The Emergence and Consolidation of Opposition to Authoritarian Rule (Book Project)

Under what conditions does the opposition survive despite extreme repression? How does government violence mobilize the opposition? When are we likely to observe the consolidation of mass mobilization? My book project examines these three questions about the relationship between repression and mobilization in Chile during the Pinochet dictatorship (1973-1990).

The first main argument is that as long as state violence is directed indiscriminately or collectively to large segments of the population we are extremely unlikely to observe mass mobilization. We are also unlikely to observe public mobilization by the top targets of a capable state. If they continue organizing at all, the top targets will do so underground. In clandestinity, opposition groups with a compartmentalized organizational structure and underground organizing capability are likely to have a higher survival rate than those without those qualities. The larger the ideological distance between the regime and the opposition group the more likely the opposition group is to have a compartmentalized structure and underground organizing capability. Therefore, ideologically more extreme opposition groups will tend to have higher rates of survival than more moderate ones. This theoretical framework explains the puzzle about the Revolutionary Leftist Movement (MIR), an ultraleftist political movement that had a comparatively high rate of survival even though it was among top target of the secret police.

But even collective and largely indiscriminate violent repression does not preclude all actors from engaging in public opposition. The relatives of the disappeared organized the first documented public protests after the coup in Chile, demanding information about the whereabouts of their loved ones and denouncing the regime’s human rights record. Labor unions, despite being associated with top targets of the regime such as the Communist Party, were sometimes able to organize public demonstrations of opposition. These groups were able to publicly oppose the regime, albeit at a small scale and by making modest demands, thanks in large part to protector institutions. Catholic priests at the local level and the Cardinal’s Vicariate of Solidarity at the national level played the role of a protector institution in this case—an organization that the government needs for legitimacy and that is willing to lend some safeguards to the opposition. Protector institutions lower the cost of high-risk activism and increase the cost of state repression. In communities that suffered high levels of repression but where the local priest was pro-opposition, the relatives were able to organize public protests. Localities with pro-Pinochet priests remained demobilized. This explanation helps solve the puzzle of public demonstrations in some localities at a time when the state was using indiscriminate and generalized collective violent repression against its citizens.

Finally, under what conditions are we likely to observe the consolidation of a mass movement? I argue that mass mobilization emerges through a process of alliance formation. Without alliances mass mobilization is not possible because there is usually no single group in society with the convening power to rally hundreds of thousands or millions of people. There are two main competing considerations for any civil society group confronted with the choice to form or join an alliance. The first is reducing the probability of suffering violent repression and the second is achieving political success. Therefore, a reduction in violent repression allows civil society organizations to forge alliances, making mass mobilization more likely.

I develop and test these arguments at the subnational level in the case of the Pinochet dictatorship in Chile from 1973 to 1990. The manuscript presents data from four original datasets based on tens of thousands of pages of archives from Chile and the United States, and more than 50 interviews with the protagonists of the conflict conducted during eight months of fieldwork in Chile. The datasets contribute to our understanding of the 18-year period of Pinochet’s rule at different levels of analysis. It triangulates data from various sources and overcomes important biases in the literature such as the focus on large and successful movements, the aggregated measurement of repression, and the assumption that the Catholic Church is a unitary actor. I partly address these challenges by dramatically expanding the range of opposition groups under study, and by including not only major movements, but also small and failed attempts. I also use micro-level and time-series data, some of which allow me to employ a regression discontinuity design. These methods and data help address the endogenous relationship between repression and mobilization.

Working Papers

“Explaining the Survival and Resilience of Opposition Groups in Chile during the Pinochet Dictatorship” (under review)

Why do some opposition groups survive government repression while others get eliminated? This paper argues that a group’s ideology conditions its organizational structure and underground organizing capacity, in turn affecting survival. Extreme groups tend to develop a compartmentalized structure with militants skilled in underground organizing. Compartmentalization and underground organizing decrease the probability of capture, as well as mitigate the downstream effects of captures. Using a novel dataset of individuals on Pinochet’s wanted lists and the victims of the Chilean dictatorship, this paper demonstrates that the victimization of ultraleftists is significantly lower than that of more moderate—but similarly targeted—groups. Archival and interview data show that differences in survival are due to organizational structure and skills, and that these characteristics flow from ideology. In contrast to other research on repression, this study compares the intended-to-repress and repressed populations to better understand the heterogeneous effects of violence.

