Sage Journals: Discover world-class research

Abstract

People living in communities where protests occur are exposed to various forms of violence, including state-perpetrated violence, such as police brutality, as well as violence from protesters, such as riots and looting. Does exposure to such protests influence people’s support for violence, whether from the state or protesters? To address this question, we analyze data from a multi-year panel study conducted in Chile (2016–2022) to examine how the massive demonstrations of 2019 shaped attitudes toward violence. Using a difference-in-differences design, we find that exposure to protests decreases support for state repression, but only for a short time. However, such exposure does not significantly influence support for the violence perpetrated by protesters. This nuanced understanding of attitudes toward violence provides new evidence to explain the support for different forms of violent actions during social movements.

Keywords

Political attitudes protests violence Latin America

Introduction

Since 2017, there have been more than 800 significant anti-government protests worldwide.¹ A common characteristic of these protests is the high levels of violence observed, particularly in large-scale demonstrations, where violence is often perpetrated by both the state and the protesters. In response to massive protests, governments frequently employ strategies that result in violent repression of social movements. State violence, especially police brutality, is widespread throughout the world (Ortiz et al., 2022; Whitehead et al., 2010; Zhukov, 2023) with effects on decreasing sympathy for state actors (Balcells et al., 2021), increasing political participation (Ang and Tebes, 2024; Hager and Krakowski, 2022), and also creating a backlash against violent repression and foster dissent (Curtice, 2021; Curtice and Behlendorf, 2021; Sutton et al., 2014).² At the same time, protesters have adopted more radical tactics, leading to a rapid escalation from non-violent protests to violent demonstrations (Ortiz et al., 2022; Pierskalla, 2010; Snow and Moss, 2014).

The question of whether people exposed to violence during protests in their communities support violent reactions from key actors—either the protesters or the state—remains under-explored, particularly from a comparative perspective.

Using the case of the 2019 Chilean social outburst, we investigate whether exposure to large protests impacts support for violence by both the state and protesters. This event offers a relevant scenario to explore such a question. During this protest cycle, the protesters faced police violence and the participants resorted to violent tactics against law enforcement and property (Palacios- Valladares, 2020; Somma et al., 2021).³ Although previous literature has explored how protests have an effect of supporting governmental strategies (Ketchley and El-Rayyes, 2021; Wasow, 2020) and protesters’ actions (Maguire et al., 2020; Thaler et al., 2023), they are usually evaluated in separate contexts and events.⁴

Using a multi-year panel survey conducted in Chile and geocoded occurrences of protests at the municipal level, we find that people who experienced protests and riots in their communities exhibited a decrease in support for state violence to repress mobilization, but only for a short time. In contrast, our results indicate that the same exposure does not affect support for violence from protesters. This suggests dual attitudes toward violence depending on the perpetrator. Considering this duality of support for violence, our study makes two main contributions. First, we examine the effects of living in a municipality where violence triggered by social movements occurs, rather than focusing on how violence is framed or presented to a broader audience. Second, this article does not center on specific groups, political leanings, or individual characteristics of protesters. Instead, the goal is to provide evidence on the dynamics of support for violence from a macro lens, exploring varying levels of support depending on the “perpetrator” (state vs protesters), an aspect not typically addressed in the literature.

The 2019 social outburst in Chile

Since the return to democracy in 1990, Chile has experienced a prolonged period of continuous economic growth and the consolidation of democratic institutions. However, despite notable improvements in various socioeconomic indicators, Chile continues to exhibit significant levels of inequality, positioning it among the most unequal countries in Latin America.

In 2019, a fare hike of 30 Chilean pesos (about USD 0.04) on Santiago’s subway tickets quickly triggered an unusual wave of protests across the country. Although these demonstrations initially began as high school students’ protests in Santiago subway stations, they quickly escalated to massive protests in various cities throughout the country. This situation led to a social outburst (Estallido Social) during the last quarter of 2019, characterized by widespread outrage over rising living costs, low wages, and inadequate healthcare, education, and pension systems, among other diverse demands. These protests became the largest demonstrations since the end of the dictatorship in 1990 (COES, 2020), prompting the government to impose extraordinary measures, such as curfews and a state of emergency in several parts of the country.

