Sage Journals: Discover world-class research

Abstract

In October 2016, the Colombian electorate narrowly rejected in a plebiscite the final agreement to end the conflict with the longest-running armed insurgency in the Western Hemisphere, the FARC. The plebiscite’s result provides a unique opportunity to assess dynamics in civil conflict termination preferences. I exploit the unexpected victory of the No vote, observed during the AmericasBarometer fieldwork, to estimate the effect of the uncertainty about the trajectory of the conflict generated by the outcome of the plebiscite. The unexpected defeat of the peace plebiscite did not measurably change the expressed support for the recently rejected agreement. However, it increased the public support for a negotiated rather than military settlement to the conflict, as well as respondents’ willingness to give concessions to FARC members. These findings inform the broader literature on civilian preferences toward civil conflict termination and compromise.

Keywords

Conflict termination civilian preferences peace agreements wartime public opinion

On September 21, 2016, the President of Colombia Juan Manuel Santos delivered his statement for the general debate of the UN General Assembly. Santos stood on the podium and told the world: “After more than half a century of internal armed conflict, I come back to the United Nations today, on the International Day of Peace, to announce, with all the strength of my voice and of my heart: The war in Colombia has ended.” Almost 2 weeks later, the administration presented the Colombian electorate a single-question ballot: “Do you support the final agreement to end the conflict and build a stable and lasting peace?” Against most pollsters’ predictions, 50.2% of voters rejected the peace agreement reached by the Colombian government and the Revolutionary Armed Forces of Colombia guerrilla (FARC). The victory of the No vote left national and international audiences in shock. After 4 years of negotiations and almost 2 years of a ceasefire from the rebel side, with some intermittent hostile actions, a lingering question materialized: “what now?”

In this article, I exploit the unexpected plebiscite results during the fieldwork of the AmericasBarometer to explore the effect of the exposure to uncertainty generated by the referendum defeat on conflict termination preferences. The identification strategy in this article relies on the salience and unexpectedness of the results of the plebiscite that split survey respondents into as-if-random treated and control groups. The findings of this analysis suggest that the uncertainty generated by the unexpected results of the vote informed individuals’ conflict termination preferences. Overall, the levels of support for a negotiated settlement to the conflict and willingness to give concessions to FARC combatants increased after the plebiscite. These results are robust to a range of model specifications and checks. While respondents’ expressed support for the signed agreement between the Santos government and the FARC remained unchanged, the findings contribute to larger debates in the conflict termination literature. These suggest that civilians’ support for negotiated resolutions to civil war is responsive to external cues, in line with previous findings (e.g., Fabbe et al., 2019). Importantly, these results also indicate that individuals are willing to update their preferences about the conditions and content of such agreements.

Peace processes and civilian conflict termination preferences

While most research about peace seeking has focused on the interplay between armed actors and governments, a growing body of the literature looks at the role of civilians and public opinion in peace processes (see Haass et al., 2022, for an overview). Generally, research has looked at the legacies of violence, proximity to the conflict, and victimization on social and political outcomes (Gallego, 2018; Haass and Ottmann, 2022; Weintraub et al., 2015). A key focus of this scholarship lies in the degree to which conflict-time experiences result in conciliatory and pro-social dispositions (Bauer et al., 2018; Hartman and Morse, 2020; Voors et al., 2012), instead of intransigent and confrontational ones (Balcells, 2012; Grosjean, 2014). Definitive answers have been elusive in this stream of work. Still, these dispositions have been associated with public preferences for conflict termination, suggesting that individuals’ support for negotiated resolutions is flexible (Fabbe et al., 2019) and linked to safety-maximization (Tellez, 2019a), as well as proximity, exposure to violence, trust in institutions, and preferences for punitive responses to crime (Hazlett, 2020; Montoya and Tellez, 2020).

Naturally, much of the research looking at civilian war termination preferences focuses on proximity to the conflict and victimization as explanatory characteristics for approval of negotiated transitions out of conflict. This research argues that civilians who bear the burdens of the war, have observed its consequences, or anticipate the potential for experiencing future harm showcase higher levels of support for a negotiated end to the conflict and willingness to compromise (e.g., Bakke et al., 2009; Hazlett, 2020; Kreiman and Masullo, 2020; Tellez, 2019a). At the same time, pursuing negotiated peace can also be a costly effort. Parties to the conflict need to sort how power ought to be shared (Mukherjee, 2006), what concessions to give (Matanock and Garbiras-Díaz, 2018), whether to accept external parties to ensure implementation (Karreth et al., 2023), amongst others. These costs can also influence civilian support for negotiated peace, which could explain why some civilians would choose not to support negotiations.

