Effective Advocacy For Animals Through Protest

Many animal advocacy groups use protest as a way of gaining media coverage and drawing attention to the plight of animals. Last year, 500 people gathered at Smithfield Market, the largest wholesale meat market in the U.K., as part of an Animal Rebellion demonstration against the meat industry ⁠— and received widespread media coverage. The Humane League has held protests outside of McDonald’s stores to draw attention to large-scale animal cruelty. Animal Charity Evaluators has estimated that there are 40 to 80 animal advocacy protests per week in the U.S. alone. 

But do protests actually have any impact? 

There is some evidence that they can: on public opinion, on voting habits, on legislation, and on public discussion around an issue. That being said, some people worry that protests can actually backfire ⁠— that they draw negative media attention and turn people away from the issues that the protesters are highlighting. This is definitely a valid concern, and the evidence suggests it’s probably true that certain kinds of extreme protest can backfire. We also want to try to figure out what factors make protests more likely to be successful, and the research we’ve done indicates that protests are more likely to succeed if they are well-organized, if the general public sides with the protesters, and if they are peaceful. The story here is fairly complicated, and we at Social Change Lab have been looking at the evidence to try and draw conclusions around when protest can be effective, and the mechanisms by which protest can have an impact.

Can Protests Really Have An Impact?

There is a moderate amount of empirical evidence that suggests protest can impact public opinion and voting behavior — some notable research on this topic includes a paper by Madestam et al. and a paper by Teeselink and Melios, which both use quasi-experimental methods. Randomised Control Trials, the gold standard for causal inference, are difficult to do when it comes to analyzing the effectiveness of protest. For obvious reasons, social scientists aren’t able to randomly assign protests to certain locations but not others to see how effective they are in changing public opinion! So, what sort of evidence should we be looking for?

There are a few types of evidence that we can use to figure out whether protests are actually able to influence public opinion. Quasi-experimental designs are a useful form of evidence for understanding the impact of protest ⁠— they can give us information similar to what we would get from an experiment without actually having to conduct an experiment. For example, suppose that lots of protests are due to take place across a country, but some of them happen to be poorly attended because of heavy rainfall, whereas others are well attended because of good weather. This is a bit like having a treatment group and a control group as you would in a conventional experiment, and we can analyze whether public opinion is different in the rainy locations compared to the rain-free locations. The simple way of putting it is that this is analogous to an experiment that tests whether the size of a protest affects public opinion. We can also look at actual experiments that take place by showing people news coverage of protests: does one group that is assigned to a ‘treatment group’ wherein they watch news reports about a protest develop different opinions about the protest issue than a ‘control’ group that doesn’t watch a news report?

Let’s start by looking at two of the studies that use quasi-experiments. In one paper, Madestam et al. (2013) use the rainfall method mentioned earlier to check whether Tea Party protests in the U.S. had any impact on voting behavior and/or public opinion ⁠— and they found that they did, in both cases. In ‘non-rainy’ districts (where lots of people turned up to protest), survey respondents were 6% more likely to express support for the Tea Party movement than in districts that had significant rain on days when a Tea Party protest occurred. This equates to a 45% increase in the number of Tea Party supporters a year and a half later in non-rainy districts. The non-rainy districts also ended up with more people voting for Republicans (the party that the Tea Party supporters are associated with). Overall, counties that had a protest in good weather ended up with a 7% increase in the vote share for the Republicans in the subsequent election, which is a reasonably significant swing given the general tight margins of electoral politics.

Another study that uses the rainfall method is a paper by Teeselink and Melios that looks at the effect of the Black Lives Matter protests on the vote share of the Democrats. Here, they found that the protests shifted peoples’ perceptions of racial disparities, with a 1% increase in the number of people going out to a Black Lives Matter protest being linked to a 5.6% increase in the Democratic vote share in the subsequent election. As the average attendance of Black Lives Matter protests was 0.5% of the county’s population (in counties with at least one protest), the actual boost to the Democratic vote share was between 1.6 to 2.8 percentage points in counties where a protest took place.

The other popular method that lets us begin to figure out whether protests are an effective tool in changing voting behavior and public opinion are randomized control trials. In this case, these are experiments that involve news vignettes to see whether a treatment group that is assigned to read a news article about a protest has a different view on the issue the protesters are highlighting than a control group that doesn’t read the news article. One study by Budgen shows that people are fairly receptive to peaceful protests ⁠— when shown a news report about peaceful environmental protesters in the U.S., the treatment group was significantly more likely to say they supported the protesters than the control group, as can be seen in the figure below. There are two caveats here: firstly, the treatment was only effective in Democrats and Independents; Republicans were not responsive to the news reports. The second caveat is that violent protests did not significantly increase support for the protesters — although there was also no evidence of a ‘backfire effect,’ where violent protest led to respondents taking the opposite position to that of the protesters.

Protest can also be effective in increasing the salience of an issue by making people more aware of a problem or campaign. Carey et al. (2014) found that a 2006 protest campaign in the U.S. drawing attention to the hardships suffered by undocumented immigrants was effective in increasing the salience of immigration among Latinx respondents ⁠— they used survey data to demonstrate that people living in areas where protests took place were more likely to say they regarded immigration as one of the most important issues.

