Peace Science Digest

Movement Violence Can Lead to a Decline in Public Support

Photo credit: Wikimedia Commons

This analysis summarizes and reflects on the following article: Muñoz, J., & Anduiza. E. (2019). “If a fight starts, watch the crowd”: The effect of violence on popular support for social movements. Journal of Peace Research, 56(4), 485-498.

Talking Points 

In the context of the 15 May Movement (15-M) in Spain,

  • In Barcelona, a four-day-long riot associated with 15-M led to an overall 12-point decline in public support for the movement.
  • The decline in public support for the movement was conditional on individuals’ predispositions towards the movement: while support declined significantly among weak supporters, opposers, and those indifferent to the movement, it did not decline significantly among core supporters.
  • How individuals select and process information is important to understanding why levels of support dropped across all categories except core supporters—these individuals were more receptive to the movement’s justification for the use of violence.


The choice between violence and nonviolence is available to any protest movement. Opting to engage in violence is more costly to the movement because it increases the chance of state repression and also reinforces the claims of those who oppose the movement. The academic research on this topic shows that nonviolent movements are more successful in achieving their long-term goals, whether influencing policy or bringing about regime change. Many researchers theorize that broad-based public support for protest movements is instrumental to their success and that the use of violence may weaken this support.

The authors of this research were able to test the second half of this theory empirically by using survey data collected from May to June, 2016, in Barcelona, Spain. An unexpected riot erupted over the eviction of a squat group linked to the 15 May Movement—15-M or indignados. While this movement garnered high levels of support (upwards of 65% approval in Spain) at its conception, evidence from this article shows that support plummeted 12 percentage points following the violent riots in May 2016.

By 2010, Spain had one of the highest unemployment rates in Europe and particularly high youth unemployment at 43.5%. The government introduced economic reforms, which the labor unions rejected due to concerns that they infringed on workers’ rights. This kickstarted a series of mostly nonviolent protests against the reforms. 15-M emerged out of this movement and started to occupy central squares of the country’s largest cities. Shortly thereafter, the movement de-centralized, and a variety of local initiatives developed with similar political and socio-economic goals.

In Barcelona around this time, a group of people occupied and started a “free place project” in a former bank. Named Banc Expropriat (Expropriated Bank), this project rejected the idea of state or private property and used the space to host food banks, free shops, libraries, and a meeting spot for the local 15-M activists. The city government protected this space for a time, paying the owner of the building approximately $70,000 in rent to avoid a political confrontation. This changed in January 2016 when the government stopped paying rent on the building. On May 23, 2016, Catalan police carried out an eviction order on Banc Expropriat. This resulted in a four-day-long riot in which dozens of police and protesters were wounded and property was destroyed. 

It was a coincidence that the authors of this article were running a survey on support for 15-M in Barcelona at the time. Their survey yielded a total sample of 1,500 respondents who were older than 18, with balanced distribution by age, gender, and location across 73 neighborhoods. The authors separated the responses into two categories: those that took place before the riots and those that took place after, opting not to use responses taken during the time period the riots were taking place. Importantly, the authors sorted responses according to individuals’ initial levels of support for 15-M, determined by their previous voting record in the 2015 general election: core supporters (those who voted for a pro-15-M party), weak supporters (those who voted for the center-left party), those who were indifferent (those who did not vote), and opposers (those who voted for center-right or right parties).

Across all categories, public support for 15-M dropped an average of 12 percentage points from before the riots to after the riots. When examining responses by levels of support based on respondent voting behavior, the authors found a significantly larger decline in support towards the movement among those in the “indifferent” (34-point drop), “weak supporters” (16-point drop), and “opposers” (13-point drop) categories. Those in the “core supporters” category did not express a significant decline in support for the movement as result of its violence.

The findings here demonstrate that the use of violence by protest movements can cause a decline in public support—but that this decline is conditional on individuals’ predispositions towards the movement. To explain this, the authors suggest that core supporters are more receptive to the movement’s justification for the use of violence. This suggestion is supported by research on how individuals select and process information. Namely, their predisposition toward the movement influences how they take in information about it, as people are more likely to expose themselves to (selective exposure), pay attention to (confirmation bias), and process (motivated reasoning theory) information that confirms their existing beliefs, ignoring or discrediting information that contradicts these.

Informing Practice

Maintaining nonviolent discipline is not only a moral choice but also a strategically wise one for reaching a movement’s goals. In particular, this research offers even more support for the claim—advanced by Erica Chenoweth and Maria Stephan in Why Civil Resistance Works—that maintaining nonviolent discipline is key to sustaining the broad-based support critical to movement success. Protest movements risk losing public support by engaging in violence. However, they may not lose the support of their most ardent supporters. This produces some challenges for encouraging nonviolence if members are emboldened to act violently by their core supporters.

This dynamic has played out in the Digest’s hometown of Portland, Oregon, where violent confrontations between far-right, white supremacist organizations and antifa periodically dominate news coverage. Core supporters of each movement believe in what they see as a justified use of violence and deploy recorded instances of the other side’s violence to mobilize support. This is not to draw a false equivalency between the sides of this conflict. The far-right organizations advocate for a deeply racist, homophobic, and sexist ideology that is antithetical to a peaceful society. They frame themselves as victims of antifa violence, despite evidence of them instigating and/or seeking violent confrontations. Yet, this framing is effective in gaining support from a broader conservative audience, only reinforced by a tweet from U.S. President Donald Trump threatening to list antifa as a terrorist organization. Paired with a “fake news” crisis, which capitalizes on individuals’ predispositions and the tendency to accept only information that confirms their existing beliefs, the result is an encroaching acceptance of violence and white supremacist ideology as a means of political expression. At the same time, applying insights from the present research to this context reveals that the choice of some antifa activists to respond violently to white supremacists may serve to diminish broad-based public support for anti-white supremacist/anti-fascist movements. While being against white supremacy is and should be an obviously easy position to take, antifa’s violence may make otherwise sympathetic folks less likely to come out in the streets to join protests against white supremacy or be otherwise associated with them. Further, it gives other, less sympathetic people cover when making absurd comparisons between “both sides.” In the end, this possible loss of public support due to the use of violence will only weaken anti-racist/anti-fascist forces, making it harder to effectively achieve their goals.

Continued Reading 

Fisher, M. (2013, November 5). Peaceful protest is much more effective than violence for toppling dictators. The Washington Post  

Peace Science Digest. (2017, September). How do “violent flanks” affect the outcomes of nonviolent campaignsRetrieved September 5, 2019, from  

Peace Science Digest. (2017, June). Nonviolent resistance and government repression. Retrieved September 5, 2019, from  

Peace Science Digest. (2019, March). Making civil resistance work against rightwing populism. Retrieved September 5, 2019, from  

Kolbert, E. (2017, February 27). Why facts don’t change our minds. The New Yorker  

Singal, J. (2019, July 10). Researchers conducted six studies tinvestigate how best to challenge science deniers. The British Psychological Society Research Digest. Retrieved September 5, 2019, from  


International Center on Nonviolent Conflict:  

Albert Einstein Institution: 

Nonviolent Action, United States Institute of Peace: 

Metta Center for Nonviolence 

Keywords: Spain, protest movements, nonviolent/civil resistance, nonviolence, riots, public support 

The following analysis appears in Volume 4, Issue 4 of the Peace Science Digest.