Sage Journals: Discover world-class research

Abstract

Civil conflicts typically end with negotiated settlements, but many settlements fail, often during the implementation stage when average citizens have increasing influence. Citizens sometimes evaluate peace agreements by voting on referendums or the negotiating leaders, and, almost always, they decide whether to cooperate. Yet, despite their role, we do not know much about how citizens form attitudes toward peace agreements. In this article, we assess how citizens form attitudes toward settlements, specifically the policy provisions that emerge from them, which are central in shaping the post-conflict context. These are complex policy changes, involving deeply factionalized actors, and the citizens evaluating them are often focused on rebuilding their lives. We therefore theorize that citizens use stark cues from political elites with whom they have affinity to form their attitudes. We test our theory using survey experiments in Colombia. We find that citizens rely on political elites’ cues to decide their stance on the settlement’s provisions. These cues appear to supply easily-accessible information that respondents use over other information. In contrast to work positing that peace agreements are exceptional and weary citizens are stabilizing forces, our results suggest that even these decisions are politics as usual, where divisions among political elites drive the outcome.

Keywords

attitudes Colombia elite cues peace agreements Revolutionary Armed Forces of Colombia

Introduction

Intrastate wars have produced many more deaths than interstate conflicts over the past half-century (Fearon & Laitin, 2003; Pettersson, Högbladh & Öberg, 2019). Most civil conflicts now end with negotiated settlements, but many settlements fail, often during their implementation (Collier et al., 2003; Jarland et al., 2020). The settlement’s terms matter to successfully stabilizing peace (Hartzell & Hoddie, 2007) – but the implementation process also has a profound effect on conflict recurrence (e.g. Nilsson, 2012).

What factors shape the terms and the implementation of a settlement? Increasingly, citizens’ attitudes matter because they are involved in peace processes, particularly in implementation (see Footnote 5). Yet, we still do not know much about how citizens form attitudes toward peace processes and agreements. A widespread theory about support for settlements focuses on exposure to violence, but the evidence is inconclusive.¹ Other factors, such as rural vs. urban residence, have some purchase but merit more exploration (Tellez, 2019).

We shift the focus from some existing structural arguments and posit that citizen support depends crucially on elite cues. Our theory’s basic premise is that acquiring information on provisions emerging from settlements is costly because these are part of complex policy changes, involving multiple, often deeply factionalized, actors.² Under these conditions, we argue that citizens use heuristics such as political elites’ cues that are clear because political factions are starkly articulated around peace processes (Sniderman, Brody & Tetlock, 1993; Kahneman, 2011).

However, our theory, and the general idea that citizens follow elite cues on settlements, is not a foregone conclusion. Other work hypothesizes that peace processes are unique settings with different dynamics from other political processes (e.g. Kalyvas, 2006; Wood, 2003). Consider two prevalent models: first, citizens are seen as war-weary and therefore supportive of any valid peace process (e.g. Zartman, 1995), collectively observing and rewarding compliance with implementation by combatant parties (Przeworski, 2005; Fearon, 2011; Wantchekon, 2004). In this view, citizens support settlements and provide the mechanism for ‘self-enforcing’ agreements. Second, and in contrast, citizens are seen as having heightened negative emotions that lead them to oppose any compromise in factionalized contexts.³ In this view, citizens are unlikely to support settlements and, instead, provide a mechanism for undermining them. While we find some evidence that citizens are generally supportive of agreements – perhaps due to war weariness – our theory differs from these views by arguing that citizens follow different factions that support and oppose the process. Our framework also contrasts more broadly with ideas that citizens think differently during peace processes, that these are somehow extraordinary circumstances that change how they operate, either fixing their views or encouraging them to sort through information fully. Thus, our theory can also explain significant variation in support for different terms as implementation occurs, something the broader literature often cannot do (see Online appendix B).