Analyses suggest that both economic and political factors played crucial roles in igniting these protests (Cox et al., 2023; Somma et al., 2021). Economic grievances, particularly regarding access to healthcare, education, and pensions, were significant drivers of citizen discontent (Somma et al., 2021). The growing divide between ordinary citizens and elites, marked by diminished party affiliation, declining trust in institutions, and decreased voter participation since 1990 (Bargsted and Maldonado, 2018; Morales Quiroga, 2020), further fueled public mobilization. Additionally, scholars argue that recent debates around social hierarchies based on gender and ethnicity exacerbated dissatisfaction with political elites, contributing to increased participation in the protests (Reyes-Housholder and Roque, 2019; Somma et al., 2021). Once people took to the streets, social media also played a role in fostering protest engagement (Scherman and Rivera, 2021).

One of the defining features of the so-called “Chilean Spring” was the unprecedented levels of violence from both protesters and the state (Somma et al., 2021). Peaceful demonstrations, which included self-organized town meetings known as cabildos, improvised artistic events, the “march of one million” people, and cacerolazos, were carried out alongside violent actions, including criminal activities. Violent incidents included looting and arson attacks on public transportation, and other public and private infrastructure, most of which had not been seen in post-authoritarian Chile.

In response to public disorder, the government declared a state of emergency, enacting measures such as the night-time deployment of military personnel and curfews to keep people off the streets at night for the first time since the dictatorship (Thaler et al., 2023). To suppress the massive protests, law enforcement officers responded with extreme brutality (Somma et al., 2021). During the peak of the protest, authorities and human rights organizations documented the scale of the repression, reporting at least 34 deaths, more than 3000 injuries, and 20,000 arrests (Dammert and Sazo, 2020). The use of kinetic impact projectiles resulted in severe ocular trauma, visual impairment, and permanent disability (Rodríguez et al., 2021). Findings indicate that police forces deliberately harmed protesters and were responsible for 75% of recorded cases of state violence (Amnesty International, 2020). As a result, for several weeks, the country experienced an unprecedented spiral of violence associated with political and social unrest (Castro, 2022; Gonzalez and Le Foulon, 2020). The government’s violent response to the 2019 protests led many Chileans to draw parallels between the authorities’ actions and the Pinochet dictatorship. The parallel between past and present episodes of violence emerged in both political discourse and cultural expression. Protesters chanted “Piñera, asesino igual que Pinochet” (“Piñera, a murderer just like Pinochet”), while signs equating 2019 with the 1973 coup appeared in the streets. Resistance songs from the dictatorship era also resurfaced, with Víctor Jara’s El Derecho de Vivir en Paz (“The Right to Live in Peace”) becoming the movement’s anthem (Thaler et al., 2023). This resurgence of historical parallels underscores the deep tensions in Chilean society, where past violence remains a powerful lens for interpreting recent events of violence.

Research design

Data

To measure the outcomes—support for state violence and violent protests—we use an in-person, multi-year panel study conducted by the Center for Social Conflict and Cohesion Studies in Chile. This panel survey is based on a probabilistic, stratified cluster, and multistage sampling design.⁵ The first wave was representative of approximately 77% of the country’s total population and 93% of the urban population.⁶

We rely on this panel survey to capture people’s attitudes toward the use of violence in the context of protests. Given that the social outburst began as a student mobilization and that previous social movements in Chile have been led by students, our main outcomes of interest focus on student protests. To measure support for state violence, we use the level of support for the following statement: the police forcibly evict students from a high school in occupation. To measure support for violent protests, we use the level of support for the following statement: students throw stones at the police during a demonstration for education. Both statements are measured on a 5-point Likert scale. These statements capture the two dimensions we are interested in, top-down and bottom-up violence, during a protest context. Since we have panel data, we can study how these attitudes evolve over time, measured using the same questions before and after the social outburst. See Appendix B in the supplementary information for further information on the question wording and operationalization of the survey items included in the analysis.