That said, we know little about how susceptible public preferences are to changes to the status quo. In this study, I evaluate the effects of an exogenous shock on levels of certainty about the course of the conflict in Colombia under a set of empirical expectations derived from this literature. The key feature in this setting is the uncertainty about the possible reactions by the government and the FARC guerrilla following the peace agreement’s rejection. In this case, the unexpected result of the Colombian peace plebiscite serves as a source of exogenous variation in the levels of certainty about the trajectory of the conflict and the possible responses by the parties to the agreement.

I argue that after a period of relative stability and expectations informed by one of the most codified peace agreements ever drafted (Quinn and Joshi, 2019), the potential for a return to confrontation, or even unknown alternatives, can shift civilians’ dominant interpretation of the cost calculus between a continuation of the war and a negotiated peace. Considering the evidence indicating that external cues can influence (a) attitudes toward compromise and the termination of conflicts through negotiation and (b) safety-maximization is an important feature in conflict opinion formation (Fabbe et al., 2019; Tellez, 2019a), I anticipate that the unforeseen outcome of the peace referendum altered respondents’ conflict termination preferences. Specifically, I expect respondents would update their preferences by increasing their support for the signed peace agreement, a negotiated end to the conflict, and becoming more willing to provide accommodations to FARC combatants.

The Colombian peace plebiscite

The peace process with the FARC is one of Colombia’s most relevant historical events. For decades, Colombia experienced a civil war between government forces and an array of left-wing guerrillas and right-wing paramilitary groups. Previous administrations achieved peace agreements with other factions of the conflict. Although most modern Colombian presidents took part in peace processes with one of the many armed groups active in the country, peace with the largest and longest-standing group, the FARC, had never materialized (Tellez, 2019b).

In mid-2012, the government announced the convening of formal peace talks with, at the time, the world’s oldest guerrilla group. Four years later, after what Flores and Vargas (2018) describe as a “peace process almost as turbulent as the war that preceded it,” the two parties to the talks announced the final agreement that stipulated the conditions under which FARC members would lay down their arms. President Santos’s government decided to put the final deal to a vote to address prospective legitimacy gaps (Matanock and García-Sánchez, 2017).

The general perception was that the plebiscite was going to be a formality. Though research suggests that support for the peace process overall was higher than support for its components (Matanock and Garbiras-Díaz, 2018), opinion polls prognosticated a Yes vote victory with a considerable margin (see SFA1). President Santos stated: “I think I know the Colombian people (…) I don’t have a plan B in case the plebiscite is not approved, because I think there will be no need for a plan B.”

To the shock of most vote observers, the October 2 plebiscite results did not correspond to the forecasts. The No option won by 56,000 votes—50.2% to 49.8%. Following the plebiscite, the Santos government moved to negotiations with the FARC and the No vote promoters. The new round of talks resulted in a revised final agreement signed on November 24, 2016, and ratified by the legislature later that week without a second popular vote.

Methods and data

Unexpected event during survey design

The Unexpected Event during Survey Design (UESD) exploits the happening of a salient and unexpected event, or treatment variable D, during the fieldwork of a survey to assess its effect on the outcome of interest Y. The rationale is that the event can naturally split respondents into as-if-random groups. In other words, the researcher can estimate the differences between the pre- and post-event respondents to assess the effect of D. This strategy has been used to determine the impact of terrorist attacks (Balcells and Torrats-Espinosa, 2018; Boydstun et al., 2018), corruption scandals (Ares and Hernández, 2017), and electoral results (Giani and Méon, 2021), among others.

Central to the validity of the identified estimates, there are two assumptions. First, excludability dictates that the only difference between the pre- and post-event groups stems from their state on D_i. This assumption could be violated if unrelated time trends or parallel events were to exist. Second, temporal ignorability states that individual potential outcomes Y_i should be independent of when the interview occurred. This assumption could be violated if there were significant differences in “pre-treatment” covariates (reachability bias) and the outcome variables were not robust to correcting for imbalance (conditional ignorability) (Muñoz et al., 2020).

