Social movements often use protests and other collective actions to draw public attention to their cause, yet the psychological reactions to such actions from their targeted audience is not well understood. This research investigates uninvolved bystanders’ immediate responses to collective action using a quasi-experimental field study designed around a racial justice protest that took place at a large public university in the United States. We surveyed two student samples exactly one week apart at the same time and location, first in the absence of protest and then again at the time of a racial justice protest (Total N = 240). We found that participants who believed that racism was not a problem on campus had more negative attitudes toward racial justice protests and protesters, as well as lower support for anti-racist efforts on campus on the day of the protest, compared to the day without a protest. These findings provide initial evidence that a protest encounter may trigger a backlash effect amongst those who have the most resistant attitudes toward social change.
Racial justice movements often have a strong presence on college campuses. Since the Black Lives Matter movement emerged in 2013, a series of protests have occurred in universities across the United States to push for fair treatment of racial minorities.
Protests typically aim to gain the attention of broader society and mobilize greater support for their cause. We can observe this from the range of tactics and contentious actions that protesters typically undertake to disrupt everyday life. Thus, being confronted by a protest might trigger backlash among some people. We therefore aimed to understand how people react when they directly encountered a protest in their social environment.
We surveyed members of a campus community exactly a week before, and during a racial justice protest that occurred at a U.S. university. The research distributed brief anonymous surveys to participants who happened to be at the location of the protest. We found that among people who had relatively weaker perceptions of campus racism, they reported more negative attitudes toward racial justice protests and protesters, as well as less support for anti-racist efforts on campus when they physically encountered a racial justice protest (compared to when they did not encounter a protest).
Although protests are a means toward social and political change, it might trigger immediate backlash among people who do not perceive a strong sense of injustice. By focusing on the attitudes of people who encounter a protest, this study contributes to our understanding of how protests can influence the attitudes of broader society.
Social movements employ many approaches for achieving their goal of social change. From holding mass street demonstrations requiring roadblocks, to occupying buildings with sit-ins, and organizing strikes that unite employees in work stoppages, collective action often challenges people to confront issues that they might otherwise not consider. This wide repertoire of tactics typically seeks to
Various models of social change have argued that the success of social movements relies on public support for the movement and its goals (e.g.,
In the present research, we take a somewhat different approach by examining bystanders’ attitudes in
Research from related disciplines, mainly political science, sociology and communication studies, provides indirect evidence that exposure to protests (via one’s proximity to protest locations and media consumption) can influence public attitudes (e.g.,
The question of whether and how protests may have a proximal impact on bystanders is especially important because related psychological research has suggested that the immediate experience of being interpersonally confronted with prejudice and discrimination tends to be negative (e.g.,
One challenge of studying the experience of directly encountering a protest is that the phenomenon is difficult to authentically capture in a traditional laboratory setting. However, even when it is not feasible to conduct randomized controlled experiments, it is sometimes possible to creatively use the timing of naturally-occurring real world events to examine or create “natural experiments” to assess the impact of such events (
Taking inspiration from naturally-occurring events to study attitudes in response to protests, in the present research, we implemented a between-subjects quasi-experimental field study designed around a racial justice protest that occurred at a university campus in the U.S. Knowing the time and location of the protest in advance, we examined the social attitudes of the same population of people twice – once when there was
Although little research has examined the impact of direct encounter with protests, past research in political science and sociology has indirectly linked protests to public opinion. This work has typically used geographical proximity to local protest events based on geocoded data as predictors of attitudes reported in public opinion polls (e.g.,
In early attempts to study how members of the public directly respond to protests,
To our knowledge, since this early work, there has been limited empirical research on people’s proximal reaction to physically encountering protests. The one recent exception is
For movement organizers and its supporters, the hope is that collective actions will shift public attitudes to move in line with the goals of a movement. Indeed, sociological and political science research has documented the power of sustained social movements in producing large-scale structural change (e.g.,
Collective action for social change is often a confrontation about injustice. Past research has found that individuals who claim discrimination are typically negatively evaluated (
Given that challenges to the status quo can generally elicit resistance, we believe that directly encountering a protest may promote negative perception of protests and negative evaluation of protesters, as well as reduce support for social change efforts. Yet, not all bystanders of a protest are expected to react in a uniformly negative manner toward protests. Prior research has shown that perceived injustice is an important antecedent to being in solidarity with disadvantaged groups and engaging in collective action among bystanders (
Perceiving injustice, such as discrimination, is one of the critical first steps towards challenging the status quo, and thus, toward achieving social change (e.g.,