In the case of state violence, support or lack of support might be explained by the nature of the protest (i.e., occupying a school). Similarly, in the case of protesters’ violence, support or lack of support might be influenced by the target of the violence (i.e., the police). Therefore, as a robustness check, we use alternative measures for both outcomes: one that captures state repression against a peaceful mobilization, and another that captures violence against a non-state actor, such as blocking a public street. Our results are robust to both alternative outcomes for capturing support for violence (see Appendix C in the supplementary information).⁷

To measure the treatment—exposure to protests during the 2019 social outburst—we use the Armed Conflict Location and Event Database (ACLED) (Raleigh et al., 2010). This dataset contains protests and riots at the municipal level on a daily basis. Using this data, we compute the total number of events that occurred at the local (i.e., municipal) level in Chile during the 2019 protests. We employ three different time frames to measure the treatment: (i) The first week after October 18, 2019, as the most intense period of the social outburst occurred during the initial days, (ii) the first month after October 18, 2019: since the 2019 wave of the panel study was conducted in late November 2019, we use this as the cutoff to capture events, ensuring that the treatment is measured before the survey fieldwork, and (iii) the first five months after October 18, 2019, marking the onset of COVID-19 quarantines in March 2020. No matter which time frame we use to measure exposure, our results remain consistent. We use the first (i.e., 1 week) in the manuscript and the other two as robustness checks in Appendix D in the supplementary information.

Although one might argue that the entire country was exposed to the social outburst due to the national salience of the event, there were important geographic differences in the magnitude and intensity of the protests and riots (Somma et al., 2021). As a result, we approach this as a dose-response design (Rosenbaum, 2003), using differential exposure to understand causal effects. In particular, we focus on protest shocks, defined as large-scale exposure to violence during the 2019 social outburst.⁸ We define exposed municipalities as those where the number of protests and riots is one standard deviation above the mean, with the rest serving as control municipalities (see Severino and Visconti (2025) for a similar design). Survey participants living in exposed municipalities are considered exposed subjects, and the same logic applies to the control group. As a robustness check, in Appendix E in the supplementary information, we employ a continuous measure of exposure (i.e., the total number of protests and riots) rather than a binary indicator for protest shock exposure. In Appendix F, we estimate the effects across varying levels of exposure by transforming the continuous treatment variable into quintiles and including it as a categorical factor in the regression analysis. This approach allows us to examine how the effects vary across different levels of protest exposure, using the lowest quintile as the reference group. Our main conclusions remain unchanged when using either alternative specification.

Because the unexposed and exposed people might live close to each other, potential spillovers are a relevant concern. This means that citizens who did not experience a protest shock might interact with those who actually did. However, if that is the case, such interactions would likely bias the effects toward zero. This assumption is common in natural experiment settings, where interference is possible. Therefore, any effect should be considered a conservative estimate (Keele et al., 2015).

Also, we acknowledge that municipalities exposed to a protest shock may differ from those that were not exposed, and these differences could explain varying attitudes toward violence in those localities. However, a difference-in-differences design relies on the parallel trends assumption, which means that exposed and non-exposed groups should have followed the same trajectory in the absence of the treatment. We provide evidence in the Results section of a common trend before the 2019 social outburst in Chile.

Empirical strategy

We use a difference-in-differences (DiD) design to estimate the effect of exposure to protests. Our analysis relies on a multi-year panel survey, using data from 3 years before (2016, 2017, and 2018) and 4 years after (2019, 2020, 2021, and 2022) the social outburst.

A static two-way fixed effects model does not allow us to observe changes in treatment effects over time or provide evidence regarding parallel trends. Consequently, we estimate a dynamic two-way fixed effects model, where we evaluate dynamic effects by regressing the outcome on time-relative-to-treatment dummies, along with municipality and time-fixed effects.⁹

Formally, we employ the following model:

Y_{i t} = α_{i} + φ_{t} + \sum_{τ \neq 2018}^{} β_{τ} D_{i t}^{τ} + {ω X}_{i t} + ε_{i t}

(1)

where Y is the five-point scale measuring support for state repression and support for violent it protests. α_i are municipality fixed effects, and φ_t are year fixed effects. D^τ is a dummy variable indicating whether exposure occurred τ years earlier or will occur τ years ahead. β_τ represents the effect of exposure for each τ year relative to the year before exposure to the social outburst (2018). The coefficient of interest is β_τ. X corresponds to a set of placebo covariates (i.e., individual characteristics not affected by exposure: respondents’ gender, age, and education).¹⁰ Finally, standard errors are clustered at the respondent level. As a robustness check, we also cluster the errors at the municipality level, and the main conclusions hold (see Appendix J in the supplementary information).