Data

I employ data from the 2016 AmericasBarometer Colombian rounds. The survey consisted of face-to-face interviews with a total of N = 1563 respondents drawn from a national probability sample of voting-age individuals fielded between August 3, 2016, and October 29, 2016. I focus on country-specific modules regarding the respondents’ conflict termination preferences and willingness to provide concessions to combatants as outcome measures. Additionally, I utilize a set of individual-level variables for covariate adjustment.

I employ age, sex, years of education, support for the incumbent (Santos), income group, and an urban/rural marker. In addition to demographic covariates, the survey contains victimization and proximity to the conflict markers, which I leverage in the analysis. I derive a measure of conflict exposure—conflict-zone—from the municipality where the respondent resides. The conflict-zone identification, used in previous studies (see Matanock and García-Sánchez, 2018; Tellez, 2019a), is based on the government’s Espada de Honor security program. The direct measure of victimization comes from the self-reported membership of the Colombian Registry of Victims (RUV). Finally, I obtain a proximity-to-the-conflict metric from self-reported victimization of the subject’s relatives by FARC members.

Figure 1 illustrates the split into pre- and post-event groups—(D₀ = 1361 and D₁ = 202)—generated by the timing of their interviews. Supplemental Table A2 summarizes the balance of the set of pre-treatment covariates utilized in the analysis. The core of the variables is balanced between the pre- and post-plebiscite groups. The only statistically significant differences between the two sample groups relate to a larger proportion of respondents from conflict-zones. I address potential threats to the assumption of the ignorability of the treatment assignment emanating from the regional composition of the groups by assessing the robustness of the estimates to alternative specifications proposed by Muñoz et al. (2020). The latter is addressed in Supplemental Appendix A.

Figure 1.

Timeline of the AmericasBarometer Fieldwork. Treatment groups are assigned based on the timing of their interview in relation to the referendum. D₀ includes units from Aug 3 to Oct 2 (≈9 weeks) and D₁ from Oct 3 to Oct 29 (≈4 weeks).

Analysis and results

To analyze the sequels of the uncertainty generated by the plebiscite results, I fit a series of OLS regressions on the support for the signed agreement, the negotiated end to the conflict, and a composite measure of concessions to FARC combatants. I ran three model specifications: (a) a naive model, (b) a model with demographic, and (c) a model with demographic and conflict controls.

First, to assess conflict termination preferences, I employ a direct measure of the extent to which respondents supported the agreement reached by the Santos government and the FARC. Second, I leverage a binary variable that represents whether the respondent considers negotiation, as opposed to military force or a combination of both, to be the best option to end the conflict. Third, to assess the respondent’s willingness to provide concessions to combatants, I obtain a metric from five items that measure agreement with concession scenarios embedded in the agreement, ranging from political, territorial, and jail-time compromises (Cronbach’s α = 0.71). Table 1 presents the items and their pre- and post-plebiscite levels. I scale the measures of support for the signed agreement and concessions, in terms of standard deviations, initially measured on a 7-point scale.

Table 1.

Overview of the outcomes of interest. The table presents the average observed values for the pre- and post-referendum groups. The values for the negotiated end item represent the proportion of respondents stating such preference in each group.

Item type	Question	Pre	Post	Change
Support Santos government-FARC agreement	The government of President Juan Manuel Santos and the FARC signed a peace agreement. To what extent do you support this peace agreement?	3.7/7	4/7	+0.2pts
Negotiated end	Of the following options for resolving the conflict with the guerrilla [FARC], which do you think is the best? [Negotiation]	0.66	0.78	+12pp
Political concession	The Colombian government must guarantee that demobilized FARC members can participate in politics. To what extent do you agree or disagree?	2.8/7	3.2/7	+0.4pts
Political concession	It was agreed that political parties made up of demobilized ex-combatants of the FARC will have the same guarantees of security and access to the media that other political parties already have. To what extent do you agree or disagree?	3.4/7	3.8/7	+0.4pts
Judicial concession	It was agreed that the members of the FARC, who are responsible for atrocious crimes and who confess their crimes, pay between five (5) and eight (8) years of deprivation of liberty. To what extent do you agree or disagree?	4.4/7	5/7	+0.6pts
Judicial concession	It was agreed that the rank and file members of the FARC (not commanders) who demobilize will not go to jail. To what extent do you agree or disagree?	2.7/7	3.2/7	+0.5pts
Territorial concession	It was agreed that the members of the FARC will concentrate in some regions of the country. To what extent do you agree or disagree?	3.4/7	3.9/7	+0.5pts

Figure 2 presents the estimates for each outcome of interest from the three model specifications. The first panel presents the results for respondents’ expressed support for the signed peace agreement between the Santos government and the FARC. Contrary to my expectations, there are no discernible differences between the pre- and post-plebiscite groups in their expressed support for the signed peace agreement between the Santos government and the FARC. While the coefficients tend to the expected direction, they are not statistically significant at conventional levels.

