Zooming out, the present results confirm our previous polling results in showing that, in the short term, disruptive animal rights protests may have negative impacts. These extend to people’s attitudes towards animals, their support for changing how we treat animals in society, and their willingness to do something about it. However, we recently published new research which suggests that the initial backlash effects due to disruptive animal rights protests do not last. When doing a six-month follow-up looking at the longer-term effects of AR’s Grand National protest, we saw that people’s knowledge of the protest activities was linked with worsened attitudes towards animals immediately after the protest, but not six months later. Moreover, it is likely that any attitudinal changes due to exposure to an actual real-world protest are generally stronger than those triggered by a short vignette describing a protest. Hence, we view it as likely that the negative effects obtained here do not reflect lasting changes in people’s attitudes, but rather a temporary phenomenon mediated by a strong emotional response to the protests. At present it is not known how long such effects are expected to last. Even though these effects are temporary, we believe that the relative differences between different campaign types and messaging strategies provide important insights into possible consequences of activists’ strategic choices. Additional longitudinal data are needed to determine the extent to which AR’s disruptive tactics contribute to pro-animal shifts in the long run.
Methods
Participants
The study focused on animal lovers who are neither vegans nor animal advocates (N = 4757). These people are hypothesised to be those for whom differences in protest tactics are most likely to matter – people who are already animal advocates are likely to already have very favourable attitudes towards animals, while people with low concern for animals are unlikely to shift their views regardless of the protest campaign or message. After all, the intervention used here merely consists of a relatively short text. We ran a screening study on Prolific with 15915 participants with the goal of finding at least 5000 participants who scored (on average) at least 5 out of 7 on the Animal Solidarity Scale (see below), who are not vegans and who have not engaged in animal advocacy in the last 12 months. Specifically, we only selected participants who indicated not having:
-
Attended an event, protest or demonstration related to animal welfare or rights for animals
-
Volunteered with an animal rights or welfare organisation
-
Donated to an animal rights or welfare organisation
​
Out of the 15915 participants, 5683 (35.7%) fulfilled all our requirements and were invited to participate in the main experiment. 4757 completed the survey.
We took two measures to account for participants possibly not being very attentive. Firstly, we included an attention check that said that when asked about their favourite sports, one should select “tennis”. People who did not select “tennis” were excluded from the analysis. Secondly, we had a challenging comprehension check: Participants had to select which out of a number of things were mentioned in the text.
-
Animal activists disrupted a horse race and criticised the exploitation of animals for entertainment and food
-
Animal activists rescued sheep from a farm and criticised the exploitation of animals for entertainment and food
-
Animal activists blockaded KFC drive-thrus and criticised the exploitation of animals for entertainment and food
-
Animal activists disrupted a horse race and encouraged spectators to boo the jockeys
-
Animal activists rescued sheep from a farm and brought them back shortly after
-
Animal activists blockaded KFC drive-thrus and handed out vegan burgers for free
-
A fashion designer is having great success with a unisex mix and match suit
This is a relatively difficult comprehension check because for all treatment conditions, there is one incorrect answer that is quite similar to the correct one. 94.5% passed the comprehension check. We included performance on the comprehension check as a covariate in the analyses to account statistically for potential differences between conditions in how attentive participants were.
​
Design
Participants were randomly assigned to one out of 10 conditions. In each experimental condition, participants read a short description of one out of three campaign types (a horse race disruption, an open animal rescue, or a KFC drive-thru blockade). Then, they were presented with a message; a short paragraph in which a spokesperson makes the case for a move away from a world in which animals are exploited for fun and food. There were three different message types: values/norms-led, problem-led, and solution-led. The values/norms-led messages appeal to the love people feel for animals and the desire people have to do the morally right thing. The problem-led messages stress the repercussions of animal exploitation for the animals and for nature. The solution-led messages highlight a plant-based system as a solution. The messages were constructed such that they were equal in length (+/- 5 words) and only differed in key elements needed to distinguish the different campaign and messaging types. The vignettes resembled how people might read about them in a newspaper. In addition, there was a neutral control condition without mentions of campaigns and without an animal-related message, which talked about fashion trends. This is necessary to evaluate whether messages are better than a neutral baseline condition and whether there are backfire/boomerang effects (conditions performing having a negative effect on attitudes towards animals).
Half of the problem-led and solution-led messages referenced climate issues, the other half focused on animal suffering. This allowed us to test whether one is more effective than the other. However, no differences were observed between the two versions, which is why the statistical analyses forewent this distinctin.