Results

Figure 1 illustrates the trajectory of the outcomes (i.e., attitudes toward violence from the state and protesters’ violence) over time (from 2016 to 2022). Regarding state violence, at first glance, we observe a similar trajectory for both attitudes before the social outburst. After the event, we see a divergence in 2020 and the following years. It is also noteworthy that after 2 years of declining support for state violence, the levels of support returned to those observed before the social outburst for both the control and exposed groups. Although interpreting this trend goes beyond the scope of this research, it is possible that the country shifted back to supporting more state violence as the immediate impact of the protests waned and concerns over security and the economy became more prominent among Chileans, particularly during the height of the COVID-19 pandemic (Castiglioni, 2020). In contrast, regarding support for protesters’ violence, there is no substantial difference between the control and exposed groups before and after the social outburst.

Figure 1.

Average support for violence from the state and protesters over time.

Figure 2 presents the DiD estimates using a dynamic model (equation (1)). As observed, there is evidence supporting the existence of parallel trends. The exposed and control groups followed similar trajectories from 2016 to 2018 for both outcomes of interest. However, there is a divergence in this trajectory for support for state repression. The exposed group experienced a reduction in their tolerance for state violence in the first years after the social outburst by 0.282 points (outcome on a 1–5 scale, 95% CI: [−0.438, −0.126]), though this effect seems to dissipate by 2022. Conversely, there is no evidence of a divergence in the trajectory between the exposed and control groups regarding support for violent protests. We report the results in table format in Appendix K in the supplementary information.

Figure 2.

Effects of protest shocks on support for violence from the state and protesters over time.

Finally, we explore heterogeneous treatment effects by ideology. Prior research has highlighted the importance of the left-right ideological spectrum in explaining voter behavior (Visconti, 2021; Zechmeister, 2015) and informing opinions in Chile (Cabezas et al., 2025). We do not find clear evidence that ideology moderates how people evaluate support for violence. See Appendix L in the supplementary information for further details.

Discussion and conclusions

Our findings indicate that experiencing large violent mobilizations changes people’s views on violence when it is perpetrated by the state (but for a short time). However, it does not seem to affect their opinions when the violence comes from protesters.

How can we explain these findings? We suggest three main potential explanations. First, police brutality constitutes a violation of human rights, involving the abuse of power by the political system, the state, and its agents (Reny and Newman, 2021). Societal perceptions regarding the functioning of the political system and the state can significantly influence people’s reactions to state actions. Such influence may not necessarily be present in the case of looting or other types of violence perpetrated by protesters. The 2019 mobilizations prompted an unprecedented and brutal response from state actors, particularly the police. The government’s rhetoric—former President Piñera stating, “We are at war against a powerful enemy”—only a few hours after the protests began did not help to curb either the state violence or the escalation of violence (Ledur and Levine, 2019). Documented instances of police brutality by various state agencies and academic research were highly unusual within the context of post-dictatorship Chile.

Second, at the beginning of the protests, there was strong popular support, and as a result, any repression against protesters would not be supported by the public. The 2019 protests in Chile were the largest in post-dictatorship history, with public opinion surveys indicating substantial support for these mobilizations, despite the significant disruptions to people’s lives and public order. The state employed extraordinary measures to maintain public order and discourage participation in the protests. Since many people perceived the social outburst positively—at least during the initial months—they may have viewed state actions aimed at suppressing the mobilizations negatively.