We test our theory using survey experiments during the implementation of Colombia’s peace agreement between the government and the Revolutionary Armed Forces of Colombia (FARC). We examine views of citizens most affected by the fighting, who would perhaps be most likely to see the peace agreement as different from another political process, making it a potentially hard test of our elite cues’ theory. We assess the effect of elite cues on voters’ preferences toward provisions of the settlement, dependent on their political affinity with elites.⁴

Our empirical work, including the case selected, builds on and contributes to a growing body of literature on public opinion about peace processes (Assouline & Trager, 2017, 2015; Fabbe, Hazlett & Snmazdemir, 2019; Kao & Redlich Revkin, 2018; Matanock & Garbiras-Daz, 2018; Tellez, 2019; Haas & Khadka, 2020; Masullo & Morisi, 2019; Carlin, McCoy & Subotic, 2019). Matanock & Garca-Sánchez (2017) provide observational evidence that public opinion towards the settlement between the FARC and the Colombian government aligned with elites’ positions. We further develop the theory and then experimentally test how attitudes on these complex policies develop in this factionalized environment. Moreover, in conflict contexts, the literature so far typically focuses on the negotiation, not the implementation period. Nevertheless, both are crucial to the success of agreements and subsequent peace, as citizens play a role in implementation regardless of whether they are incorporated into peace processes. We, therefore, focus on implementation.

The data show that citizens’ attitudes toward implementing provisions that emerged from the 2016 peace agreement are shaped by cues from political elites representing stark political factions, and the direction of this effect is contingent on individual affinity with those political elites. Beyond these results, a preliminary exploration of the mechanisms suggests that cues sometimes can convey a type of information (e.g. expected beneficiaries of a policy); however, such other information is unlikely to substitute for citizens’ reliance on cues from political elites.

Understanding citizen attitude formation during peace processes has important implications. As settlements are more likely to end civil conflict, they increasingly feature citizens involved through a direct vote on settlement referendums or on leaders negotiating them.⁵ Even without voting, citizens still decide whether to cooperate with settlement terms (e.g. Office of the United Nations High Commissioner for Human Rights, 2009). Shedding light on what influences citizens’ attitudes toward settlements thus speaks to a growing literature on inclusion in peace processes (e.g. Nilsson, 2012). Specifically, we show that citizens seem to split along partisan lines and reinforce elite deals or divisions, so an elite deal remains the crucial component of a stable agreement.

Our results also suggest that political elite stances and their messaging to signal their stance on these provisions can substantially influence not only citizen support but also other crucial outcomes. While malleable, and specifically reflecting elite deals or divisions according to our findings, public opinion matters due to the different forms of citizen involvement in the process (ratifying, voting and cooperating). As political elites shape attitudes on provisions, they can alter the prospects of a settlement and its implementation, which, in turn, can change policy and politics in the country going forward (e.g. Dyrstad, Bakke & Binningsbø, 2021).

How do citizens form their attitudes about settlement provisions?

Peace processes are composed of compromises between various actors, and citizens influence their provisions by voting on them directly at times, voting on the leaders who negotiated them at times, and, in all cases, deciding whether to cooperate with policy changes.⁶ But how do citizens form attitudes about the provisions?

Our theory builds on the idea that peace processes match the dynamics of attitude formation identified in other political processes where elite messaging is important,⁷ and contrasts with ideas in existing literature on post-conflict contexts that instead imply that different dynamics are at work. First, some models suggest that citizens’ attitudes may simply be more fixed after fighting. Negative emotions can be heightened, which may make reconciliation difficult and citizens less receptive to even elite cues (Bar-Tal, 2009). Any party allegiances or other normal political dynamics may be ignored if, for instance, some see a status quo stalemate as better than a failure-prone agreement (e.g. Walter, 1997), or, alternatively, if citizens are war-weary and desperate to support any agreement (Zartman, 1995). Second, given the exceptional changes in peace processes, as well as the mixed messages that accompany them, other models suggest that citizens have strong incentives to sort through information on advantages and disadvantages to make fully informed decisions on whether to support each policy. Exiting models thus suggest that forming attitudes about provisions in settlements therefore may be unique decisions as voters assess policies but also reshape their state. However, these arguments provide reason to question whether post-conflict dynamics are different from normal politics. initial work on trends in Colombian public opinion suggests that citizens’ attitudes reflect prominent political leaders’ factions (Matanock & Garca-Sánchez, 2017; Liendo & Braithwaite, 2018). In addition to these conflicting ideas, both the theory and empirical tests are underdeveloped in these contexts. This paper seeks to further develop both.⁸

Elite cues

We first draw on existing studies in political behaviour showing that acquiring information is costly and most citizens are uninformed even on important issues, and that they often rely on heuristics to form their opinions and make decisions (Tversky & Kahneman, 1974; Kuklinski & Quirk, 2000). Underpinning these ideas are theories about when people will use cognitive shortcuts vs. when they will invest more effort in acquiring and processing information (Sniderman, Brody & Tetlock, 1993; Kahneman, 2011). Citizens are likely to use cognitive shortcuts when they are distracted, when they must put cognitive effort elsewhere, and when the decisions are complex (Lupia, 1994).