​
The group Animal Rising is occupying livestock farms and liberating animals from farms. They argue that these animals endure high levels of suffering and they want to save them from being slaughtered for food. Animal Rising wants to bring attention to our broken relationship with animals and nature.
Five sheep were taken from Jeremy Clarkson’s farm over the weekend when a group of animal activists broke into the sheds and rescued the animals. Supporters of the grassroots organisation Animal Rising have claimed responsibility for taking livestock animals from the Lickety Spit farm, made famous on Clarkson’s Amazon Prime TV series.
Table 7. The stimuli used in the open rescue conditions. This is presented here to help the reader get an idea of what the different conditions looked like. The full list of stimuli can be found in the Appendix.
Outcome variables
The main outcome variables were participants’ average score on the animal solidarity scale, their support for AR’s demands, and their willingness to act (see below). The Animal Solidarity scale was developed to quantify people’s “sense of belonging, psychological attachment, and closeness felt toward other animals” (Amiot & Bastian, 2017, p. 2). Scores on the scale predict pro-animal behaviours and attitudes above and beyond previous similar scales and are thus a good proxy for how favourable somebody’s attitudes towards animals are (Amiot & Bastian, 2017).
Table: Animal solidarity scale
We measured support for AR’s demands (7-point Likert scale) by asking people about their main demands as stated on their webpage and used the average score as the dependent variable (Cronbach’s alpha = 74.5%):
-
To which extent do you oppose or support the following: - The UK should move away from animal farming and transition towards an animal-free food system
-
To which extent do you oppose or support the following: - The UK Government should commit to rewild the land and ocean as part of a broader programme of wildlife restoration and carbon drawdown
-
To which extent do you oppose or support the following: - The UK should ban factory farming
Willingness to act was operationalised as the average (Cronbach’s alpha = 84.2%) of the 7-point Likert scale scores regarding the following three items:
Several activist groups in the UK are fighting to improve animals' lives, minimise their suffering, and stop them from being exploited for food and entertainment. How willing are you to do the following in the next two months:
-
Donate to an animal rights activist group
-
Participate in a peaceful animal rights protest
-
Write to or phone your MP about animal welfare
Analysis
We carried out Bayesian linear regression analyses (Bürkner, 2017) to test the effect of the different campaigns and messages. The first analysis was a campaign type (horse race disruption, open rescue, KFC drive-thru blockade) X message type (norms/values-led, problem-led, solution-led) regression analysis. For this analysis, the neutral control condition was removed from the dataset to allow for a typical crossed factorial analysis. This analysis was used primarily to test for main effects in order to assess whether certain campaigns or message types are generally more persuasive. The analysis included the following covariates: 1) Accuracy on the comprehension check (correct vs. incorrect). 2) Awareness of the Grand National horse racing protest in April 2023 (Likert scale; 1 = Nothing at all, 2 = A little, 3 = A moderate amount, 4 = A lot, 5 = A great deal). 3) Awareness of the open rescue of sheep in May 2023 (5-point Likert scale as above). One such regression analysis was carried out for each of the three outcome variables (Animal Solidarity scores, support for AR’s demands, willingness to act).
​
The second analysis simply treated all conditions as unique, i.e. it ignored which campaign or message type they belong to (hence, the analysis had a single factor with 10 levels). Here, the control condition was the baseline condition, such that the coefficients give the effect of the treatment conditions relative to the control condition.
Both analyses used mildly informative priors, partly based on the pre-test on Animal Solidarity scores, partly based on the expectation that any treatment effects in the present study should be rather small. In particular, we set the priors for the intercepts to a normal distribution with a mean of 5 and a standard deviation of 1, whereas the priors for all the treatment effects were set to a normal distribution with a mean of 0 and a standard deviation of 0.25. The crossed factorial analysis used sum-to-zero coding and used estimated marginal means comparisons to evaluate to what extent given pairs of conditions differed. The analysis testing the effect of each of the 9 treatment conditions relative to the control condition used dummy-coding with the control condition as the baseline.
About Social Change Lab​
Social Change Lab conducts empirical research on disruptive protest and people-powered movements. Through research reports, workshops, and trainings, we provide actionable insights to help movements and funders be more effective. You can find all of our research projects and resources on our website. You can contact us at info@socialchangelab.org
Appendix
Messages
Note: The bit in the campaign description that is in smaller font came in-between the quotes. It is included in the top part to make it easier to see the different messages.
Version 1 - Climate mentioned, suffering not mentioned
Version 2 - Climate not mentioned, suffering mentioned