Finally, a third potential interpretation relates to how individuals attribute responsibility and accountability for violent acts. It is possible to argue that the asymmetry observed in the results may be shaped by perceptions of responsibility in the development of violent actions. Protester violence, especially during the early phase of the Estallido Social, may have been seen as the work of “a few bad apples” within a large, loosely organized, without clear leaders, and heterogeneous movement. This perception might have pushed individuals to disconnect the actions of violent protesters from the goals of the movement as a whole, especially during the first weeks of the movement, when the mobilizations were gaining public and elite support, and momentum. For instance, groups like Primera Línea, a relatively small group of protesters known for direct confrontations with police, were perceived by some as violent agitators, even by people supporting the movement, but a necessary defensive force against police brutality by others (Claude, 2020). In contrast, violence committed by the police was more easily attributed to the state and its authorities, given their institutional role and hierarchical structure. As a result, citizens were more likely to attribute direct responsibility to the government for the violent actions, viewing them not as isolated incidents but as part of a broader strategy of repression.¹¹

Regarding the short duration of new attitudes toward state violence, a possible explanation is that exposure to protest shocks did not affect people’s deeply held political beliefs but instead had a more transient impact on their sentiments and attitudes toward politics.

Some may argue that everyone in the country was exposed to the violence through extensive media coverage and social media dissemination of police brutality victims. However, individuals living in areas experiencing high levels of violence encountered it directly, not just through the media. First-hand experiences, including those through family, friends, and acquaintances, likely influenced people’s attitudes and opinions more profoundly than media exposure alone. Even though extensive media coverage made everyone aware of violence, these differences in direct exposure may have triggered stronger negative perceptions of the state’s actions, particularly in the months following the onset of the protests.

A major concern for the validity of our identification strategy is the onset of the COVID-19 pandemic, which occurred in the months following the start of the mobilizations. To further assess potential confounding, in Appendix M in the supplementary information, we test whether municipalities exposed to protest shocks were also more affected by the pandemic. The results show no significant relationship between protest exposure and COVID-19 case rates, suggesting that municipalities more affected by protests were not systematically more (or less) affected by the pandemic. Additionally, there is no evidence that pandemic-related measures, such as lockdowns and curfews, were implemented strategically to demobilize protesters. While these policies naturally led to a decline in the frequency and scale of public demonstrations, they appear to have been driven primarily by public health concerns, guided by epidemiological indicators such as infection rates and ICU capacity (Pavani et al., 2022). Therefore, this weakens the interpretation that these measures were politically motivated acts aimed at demobilizing or repressing the population.

Overall, this research provides new evidence on citizens’ perceptions of violence in the context of large demonstrations. It also contributes to a better understanding of the short- and long-term dynamics of public opinion during mobilizations and how massive events reshape attitudes. These findings offer a solid starting point for examining how the level of violence or tactics used by law enforcement, as well as violent actions from protesters, affect public support for violence from a comparative perspective.

Supplemental Material

Supplemental Material - Exposure to protests and support for different forms of violence: Evidence from the 2019 social outburst in Chile

Supplemental Material for Exposure to protests and support for different forms of violence: Evidence from the 2019 social outburst in Chile by Sebastian Rivera, Fernando Severino, and Giancarlo Visconti in Research & Politics

Supplemental Material

Supplemental Material - Exposure to protests and support for different forms of violence: Evidence from the 2019 social outburst in Chile

Footnotes

Acknowledgements

We thank Francisca Castro, Rodrigo Medel, Piero Stanig, and MPSA 2024 seminar participants for their comments and suggestions.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by ANID-Millennium Science Initiative Program [Grant Number NCS2024_007].

ORCID iDs

Sebastian Rivera

Fernando Severino

Giancarlo Visconti

Supplemental Material

Supplemental material for this article is available online.

The replication files are available at:

Carnegie Corporation of New York Grant

This publication was made possible (in part) by a grant from the Carnegie Corporation of New York. The statements made and views expressed are solely the responsibility of the author.

Notes

References

Amnesty International (2020) Eyes on Chile: police violence and command responsibility during the period of social unrest. https://www.amnesty.org/en/documents/amr22/3133/2020/en/

Ang

Tebes

(2024) Civic responses to police violence. American Political Science Review 118(2): 972–987.

Balcells

Dorsey

Tellez

(2021) Repression and dissent in contemporary cat- alonia. British Journal of Political Science 51(4): 1742–1750.