The present study focused on people’s reactions to a racial justice protest on a university campus in the United States. A considerable number of Americans believe racism is no longer a problem (
Given that we employed a quasi-experimental between-subjects design to examine whether encountering a protest would shift people’s attitudes, it was important to test whether participants recruited on the day of the protest and on the day without the protest differed on relevant individual difference variables that are not explicitly related to racial attitudes, but are known to shape people’s social and political attitudes more generally. Thus, in addition to our primary variables, we included ancillary measures of people’s general ideologies and moral motives that are well-established in prior research as largely stable individual differences: system justification (
System justification refers to a motivation to defend and legitimize the status quo in a society, which can subsequently contribute to the stability of oppressive social systems. Prior research has shown that system-justifying motives promotes lower support for protests for social change that challenges systems of inequality (
People typically have intuitions that guide their moral judgments in everyday life. These intuitions are often experienced as absolute truths or rules that organize society. Two core motives that guide moral regulation of group-based behaviors are social order and social justice (
The present research employed a quasi-experimental between-subjects design to study the effect of an anti-racism protest on uninvolved bystanders at a university. Universities have long been a site of activism (
Three pairs of research assistants were stationed at the targeted locations during both sessions of data collection, each lasting about an hour. The research assistants introduced themselves as researchers from the psychology department, and asked participants to complete a brief anonymous survey on their attitudes toward activism in exchange for candy. During the day of the protest, research assistants began approaching participants as soon as the protesters made their way through the targeted locations. Given the short amount of time during which participants were exposed to protesters at these locations, we aimed to sample as many participants as possible. The protest itself unfolded as anticipated; the protesters marched around campus while carrying banners (e.g., “Stand against racism”), and chanting slogans (e.g., “Racism at [university name] has got to go”, “Black lives matter”) to draw attention to their cause. The stated goal of the protest was to raise awareness of racial issues on campus. Local media reports estimate that between 60 to 100 people, mostly university students and some staff members, participated in the protest.
There were 162 participants recruited on the no-protest day and 78 participants recruited on protest day at a large public university in the Northeastern U.S. (total
Participants were asked “How much of a problem is racism on the [name of university] campus?” on a scale from 1 (
Participants were asked, “How much do you oppose or support anti-racist efforts on campus? (e.g., protests, discussion panels, workshops, diversity trainings)” on a scale from 1 (
Participants were asked: “People have different opinions on protests about racial issues. To what extent do you agree or disagree with the following opinions? I think protest about racial issues are...” followed by 9 reaction itemsiii [i.e., “disrespectful, dangerous, counterproductive, silly, annoying, legitimate (R), important (R), justified (R), effective (R)”] measured on a scale from 1 (
Participants were asked: “People have different opinions about racial justice protesters. To what extent do you agree or disagree with the following opinions? I think people who are protesting about racial issues are...” followed by 4 items [i.e., “admirable, determined, troublemakers (R), hypersensitive (R)”] measured on a scale from 1 (
Given our quasi-experimental between-subjects design, it was important to test whether participants on the protest and no-protest day differed on relevant individual differences. Thus, in addition to our main measures, we included measures of people’s general ideologies and social motives that are well-established in prior research as largely stable individual differences: system justification (
To measure system justification, we used two items developed by
Participants were asked to report their age, race, and university affiliation (e.g., undergraduate student, graduate student).
The means, standard deviations, and intercorrelations among the main variables on the no-protest day and protest day are displayed in
Variables | 1 | 2 | 3 | 4 | |
---|---|---|---|---|---|
1. Perceived campus racism | 2.99 (1.05) | – | |||
2. Support for anti-racist efforts on campus | 4.20 (1.00) | .27*** | – | ||
3. Negative reactions toward protests | 1.96 (0.74) | -.33*** | -.56*** | – | |
4. Negative evaluation of protesters | 1.89 (0.75) | -.32*** | -.53*** | .75*** | – |
***
Variables | 1 | 2 | 3 | 4 | |
---|---|---|---|---|---|
1. Perceived campus racism | 3.15 (1.20) | – | |||
2. Support for anti-racist efforts on campus | 3.88 (1.12) | .60*** | – | ||
3. Negative reactions toward protests | 2.17 (0.91) | -.53*** | -.72*** | – | |
4. Negative evaluation of protesters | 1.99 (0.93) | -.56*** | -.71*** | .86*** | – |
***
First, we assessed whether participants recruited on protest day and the no-protest day differed on key demographics (i.e., age, race) and on individual difference measures that we expected to be largely stable (i.e., system justification, social order, social justice). We used a chi-square difference test for the categorical variable of race and independent
As expected, system justification was related to more negative reactions toward racial justice protests, system justification was related to more negative reactions toward racial justice protests (
Participants on protest day and the no-protest day did not significantly differ in the extent to which they perceived racism to be a problem on campus (protest,
There was a significant difference in support for anti-racist efforts on campus between participants on protest day and the no-protest day,
Support for anti-racist efforts on campus as a function of the interaction between perceived campus racism and condition (protest vs. no-protest day).
