What Kinds of Protests Actually Work?

How activists can both create pressure for change and win over support

Posted Nov 21, 2020

By Eric Shuman and Eran Halperin, Ph.D.

The killing of George Floyd at the hands of the police this summer (along with the deaths of Ahmaud Arbery, Breonna Taylor, and many others) ignited one of the largest protest movements in American history. Between May 26 and August 22, there were more than 7,750 demonstrations linked to the Black Lives Matter (BLM) movement across more than 2,440 locations in all 50 states and Washington, D.C. (ACLED, 2020).

 Life Matters
Source: Pexels; Credit: Life Matters

While the overwhelming majority of these demonstrations were peaceful, there were also numerous clashes with the police, rioting, and looting connected in some way with the protests. This sparked debate both among activists and the general public about what kind of tactics are more effective. Are the outbursts of aggression a necessary evil in order to generate pressure for tangible change? Or should activists seek to be as peaceful and moderate as possible in order to win more support? Or is there perhaps some way to combine these two perspectives?

While these debates were often conducted based on intuition, with each side drawing on their own experiences and readings of history, in recent years, this debate is also reflected in the scientific literature. Interestingly, there seems to be support for both sides. Research in sociology and political science has found evidence that riots and other extreme, more disruptive, tactics can sometimes spur change. For example, political scientists have linked the 1992 Los Angeles riots to positive changes in urban poverty public policy (Enos, et al., 2019). And in general, riots and other relatively extreme forms of protest are thought to generate pressure to address the problems causing the violence (Biggs & Andrews, 2015).

However, research also suggests that extreme tactics generate backlash and that moderate and peaceful protests are more effective at winning public support. Researchers studying different protest movements (animal rights, racial justice, partisan) recently found that compared to protests that used extreme tactics, including violence or the threat of violence, moderate peaceful protests generated more popular support and willingness to join the movements (Feinberg et al., 2020). The main reason for this was that people simply couldn’t identify with more radical protesters or see them as people like themselves.

This leaves activists with something of a dilemma: disruptive, potentially destructive actions that could apply pressure to bring about change are also likely to harm support and potentially turn public opinion against them. Might there be a third way leading out of this dilemma? This is what we investigated in a recent paper published in the Journal of Personality and Social Psychology (Shuman, Saguy, van Zomeren, & Halperin, 2020). Our starting point was that much of the discussion about protest tactics has been caught in a binary of comparing moderate, nonviolent to radical, violent action. We suggest there is a third type of action, which we term nonviolent nonnormative action, that might be particularly effective in generating public support.

Most nonviolent protests are normative, that is, they are in line with what people think is acceptable for a protest, e.g., peacefully assembling and marching. Similarly, most violent protests are considered nonnormative, as violence is seen as an automatic violation of what is acceptable. However, protests can also be both nonviolent and nonnormative – that is, they can step outside the bounds of what is considered normal or acceptable behavior, but do so in a nonviolent way. Classic examples of this include refusing to pay fines or taxes, striking, boycotting, engaging in sit-ins, or otherwise occupying public space. We thought this type of protest tactic might be particularly effective because its nonnormative nature generates pressure for change, while its nonviolent nature helps to maintain public support.

In our research, we experimentally compared these three different protest tactics in three studies—one in the U.S. in the context of BLM, and two in Israel in the context of Arab-Jewish relations. In the U.S. study, participants read one of four different news articles (the news articles were specially created for the research and were not real articles). One article reported on a peaceful BLM march (nonviolent normative protest), the second on a BLM march that degraded into a riot (violent nonnormative protest), and the third on a BLM strike on paying any municipal fines and fees (nonviolent nonnormative protest), and finally some read about no protest at all. In this and the other studies, we found that people's support for the protest’s demands for police reform was highest after being exposed to the nonnormative nonviolent protest. Even more interesting, the effect of nonnormative nonviolent methods of struggle was primarily on people who had been initially resistant to the protest's agenda. So actions like strikes and sit-ins were more effective than peaceful demonstrations and petitions.

Our research shows that nonnormative nonviolent protests are so effective because they balance two important factors. First, nonnormative nonviolent protests are seen as disruptive and thus demanding at least some sort of action to address it. Further, they are also seen as constructive – participants saw the protesters as having goals of achieving something beneficial for their cause and not trying to harm their opponents. Because participants felt the protests demanded immediate action and saw the protesters as having positive intentions, they responded with support for the protesters’ policy goals.

In addition to these experimental studies, we also analyzed two-real world protests as they were evolving (protests involving the disabled community in Israel and the March for Our Lives protests in the U.S.) and found that the more participants perceived the protests as having both of these elements—disruption and constructive intentions—the more supportive they became of the protest’s goals.

Thus, our research indicates that nonviolent nonnormative tactics can be an effective means for advancing social change because they are disruptive in a way that generates pressure for political change but also lead people to view the protesters as constructive. Notably, we found this effect mainly among people who were more resistant to the protest, suggesting nonnormative nonviolent action can be a powerful strategy when movements are facing resistance.

As so many modern social movements face opposition and resistance from those advantaged by the current status quo, we urge advocates to consider nonviolent nonnormative tactics as another powerful tool in their toolbox. While it can be easy to slip into the mindset that the only two options available are either violence or protest that might be too peaceful to challenge the status quo, our research shows that nonnormative nonviolent protest can be a third way that creates pressure for immediate action while also winning support for change in line with the protesters' goals.

 Eric Shuman
Source: Eric Shuman

Eric Shuman is a doctoral researcher at the University of Groningen, co-supervised by Martijn van Zomeren, Tamar Saguy, and Eran Halperin. He is currently based at the PICR Lab at the Hebrew University in Jerusalem. His research focuses on the psychology of collective action, protests, and social movements. He is particularly interested in understanding how different types of action can advance intergroup equality, and what psychological processes are necessary in order for historically advantaged group members to become more supportive of equality. He also works as a consultant at aChord: Social Psychology for Social Change. aChord Center is a non-profit organization that develops innovative, evidence-based, practical knowledge and tools to improve intergroup relations by utilizing cutting-edge social psychological theory and data.


Armed Conflict Location & Event Data Project. (2020). Demonstrations & Political Violence in America: New Data for Summer 2020. https://acleddata.com/2020/09/03/demonstrations-political-violence-in-america-new-data-for-summer-2020/

Biggs, M., & Andrews, K. T. (2015). Protest campaigns and movement success: Desegregating the US South in the early 1960s. American Sociological Review, 80, 416-443.

Enos, R. D., Kaufman, A. R., & Sands, M. L. (2019). Can violent protest change local policy support? Evidence from the aftermath of the 1992 Los Angeles riot. American Political Science Review, 113, 1012-1028.

Feinberg, M., Willer, R., & Kovacheff, C. (2020). The activist’s dilemma: Extreme protest actions reduce popular support for social movements. Journal of Personality and Social Psychology.

Shuman, E., Saguy, T., van Zomeren, M., & Halperin, E. (2020). Disrupting the system constructively: Testing the effectiveness of nonnormative nonviolent collective action. Journal of Personality and Social Psychology.