Banaszak

Ondercin

(2016) Public opinion as a movement outcome: the case of the US women’s movement. Mobilization: An International Quarterly 21(3): 361–378.

Barceló

(2018) Batons and ballots: the effectiveness of state violence in fighting against Catalan separatism. Research & Politics 5(2): 2053168018781747.

Bargsted

Maldonado

(2018) Party identification in an encapsulated party system: the case of postauthoritarian Chile. Journal of Politics in Latin America 10(1): 29–68.

Branton

Martinez-Ebers

Carey

Jr , et al. (2015) Social protest and policy attitudes: the case of the 2006 immigrant rallies. American Journal of Political Science 59(2): 390–402.

Cabezas

Navia

Rivera

(2025) The moderating effect of ideological identification on how economic news and economic growth impact individual economic perceptions: evidence from Chile. Political Studies Review 23(1): 171–189.

Castiglioni

(2020) La política chilena en tiempos de pandemia. Entre la (des) movilización social y la crisis sanitaria. Nueva Sociedad 287: 68–79.

10.

Castro

(2022). The backlash of state coercion: varieties of repression and their effect on mobilization. OSF Preprints. January, 25.

11.

Claude

(2020) Retrato de un clan de la Primera Línea. CIPER Chile. https://www.ciperchile.cl/2020/01/06/retrato-de-un-clan-de-la-primera-linea

12.

COES (2020) Observatory of conflicts - cumulative dataset. Harvard Dataverse (V7). DOI: 10.7910/DVN/GKQXBR

13.

Cox

González

Le Foulon

(2023) The 2019 Chilean social upheaval: a descriptive approach. Journal of Politics in Latin America 16(1): 68–89.

14.

Curtice

(2021) How repression affects public perceptions of police: evidence from a natural experiment in Uganda. Journal of Conflict Resolution 65(10): 1680–1708.

15.

Curtice

Behlendorf

(2021) Street-level repression: protest, policing, and dissent in Uganda. Journal of Conflict Resolution 65(1): 166–194.

16.

Dammert

Sazo

(2020) Scapegoats of the 2019 Chilean riots: from foreign intervention to riff-Raff involvement. SAIS Review of International Affairs 40(2): 121–135.

17.

Disi

(2021) The nearness of youth: spatial and temporal effects of protests on Political Attitudes in Chile. Latin American Politics and Society 63(1): 72–94.

18.

Edwards

Arnon

(2021) Violence on many sides: framing effects on protest and support for repression. British Journal of Political Science 51(2): 488–506.

19.

Enos

Kaufman

Sands

(2019) Can violent protest change local policy support? Evidence from the aftermath of the 1992 Los Angeles riot. American Political Science Review 113(4): 1012–1028.

20.

Gonzalez

Le Foulon

(2020) The 2019–2020 Chilean protests: a first look at their causes and participants. International Journal of Sociology 50(3): 227–235.

21.

Hager

Krakowski

(2022) Does state repression spark protests? Evidence from secret police surveillance in communist Poland. American Political Science Review 116(2): 564–579.

22.

Keele

Titiunik

Zubizarreta

(2015) Enhancing a geographic regression discontinuity design through matching to estimate the effect of ballot initiatives on voter turnout. Journal of the Royal Statistical Society - Series A: Statistics in Society 178(1): 223–239.

23.

Ketchley

El-Rayyes

(2021) Unpopular protest: mass mobilization and attitudes to democracy in post-Mubarak Egypt. The Journal of Politics 83(1): 291–305.

24.

LeBas

Young

(2022) Repression and dissent in moments of uncertainty: panel data evidence from Zimbabwe. American Political Science Review 118: 1–18.

25.

Ledur

Levine

(2019) Chile Woke Up. Reuters. https://www.reuters.com/graphics/CHILE-PROTESTS/0100B32527X/

26.

Maguire

Barak

Wells

, et al. (2020) Attitudes towards the use of violence against police among occupy wall street protesters. Policing: Journal of Policy Practice 14(4): 883–899.

27.

Medel

Somma

(2016) ¿Marchas, ocupaciones o barricadas? Explorando los determinantes de las tácticas de la protesta en Chile. Politica y Gobierno 23(1): 163–199.

28.