Figure 2.

Estimated effect of the referendum results on public support for the signed peace agreement, a negotiated settlement to the conflict, and willingness to give concessions to FARC combatants under three model specifications. Estimates from OLS regression models with heteroscedasticity-consistent standard errors for negotiated end and concessions with 95% confidence intervals. The estimates for the support for the signed peace agreement and the composite concession measure are presented as shifts in standard deviations.

On the other hand, in line with my expectations, the results show that exposure to the peace plebiscite increased support for a negotiated end to the conflict with the FARC and willingness to compromise. Based on the estimates, there was around a 10-percentage-point increase in the support for a negotiated end to the conflict in the post-plebiscite period. Additionally, the composite measure of political, territorial, and jail-time concessions to FARC combatants increased by around 0.2 standard deviations. In addition to the scale measure, I test a latent measure of willingness to provide concessions based on the first component of PCA and each concession item individually. The tests render comparable results to the primary model (see Table C3 and Figure C1 of the Supplemental Appendix).

These results present a mixed picture between what can be understood as a direct measure of the Havana accords, some of its components, and more abstract termination preferences. On one side, respondents’ support for the specific agreement settled between the government and the insurgent group did not significantly change. On the other, following the plebiscite, respondents expressed more support for concessions to FARC combatants embedded in said agreement and more widespread preference for a negotiated end to the conflict with the armed group. Even more, respondents in the post-plebiscite period are more willing to believe that following the agreement, FARC members will uphold their compromise to fully demobilize and cease drug trafficking (see Figure B1 in the appendix). In other words, there is a seeming separation between respondents’ conflict termination preferences, belief that the parties will uphold the accord, and their expressed support for the agreement to be ratified.

Robustness checks

The identification strategy relies on the unexpectedness, uniqueness, and salience of the plebiscite results, as well as excludability and temporal ignorability assumptions. For instance, exclusion restriction could be threatened if respondents’ treatment state affected conceptually unrelated variables. To assess that the interview’s timing only affects the outcome variables through D_i, I employ measures of conflict-termination preferences with competing armed groups and support for other agreement provisions unrelated to FARC combatants. The AmericasBarometer included a question about conflict termination preferences with the Bacrim (Bandas Criminales), primarily perceived as legacy groups from the peace process with far-right paramilitaries in 2006. Additionally, as part of the agreement-related battery, the respondents were surveyed about their support for provisions embedded in the agreement unrelated to concessions to FARC combatants, such as support for land redistribution, the treatment of military and police personnel involved in atrocious crimes in the context of the conflict, and the legality of coca leaf plantation (see Table C2 for an overview of the items included in this measure.) The uncertainty generated by the plebiscite results would be expected to inform preferences toward the FARC and not the alternative groups. In alignment with the expectation, Figure 3 showcases the exposure to the plebiscite results did not increase support for a negotiated end to the conflict with the Bacrim nor non-FARC-related provisions of the agreement.

Figure 3.

Estimated effect of the referendum results on public support for a negotiated settlement with the BACRIM and support for non-FARC-related agreement provisions under three model specifications. Estimates from OLS regression models with heteroscedasticity-consistent standard errors with 95% confidence intervals.

Furthermore, I systematically assess and address possible violations of the identification assumptions following all recommendations and tests by Muñoz et al. (2020). The results obtained from the main models remain robust to further checks—placebo events set at different dates prior to the plebiscite, alternative specifications, multiple bandwidths (±7 and 15 days), non-response analysis, and falsification checks through previous rounds of the AmericasBarometer in Colombia (see ST A1 for an overview of the checks and results).