We contend that the conditions during peace processes warrant the use of cognitive shortcuts, due to the complexity of the issues, as well as the post-conflict context that draws their focus elsewhere. Even as fighting winds down, citizens facing an adverse economic environment centrally focus on survival (Voors et al., 2012). Moreover, the advantages and disadvantages of the policies emerging from peace agreements are often complicated and based on compromise over time between opposing parties. Thus, citizens will likely rely on cognitive shortcuts to simplify decisions throughout these contexts.

An important cognitive shortcut is cues from political elites. These cues can lead citizens to update their information about the policy (Bullock, 2011; Cohen, 2003).⁹ Elites specifically representing political camps (henceforth, ‘political elites’), supply clear cues because their starkly split positions provide salience on the issue and increase information in an existing cognitive context (Levendusky, 2010: 114). In our theory, political factionalization provides for the information environment in which political elites’ cleavages manifest and become starker, and, thus, where we expect their cues to be informative for citizens to form their attitudes.¹⁰

Post-conflict contexts tend to be characterized by factionalization and ideological debate around the issue of war settlement and peace. Political elites in these settings are usually deeply divided as they often emerge from fighting factions (e.g. ethnic polarization; see López-Pintor, 2005), and, as transitions change the balance of power, new divisions also materialize in the post-conflict context on issues such as whether to take dovish or hawkish approaches to negotiations (e.g. Feldmann, 2019). Furthermore, in contexts where settlements are ratified via referendum, this factionalization can become especially acute because these processes tend to generate clear political camps that seek to mobilize citizens for or against the settlement. Political elites who have staked out strong positions during or after the conflict send clear signals through campaigns, public statements, or official endorsements. Therefore, cues from these political elites, defined in this context as politicians who lead opposing camps on an agreement, will be particularly legible for citizens who will then be prone to rely on them to form their attitudes toward peace issues.¹¹

Hypothesis 1: (Political elite cues): Citizens will rely on political elites’ cues to form their opinion about provisions emerging from settlements.

Affinity

We argue that the influence of elite cues depends on feelings about specific political elites, and that in the factionalized context of civil conflict, many citizens will hold strong positions. Carlin, McCoy & Subotic (2019) show that trust in the negotiating parties of the peace agreement in Colombia is highly correlated with support for the settlement and for certain provisions. We expect individual affinity with an elite, stemming from shared beliefs and values, ideological proximity, perceived common interests, or less ‘rational’ reasons such as the elite’s charisma or style, to activate and direct the effect of cues on attitudes (Ortiz-Ayala & Garca-Sánchez, 2014; Beck & Jennings, 1991). Elite cues on the costs and benefits of peace agreement provisions in economic, legal, and even emotional terms are viewed as either positive or negative, perhaps especially acutely when individual predispositions are deeply held (Levendusky, 2010; Druckman, Peterson & Slothuus, 2013).

Hypothesis 2. (Affinity): An endorsement on a provision from a political figure will increase support for that provision when a citizen has affinity for that elite. Conversely, an endorsement will decrease the citizen’s support when a citizen has an aversion to the endorser.

Information

We can also test some aspects of how cuing works. We posit that political elite cues could work as information shortcuts for citizens. Under this logic, cuing could, for instance, change expectations about the provisions’ likely beneficiaries. Citizens could use this heuristic to glean information and make efficient decisions without collecting and sorting all the details.¹²

In our context, this explanation could be at work, as respondents did not know the detail about provisions emerging from the peace process. For instance, in our survey, only 14.4% of citizens could correctly identify the cap on prison sentences for those judged by the Colombian Special Tribunal for Transitional Justice (JEP), a key aspect of these provisions that was a contentious issue among political elites.¹³

Hypothesis 3. (Perceived Beneficiaries): Receiving a political elite cue will shape which groups citizens perceive as potential beneficiaries from the provision.