Morales Quiroga

(2020) Estallido social en Chile 2019: participación, representación, confianza institucional y escándalos públicos. Analisis Politico 33(98): 3–25.

29.

Ortiz

Burke

Berrada

, et al. (2022) World Protests: A Study of Key Protest Issues in the 21st Century. Berlin: Springer Nature.

30.

Palacios-Valladares

(2020) Chile’s 2019 october protests and the student movement: eventful mobilization? Revista de Ciencia Política 40(2): 215–234.

31.

Pavani

Cerda

Gutiérrez

, et al. (2022) Factors associated to the duration of COVID-19 lockdowns in Chile. Scientific Reports 12(1): 9516.

32.

Pierskalla

(2010) Protest, deterrence, and escalation: the strategic calculus of government repression. Journal of Conflict Resolution 54(1): 117–145.

33.

Raleigh

Linke

Hegre

, et al. (2010) Introducing ACLED: an armed conflict location and event dataset. Journal of Peace Research 47(5): 651–660.

34.

Reny

Newman

(2021) The opinion-mobilizing effect of social protest against police violence: evidence from the 2020 George Floyd protests. American Political Science Review 115(4): 1499–1507.

35.

Reyes-Housholder

Roque

(2019) Chile 2018: desafíos al poder de género desde la calle hasta La Moneda. Revista de ciencia política (Santiago) 39(2): 191–216.

36.

Rodríguez

Peña

Cavieres

, et al. (2021) Ocular trauma by kinetic impact projectiles during civil unrest in Chile. Eye 35(6): 1666–1672.

37.

Rosenbaum

(2003) Does a dose–response relationship reduce sensitivity to hidden bias? Biostatistics 4(1): 1–10.

38.

Scherman

Rivera

(2021) Social media use and pathways to protest participation: evidence from the 2019 Chilean social outburst. Social Media + Society 7(4): 1–13.

39.

Severino

Visconti

(2025) Immigration shocks and unfounded concerns about crime: evidence from Haitian migration to Chile. The Journal of Politics 87(3): 1210–1214.

40.

Snow

Moss

(2014) Protest on the fly: toward a theory of spontaneity in the dynamics of protest and social movements. American Sociological Review 79(6): 1122–1143.

41.

Somma

Bargsted

Disi Pavlic

, et al. (2021) No water in the oasis: the Chilean spring of 2019–2020. Social Movement Studies 20(4): 495–502.

42.

Sørensen

(2016) After the immigration shock: the causal effect of immigration on electoral preferences. Electoral Studies 44: 1–14.

43.

Sutton

Butcher

Svensson

(2014) Explaining political jiu-jitsu: institution-building and the outcomes of regime violence against unarmed protests. Journal of Peace Research 51(5): 559–573.

44.

Thaler

Mueller

Mosinger

(2023) Framing police violence: repression, reform, and the power of history in Chile. The Journal of Politics 85(4): 1198–1213.

45.

Visconti

(2021) Reevaluating the role of ideology in Chile. Latin American Politics and Society 63(2): 1–25.

46.

Wallace

Zepeda-Millán

Jones-Correa

(2014) Spatial and temporal proximity: examining the effects of protests on political attitudes. American Journal of Political Science 58(2): 433–448.

47.

Wasow

(2020) Agenda seeding: how 1960s black protests moved elites, public opinion and voting. American Political Science Review 114(3): 638–659.

48.

Whitehead

Fair

Payne

(2010) Violent Democracies in Latin America. Durham, NC: Duke University Press.

49.

World Bank (2023) “Urban Population Chile”. Retrieved from: https://data.worldbank.org/indicator/SP.URB.TOTL.IN.ZS?locations=CL. Accessed June 2025.

50.

Zechmeister

(2015) Left-right identifications and the Latin American voter. In: Carlin

Singer

Zechmeister

(eds) The Latin American Voter: Pursuing Represen- Tation and Accountability in Challenging Contexts. Ann Arbor, MI: University of Michigan Press, 195–225.

51.

Zhukov

(2023) Repression works (just not in moderation). Comparative Political Studies 56(11): 1663–1694.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.50 MB

0.00 MB

0.10 MB