Discussion and conclusion

The Colombian peace plebiscite provides a unique opportunity to assess conflict termination preference formation. Since its emergence, the Colombian civil war has paradoxically coexisted with democratic politics and institutions. The unexpected rejection of the peace agreement via democratic means somewhat fits the paradoxical nature of the conflict. Although this armed conflict has distinctive features, in many ways, the Colombian case shares characteristics with other civil wars (Matanock and García-Sánchez, 2017).

I leveraged the plebiscite’s unexpected result as an exogenous source of variation of uncertainty about the trajectory and termination of the conflict. The identification strategy provides estimates with substantial internal and external validity. In line with previous research, the results suggest that, to some extent, individuals inform their termination preferences based on external cues, in this case, from the referendum results. Somewhat puzzling, however, is that respondents’ support for the recently rejected Havana agreement did not change after the plebiscite. It is possible that the contrasting signals, that is, increased support for procedural features and components of the agreement and the unchanged support of the specific agreement framed as “the Santos’ government and FARC accord,” are a reflection of the highly politicized and polarized nature of the specific agreement. While endorsing the signed Havana accord is an indicator of conflict termination preferences, it is potentially a composite tainted by contextual and political considerations.

Further, it is important to note that this study has some limitations. First, the identification strategy relies on the ignorability of the treatment assignment and excludability assumptions. Although the sample is balanced across most dimensions, the post-referendum group had more respondents from conflict regions. I provide auxiliary evidence suggesting that individuals in these regions do not drive the observed results and results are robust to further checks. Still, this remains an important qualification for the analysis. Second, the post-plebiscite group is substantially smaller than its counterpart. Given these setup, sample size, and statistical power constraints, it is not possible to study the more local-specific dynamics to the Colombian conflict (Kreiman and Masullo, 2020) nor the heterogeneity of these effects. Finally, it is possible that these dynamics can work differently in identity-based conflicts compared to this agrarian conflict (Hirsch-Hoefler et al., 2016). That said, these findings provide evidence to inform the broader literature of civilian preferences for conflict termination and compromise.

The analysis suggests that individuals in the post-plebiscite group were, on average, more supportive of a negotiated end to the conflict and more willing to provide concessions. These results favor the pro-peace and “war weariness” arguments in conflict termination preference formation (Hazlett, 2020; Tellez, 2019a). In most instances of civil strife, the burdens of conflict are not evenly distributed amongst all individuals. In this case, we observe how individuals react to exogenous changes to these costs. The unexpected victory of the No vote took the political elite and public by surprise. This case provides a unique opportunity to study sudden changes in the balance during civil war negotiations and how these can affect public preferences for conflict termination.

The peace plebiscite in Colombia illustrates the relevance of seeking to understand the dynamics behind opinion formation in civil conflict. While the government conceived the vote to validate the agreement as a formality to address legitimacy gaps, the electorate attached diverse considerations when casting their votes during a period of relative stability.

Supplemental Material

Supplemental Material - The unexpected results of the peace referendum changed conflict termination preferences in Colombia

Supplemental Material for The unexpected results of the peace referendum changed conflict termination preferences in Colombia by Sebastian Ramirez-Ruiz in Research & Politics

Footnotes

Acknowledgments

I would like to express my gratitude to Simon Munzert, Francesca Minetto, Lisa Oswald, and the participants of the Latin American Peace Science Society Annual Conference and AFK-Empirical Peace and Conflict Workshop for their valuable feedback on this manuscript.

Declaration of conflicting interests

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

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Sebastian Ramirez-Ruiz

Supplementary material

The supplementary material for this article can be found at .

References

Ares

Hernández

(2017) The corrosive effect of corruption on trust in politicians: evidence from a natural experiment. Research & Politics 4(2): 205316801771418.

Bakke

O’Loughlin

Ward

(2009) Reconciliation in conflict-affected societies: multilevel modeling of individual and contextual factors in the north caucasus of Russia. Annals of the Association of American Geographers 99(5): 1012–1021.

Balcells

(2012) The consequences of victimization on political identities: evidence from Spain. Politics & Society 40(3): 311–347.

Balcells

Torrats-Espinosa

(2018) Using a natural experiment to estimate the electoral consequences of terrorist attacks. Proceedings of the National Academy of Sciences of the United States of America 115(42): 10624–10629.

Bauer

Fiala

Levely

(2018) Trusting former rebels: an experimental approach to understanding reintegration after civil war. The Economic Journal 128(613): 1786–1819.