“The Hydra Effect: When Repression Creates New Opposition against Authoritarianism”

State repression has been found to demobilize the opposition, in other cases it has a positive effect on mobilization, and in yet other studies, dissidents have been found to avoid repression altogether. Rather than adjudicating between these findings, this paper takes the position that repression has a multiplicity of effects, and that observing certain patterns over others depends on the conditions. Repression causes new nonviolent and public opposition when potential dissidents are able to secure some level of protection from local protector institutions. “Protector” institutions are those that the government needs for legitimacy and that are willing to lend some safeguards to the opposition. I test this hypothesis in the case of Chile during the Pinochet dictatorship (1973-1989). After geocoding all the Catholic temples in the most populated state, along with the political tendency of each priest at the year-temple level, and providing indicators for whether or not new opposition groups formed, this paper finds that the likelihood of mobilization dramatically increases when the group can count on the support of a local Catholic priest, controlling for repression, the opposition group’s ideology, the political party of the mayor, and whether the locality is mainly rural or urban. Interview and archival data also show that in communities where the local priest was pro-Pinochet, opposition groups were not able to organize even though they had attempted to do so.

“Repression and the Onset of Nonviolent and Violent Movements”

Do levels of government repression condition the tactics that opposition groups use to wage political struggle? Conventional wisdom suggests that popular uprisings against repressive regimes are more likely to be violent than nonviolent, and that if a peaceful struggle happens to emerge in such a difficult environment, it will be quickly put down or morph into an armed struggle. This article tests the conventional wisdom of violence begetting violence with the Nonviolent and Violent Outcomes Dataset (NAVCO 2.0), which contains yearly information on 250 nonviolent and violent mass movements for regime change, anti-occupation, and secession from 1945 to 2006, as well as with repression data from the CIRI Human Rights Data Project and the Political Terror Scale. NAVCO 2.0, however, is adjusted to include all the negative cases—the country-years with no nonviolent or violent campaigns—to be able to compare the conditions of onset and no onset countries. Close analysis of the level and type of government repression in the years prior to major social movements reveals a more nuanced pattern than that suggested by the conventional wisdom. While in highly repressive countries the dominant form of popular struggle that emerges is armed rebellion, the proportion of primarily nonviolent movements does not substantially vary by level of lagged repression. When tactics are disaggregated to distinguish between nonviolent movements and nonviolent campaigns with violent flanks, it becomes apparent that in the most repressive countries there will be a much higher proportion of armed insurrections and mixed movements than purely nonviolent campaigns. This research shows that nonviolent struggles do emerge under very difficult conditions of government repression, though these primarily nonviolent movements tend to have violent wings that are often used as defense groups. In addition, the patterns of lagged repression and movement onset suggest that there is a threshold after which the level of government oppression selects out civil resistance.



“Chapter 3: Can We Live Together? Citizen Insecurity as a Threat to Democratic Coexistence,” in Kevin Casas Zamora, The Besieged Polis: Citizen Insecurity in Latin America, Latin America Initiative Report, Brookings, July 8, 2013  (with Kevin Casas-Zamora).

Policy and Opinion

What is happening in Venezuela? Political Violence @ A Glance, February 2, 2019.

“Repression, Resilience, and Mass Movements: A Page from Chilean History” ICNC Minds of the Movement Blog, December 1, 2017.

“Conversations with Experts on the Future of Central America,” Latin America Initiative Report, Brookings, November 19, 2012 (with Diana Negroponte and Alma Caballero).

“Venezuela: What Future for Chavismo without Chavez?” Opinion Editorial, Brookings, July 25, 2012 (with Diana Negroponte).

“The 2012 Venezuelan Elections: Hopes for Legitimacy,” Opinion Editorial, Brookings, March 9, 2012 (with Diana Negroponte).

Works in Progress

“Can Nonviolent Discipline be Taught in the Classroom? An Evaluation of the United States Institute of Peace’s Civil Resistance Educational Initiative”

“Generalizing from Unequal Probability Sampling” (with Peter Aronow and Fredrik Savje)

“Transparency in the Social Sciences” (with Peter Aronow, Tommaso Bardelli, Josh Kalla, and Hilton Simmet)

“Strategic Watersheds in the Life of a Rebel Group: When Armed Groups Incorporate Nonviolent Resistance or Splinter” (with Victoria McGroary)