What do politicians know about citizens’ preferences for peace agreement provisions?

EGAP researchers: Aila Matanock; Natalia Garbiras-Díaz

Other researchers: Miguel García-Sánchez

Key takeaways: Legislators are key actors in the implementation of peace agreements. Yet they can be misinformed about citizens’ attitudes toward the provisions in these settlements and their own policy positions may not be responsive to information about the true level of support among their constituents.

Geographical Region: Latin America

Type of study: Survey experiment

Preparer: Mark Williamson

August 6, 2024

Executive Summary

This study investigates what legislators know about their constituents’ attitudes toward the implementation of peace agreements. These elites play a crucial role in translating the provisions of a peace accord into tangible legislation, yet it is unclear how citizens’ preferences shape legislators’ decisions about whether to support specific policies under the agreement. Looking at Colombia’s implementation of its 2016 settlement, the authors find that members of Congress greatly underestimate popular support for at least one particular provision of the agreement related to post-conflict electoral reform. Importantly, these misperceptions are related to politicians’ own positions on the issue, with those who support (or oppose) the provision estimating that citizens are more likely to support (or oppose) it as well. Using an experimental intervention, the authors find that providing factual information about citizens’ actual attitudes toward the provision does not cause politicians to update their own stances on the issue.

Policy Challenge

After a peace agreement is signed, implementation of its various provisions often requires the legislative branch to pass individual pieces of legislation. Members of the legislature are therefore often the key actors determining the fate of an agreement and how faithfully it is put into law. Yet we know little about what factors legislators consider when deciding whether to support, alter or oppose a policy that was originally set forth in a peace agreement. To what extent are their positions responsive to citizens’ attitudes toward specific provisions, including those citizens most exposed to the conflict?

While a considerable literature in political science has identified a strong link between citizen preferences and legislator activity on a range of policy issues, there are reasons to suspect that this correspondence will be weaker for policies connected to a peace agreement. For one, these settlements are often highly technical and complex, so citizens may not be sufficiently well informed to hold their representatives accountable for their policy positions. And while peace agreements are often highly salient when they are first signed, the public’s interest can wane over time as the implementation process drags on. As a result, politicians may not expect to be held to account on implementation issues when they are no longer under intense public scrutiny. In post-conflict settings, elites also tend to take polarizing stances as a negotiation tactic, meaning their positions may be unrelated to citizen attitudes. Finally, while politicians can often gauge citizens’ general opinions toward a peace agreement in public opinion polls, they rarely know their preferences for specific provisions within the agreement, making it difficult for those attitudes to inform their decisions on specific implementation items.

In sum, legislators may be either uninformed about citizen preferences over peace agreement provisions or else incentivized to ignore the public’s attitudes. To assess the extent of these barriers to citizen input, the authors ask two key questions:

i) How accurate are politicians’ perceptions of citizen attitudes toward specific provisions in peace accords?
(ii) Are their positions on those provisions responsive to information about citizen preferences?

Context

The authors explore these questions in Colombia, where the government and Revolutionary Armed Forces of Colombia (FARC) signed a peace agreement in 2016 that sought to end a civil conflict that had lasted more than 50 years. After a failed plebiscite on the agreement and the ratification of a revised text in Congress later that year, the settlement remained a central issue in the 2018 presidential and parliamentary elections.

After the peace agreement was approved by Congress, the governing coalition sought to implement the provisions of the accord through individual pieces of legislation. In this study, the authors focus on one particular provision of the agreement: the creation of a special 16-seat “peace district” in the House of Representatives to increase representation of citizens living in areas that were most affected by the conflict. At the time of the authors’ data collection, proposals had been made to implement this provision, but no legislation had been passed and legislators continued to disagree about whether and how best to follow through on this policy.

This provision became gradually more popular among citizens since the agreement was signed in 2016. Survey data indicates that just over half of Colombians supported its implementation at the time of the study, a figure which rises to around 70% among citizens living in conflict-affected areas.

Research Design

To investigate elite perceptions of citizen support for the peace agreement provision, the authors partnered with a local NGO, the Misión de Observación Electoral (MOE), to embed experimental questions in the 2019 wave of the MOE’s regular survey of all members of Congress in Colombia. In total, 250 of the country’s 267 legislators elected in 2018 participated in the face-to-face survey.

Politicians were block-randomized based on their party and chamber of Congress into one of three versions of the questionnaire:

1. Control: Legislators were asked to estimate the percentage of citizens who support the provision (a) in Colombia as a whole and (b) in conflict-affected areas.
2. Information I: Legislators were asked to read a vignette describing the percentage of Colombians who support the provision (based on data from the most recent Colombia LAPOP survey conducted in 2016).
3. Information I: Legislators were asked to read a vignette describing the percentage of Colombians living in conflict-affected areas who support the provision (based on data from a survey conducted in 2017 by the Observatorio de la Democracia, a research center in charge of conducting the LAPOP surveys in Colombia, with a sample representative of these areas).

All legislators then state how strongly they agree or disagree with the creation of the 16 special seats. The authors use the estimates from the control group to descriptively characterize baseline beliefs about citizen preferences among legislators and benchmark them against the actual figures from in-country public opinion surveys that are representative both of the country as a whole and of citizens living in conflict-affected areas. They then compare average support for the specific peace agreement provision across legislators in each experimental condition to determine whether the information about public support changed politicians’ own stances on the provision.

Results

The authors first examine politicians’ beliefs about citizen preferences, finding that legislators in Colombia underestimated support for the peace agreement provision related to reserved seats. On average, the politicians surveyed estimated that 38% of Colombians supported the provision, although the true percentage was just over 54%. Similarly, among citizens living in conflict-affected areas, legislators believed that 60% supported the introduction of reserved seats in their areas, which was 8% lower than the true figure.

These misperceptions are correlated with the positions of legislators’ parties toward the agreement. Legislators from parties that were in favor of the settlement had beliefs about citizen support for the provision that were relatively close to the true level of support. In contrast, members of parties that opposed the peace agreement tended to systematically underestimate support for the provision. Misperceptions were also larger for politicians who did not represent constituents from conflict-affected areas.

Unsurprisingly, legislators’ misperceptions of public attitudes are associated with their own positions toward the peace agreement provision: those who believe citizen support for the policy is lower also report lower support themselves. Can providing information about the true level of support change their positions? The authors find that, on average, their informational intervention describing public support for the provision did not cause legislators to meaningfully update their positions on this policy. There is suggestive evidence that those legislators who belonged to parties that supported the peace agreement increased their support for the provision after the treatment, but this finding is subject to statistical uncertainty.

Lessons

This study offers lessons for policymakers working on both conflict resolution and citizen accountability more broadly. First, the results here suggest that while citizens’ attitudes may play important roles in the implementation of peace agreements, their influence does not appear to operate by changing legislators’ positions on specific provisions that need to be passed into legislation. Second, the null results of the authors’ experimental intervention stand in contrast to previous research illustrating how factual information about citizen preferences can shape politicians’ own policy positions. One explanation for this finding is that on policies stemming specifically from peace agreements, legislators may not have incentives to learn about or respond to citizen concerns in the ways that have been identified in prior research.