Boydstun

Feezell

Glazier

(2018) In the wake of a terrorist attack, do Americans’ attitudes toward Muslims decline? Research & Politics 5(4): 205316801880639.

Fabbe

Hazlett

Sınmazdemir

(2019) A persuasive peace: Syrian refugees’ attitudes towards compromise and civil war termination. Journal of Peace Research 56(1): 103–117.

Flores

Vargas

(2018) Colombia: democracy, violence, and the peacebuilding challenge. Conflict Management and Peace Science 35(6): 581–586.

Gallego

(2018) Civil conflict and voting behavior: evidence from Colombia. Conflict Management and Peace Science 35(6): 601–621.

10.

Giani

Méon

(2021) Global racist contagion following Donald Trump’s election. British Journal of Political Science 51(3): 1332–1339.

11.

Grosjean

(2014) Conflict and social and political preferences: evidence from World War II and civil conflict in 35 European countries. Comparative Economic Studies 56(3): 424–451.

12.

Haass

Ottmann

(2022) The effect of wartime legacies on electoral mobilization after civil war. The Journal of Politics 84(3): 1322–1336.

13.

Haass

Hartzell

Ottmann

(2022) Citizens in peace processes. Journal of Conflict Resolution 66(9): 1547–1561.

14.

Hartman

Morse

(2020) Violence, empathy and altruism: evidence from the Ivorian refugee crisis in Liberia. British Journal of Political Science 50(2): 731–755.

15.

Hazlett

(2020) Angry or Weary? How violence impacts attitudes toward peace among Darfurian refugees. Journal of Conflict Resolution 64(5): 844–870.

16.

Hirsch-Hoefler

Canetti

Rapaport

, et al. (2016) Conflict will harden your heart: exposure to violence, psychological distress, and peace barriers in Israel and Palestine. British Journal of Political Science 46(4): 845–859.

17.

Karreth

Quinn

Joshi

, et al. (2023) International third parties and the implementation of comprehensive peace agreements after civil war. Journal of Conflict Resolution 67(2-3): 494–521.

18.

Kreiman

Masullo

(2020) Who shot the bullets? exposure to violence and attitudes toward peace: evidence from the 2016 colombian referendum. Latin American Politics and Society 62(4): 24–49.

19.

Matanock

Garbiras-Díaz

(2018) Considering concessions: a survey experiment on the Colombian peace process. Conflict Management and Peace Science 35(6): 637–655.

20.

Matanock

García-Sánchez

(2017) The Colombian paradox: peace processes, elite divisions & popular plebiscites. Dædalus 146(4): 152–166.

21.

Matanock

García-Sánchez

(2018) Does counterinsurgent success match social support? Evidence from a survey experiment in Colombia. The Journal of Politics 80(3): 800–814.

22.

Montoya

Tellez

(2020) Who wants peace? predicting civilian preferences in conflict negotiations. Journal of Politics in Latin America 12(3): 252–276.

23.

Mukherjee

(2006) Why political power-sharing agreements lead to enduring peaceful resolution of some civil wars, but not others? International Studies Quarterly 50(2): 479–504.

24.

Muñoz

Falcó-Gimeno

Hernández

(2020) Unexpected event during survey design: promise and pitfalls for causal inference. Political Analysis 28(2): 186–206.

25.

Quinn

Joshi

(2019) Transitional justice in the Colombian final accord: text, context, and implementation. In: DeMeritt

JHR

Meernik

Uribe-López

(eds) As War Ends: What Colombia Can Tell Us about the Sustainability of Peace and Transitional Justice. Cambridge: Cambridge University Press, 207–229.

26.

Tellez

(2019a) Peace agreement design and public support for peace: evidence from Colombia. Journal of Peace Research 56(6): 827–844.

27.

Tellez

(2019b) Worlds apart: conflict exposure and preferences for peace. Journal of Conflict Resolution 63(4): 1053–1076.

28.

Voors

Nillesen

EEM

Verwimp

, et al. (2012) Violent conflict and behavior: a field experiment in Burundi. The American Economic Review 102(2): 941–964.

29.

Weintraub

Vargas

Flores

(2015) Vote choice and legacies of violence: evidence from the 2014 Colombian presidential elections. Research & Politics 2(2): 205316801557334.

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.

0.00 MB

3.92 MB