There was a significant difference in negative perception of racial justice protests between participants on the protest and the no-protest day,
Negative perception of racial justice protests as a function of the interaction between perceived campus racism and condition (protest vs. no-protest day).
There was no significant difference in negative evaluation of racial justice protesters between participants on protest day and the no-protest day,
Negative evaluation of racial justice protesters as a function of the interaction between perceived campus racism and condition (protest vs. no-protest day).
Collective action is one of the tools that social movements use to engage broader society for social change. Yet, little data exists on people’s immediate reactions to collective action. In the present research, we conducted a quasi-experimental field study around a real anti-racist protest that occurred on a university campus, which we suggest encapsulates the phenomena whereby collective action aims to incite a reaction from its audience. We found evidence consistent with a backlash effect amongst people who did
Our findings are consistent with prior research showing that people typically have negative reactions to interpersonal confrontations of injustice (e.g.,
The present research has several limitations, which we believe motivates questions for future research. For one, although the current research has the strength of external validity and applied relevance, we acknowledge that the inability in a quasi-experimental field design to achieve true random assignment to conditions precludes the strongest causal claims (
Another limitation we wish to acknowledge is the uneven sample sizes collected on the day of the protest and the day without a protest. Our sample size was smaller on the day of the protest compared to the no-protest day because we were careful to only recruit participants who were directly exposed to the protest, which depended on the number of people who happened to be at the protest site. We attempt to address the problem of unequal cell size statistically by using general linear modelling and adjusted marginal means. We also conducted non-parametric tests where appropriate (see
It is important to note that the present study focused on understanding responses to collective action that is relatively normative and peaceful, in a largely non-repressive and low-risk context. However, people’s responses to encountering collective action are likely shaped by the strategies used during collective action. Prior research found that people tend to be more supportive of normative, non-violent collective actions compared to non-normative, violent actions (
Further, the protest that we investigated was relatively small in size. The protest we focused on also occurred on a university campus and was relevant to the campus community members but not necessarily to the broader public. Thus, such a protest may be qualitatively different from large-scale national protests that occur in other public places and aim to influence public opinion. Prior work has shown that social causes with a critical mass of supporters tend to arouse greater support amongst the public (
We did not find an empowering effect of encountering the protest on our outcome measures. There may be several possible reasons for this. First, descriptively we observed that participants who perceived high levels of racism on campus reported very low levels of negative perceptions of racial justice protests and negative evaluations of racial justice protesters, as well as very high levels of support for social change efforts on campus. This pattern creates a floor or ceiling effect respectively, which may conceal the potential mobilizing impact the protest had on this group. Second, it is possible that encountering an anti-racist protest may invigorate people who perceive a high degree of racism on campus, or it may promote complacency because it signals that action is already being taken. These divergent effects may cancel each other out and lead to the observed null effect. Third, the outcomes we measured may also not effectively capture an empowerment response. For example, people who were potentially empowered by the protest may report higher levels of efficacy and anger that could motivate future collective action (
Although this study identified the potentially demobilizing effect of a protest amongst those who had weak perceptions of injustice (i.e., low on perceived campus racism), we definitely do not suggest or expect that this will always be the case. Related research has shown that protests can spread and mobilize others, giving rise to protests in other geographical regions through various channels including news media coverage of protests and the social networks of activists themselves – what scholars have called protest diffusion (e.g.,
For this study, the data is freely available (see the
The complete study materials, data, analysis code, and report of non-parametric tests are available at
The authors have no funding to report.
The authors have declared that no competing interests exist.
The authors would like to thank the Racial Justice Coalition and the Human Relations Lab from the University of Massachusetts Amherst for assisting in data collection for this project.
Our sample size was determined by the physical and time constraints around the protest event, thus we did not conduct an a priori power analysis. Post hoc power analysis calculated in G*Power (
In addition to the measures reported, we also included exploratory questions about perceived importance of racial issues to students of color and to White students on campus. These perceptions did not differ between the protest and no-protest day (
We also included “disruptive” as an item to measure negative perception of racial justice protests. However, the item was unexpectedly negatively correlated with the other items (e.g., the more people thought the protests were disruptive, the less they thought the protests were annoying) which suggests that our participants may have interpreted disruptiveness as a positive feature of the protest. Thus, we excluded it from our scale.
Given the strong correlation between negative perceptions of protesters and negative evaluations of protesters, we also combined the two measures to form a composite variable of negative reactions to protests. There was a significant interaction between condition and perceived campus racism in predicting negative reactions to protests,
It was also possible that system justification, which is conceptually the reverse of perceived campus racism, could moderate the differences in attitudes on the day of the protest and the no-protest day. Indeed, higher perceived campus racism, which we conceptualized as our moderator, was related to lower scores on system justification (