Another possible benefit of protest could be the direct impact on policymakers ⁠— if protesters are able to inspire action from legislators, they are more likely to change policies and bring about institutional change. Wouters and Walgrave (2017), in a study of 65% of all elected politicians in Belgium, show that protests have a significant effect on the beliefs of political representatives. They reported that protests affect the salience of the protest’s issue amongst politicians, the position they take and their intended actions (e.g. voting on a particular policy). They also found that the size of the protest, and unity in the protestors’ message, are the most significant factors in influencing political representatives.

What Sort Of Protest Movements Should People Support?

After carrying out an initial review of some of the literature on which protest movements are most likely to be successful, we believe there are a few characteristics of protests that animal welfare advocates should look out for when deciding which protests to support. Firstly, research from Wasow and Feinberg et al. suggests that non-violent protest is probably more effective in getting the public to lend their support to protesters. Secondly, research from Mirowsky and Ross suggests that successful protests are likely to be carried out by well-organized groups, and have lots of people in attendance ⁠—smaller protests are less likely to bring about policy changes or get the attention of the media. Thirdly, research from Agnone suggests that protest movements are more likely to be successful if they choose an issue that the public already sides with them on ⁠— for instance, animal advocates could protest against the use of cages in farming.

There is some evidence on what makes protests successful in attracting attention from the public and media, changing legislator behavior, and affecting public opinion. In the 1970s, Charles Tilly formulated his ‘WUNC’ Framework of protests (as explained by Wouters and Walgrave), arguing that protesters and social movements are more likely to be successful if they demonstrate ‘Worthiness, Unity, Numbers, and Commitment’ (WUNC). There is some empirical support for these claims ⁠— as noted above, Wouters and Walgrave conducted an experiment showing news vignettes to Belgian legislators. They found that when legislators were shown news clips where protests had large numbers of people and demonstrated ‘unity’ (meaning all protesters used similar signs and chanted the same slogans), they were more receptive to the cause of the protesters, becoming more likely to say that they agreed with the position of the protesters and that they would take action on the cause that the protesters were highlighting.

Wasow (2020) and Feinberg et al. (2020) have both looked at whether protests are more/less successful if they are more extreme. Wasow uses the rainfall method to determine that violent protests during the Civil Rights Movement depressed the Democratic vote-share in the county that protest took place in by 1.5% to 7.9%, whereas non-violent protest increased the Democratic vote-share by 1.6% to 2.5%. Feinberg et al. (2020) carried out an experiment wherein people were shown news reports depicting either extreme or moderate protest. For instance, participants were either shown a news report depicting a moderate protest involving people holding up anti-Trump signs and chanting outside of a Trump campaign event, or they were shown protesters blocking carloads of Trump supporters from reaching the event and causing a traffic jam. The figure below shows that participants who were shown the extreme anti-Trump protests were significantly more likely to say they supported Trump after viewing the clip. As mentioned above, Budgen (2020) also found that people are receptive to peaceful protests but not receptive to violent protests ⁠— although unlike the other papers, there was no evidence that the violent protests actively backfired, they were just significantly worse than peaceful protest in getting respondents to support the cause of the protest.

This evidence suggests that to maximize their effectiveness, animal welfare advocates ought to plan protests that are not perceived as extreme by the general public, with a total commitment to not use violence during protests or demonstrations. The examples of animal welfare protests at the beginning of this article organized by The Humane League and Animal Rebellion are probably more likely to change public opinion than more extreme forms of protest.

Successful protest movements are also likely to be well-organized, with clear governance and processes. Mirowsky and Ross (1981) show that if a protest movement has a formal bureaucracy, it is much more likely to be recognised by political actors as legitimate. There are clear risks to protest movements that don’t have much in the way of hierarchy and structure: infighting and disputes are common among small, grassroots organizations. Establishing clear procedures to handle disputes and deal with harmful behaviors is vital for protest movements to be recognized as legitimate and to succeed in achieving their goals.

Protest Movements are also probably more likely to succeed when their cause is popular with the public. A 2007 paper by Jon Agnone proposed the Amplification Model, suggesting that legislators become more likely to pass legislation on an issue when there public opinion favors action and there are protests taking place. There are quite a few issues relating to animal welfare that are likely to be good candidates here. A YouGov poll carried out in October 2020 found that a substantial majority of people in the EU agree that using cages in farming is cruel to the animals being farmed, and disagree with the statement ‘using cages in farming is acceptable.’

For animal advocates, protest can be a valuable way of drawing attention to an issue and shifting public opinion. We think there is good evidence that protests can both increase the salience of an issue and make the general public more sympathetic to the cause of protesters. There is also some evidence that protest can influence legislators and change voting behavior. It is possible that some types of extreme protests are counterproductive, so advocates interested in effective action should support protest movements that use nonviolent methods. Other factors that advocates should consider when deciding which protest movements they should support include whether the aims of the protest align with the views of the general public, and whether a protest movement is well-organized with clear and specific demands.