Related, we argue that if cueing provides knowledge, more factual information potentially offers an alternative mechanism to understand and then evaluate provisions. However, we anticipate that citizens prefer the clearer signals from political elites more than additional information of any kind in these contexts because these cues more easily convey either affect or the policy’s advantages and disadvantages, producing a less effort-intensive path (Cohen, 2003; Druckman, Peterson & Slothuus, 2013).¹⁴ We, therefore, do not state a standalone hypothesis about additional information, but we do evaluate its effect.

The context of the Colombian peace agreement

Colombia offers a unique opportunity to test our theory during the implementation of a hotly debated settlement to end a civil conflict. The role of voters is clear in this case. In 2016, citizens voted on and narrowly rejected the final agreement negotiated to end a war of over 50 years between the FARC and the government. A revised settlement was ratified in Congress later that year, after incorporating a set of the changes suggested by the opposition, and it has been slowly implemented since. Each provision requires approval by Congress. Citizens continue to factor into the process by, for example, voting on the legislators and the government that will be making decisions about implementation.

Despite their role, citizens still have little information about the settlement’s provisions due to its complex nature. The initial accord was nearly 300 pages long and filled with technical language about the provisions. Parts of the peace agreement have been enacted into law while others were effectively opposed and altered or abandoned. Understanding the advantages and disadvantages of specific policies emerging from the process demands considerable cognitive effort.

Colombian political elites on the government side divided into political camps during the negotiating phase (2012–2016) and remain divided around the provisions included in the settlement during implementation. The coalition led by the president who negotiated the agreement, Juan Manuel Santos, has expressed consistent support for the final settlement, while the coalition led by his predecessor, Álvaro Uribe, questions its legitimacy and has systematically sought to hinder implementation. Santos and Uribe have thus been clear representatives of those who support and oppose the settlement, respectively. Observational evidence suggests that when Uribe increased his criticisms against the peace process in 2013, his followers took more radical stances on issues such as a negotiated peace and reconciliation with former combatants (Gaviria, Ávila García & García-Sánchez, 2019).

Citizens still weigh in and ultimately have a say in the stability of the settlement through this extended implementation process. The 2018 elections, for example, reconfigured this process. The political camp opposing the agreement gained the upper hand, and, as the provisions from the peace deal were set to be approved piecemeal, the implementation slowed. For instance, to date, many of the provisions that had to be approved in Congress have been either voted against or not even been included in the agenda (Kroc Institute, 2019).

Variation in support for provisions emerging from the settlement

Before detailing the research design, we review the existing variation in public support for the provisions derived from the agreement between the Colombian government and the FARC. Attitudes vary considerably across these provisions. According to our original survey, representative of the conflict zones, 53% of Colombians expressed support for the overall peace agreement. Meanwhile, support for provisions regarding transitional justice and the political representation of FARC was lower than that of the overall agreement. The only provisions for which support was higher than the overall agreement is on rural development and the special seats for conflict regions (see Figure 1).

Overall, these data highlight that support for the provisions is not uniform and thus reinforce questions of how citizens form their opinions about the often-complex

Figure 1.

Percentage of individuals who support the peace agreement and its specific provisions

policies that emerge from and are implemented after the settlement and what sources of information they use.

Research design

We test our hypotheses with a survey experiment during the ongoing implementation of the Colombian peace process. We randomize political elite cues on two provisions that emerged from the settlement and were not yet finalized during our study of this post-conflict context. The cue randomization provides us with exogenous variation on whether subjects receive an endorsement from an elite with whom they have affinity or not, even though we cannot directly manipulate individual affinity with the endorser. Hence, as we discuss later, we block on this variable for the treatment assignment.

The sample

We embedded the experimental questions in a 2017 survey representative of areas of civil conflict in Colombia. The sample selected adults living in the territories where the government, at the time we gathered the data, was running development plans, known as Development Plans with a Territorial Focus, as part of the implementation of the peace agreement with the FARC.¹⁵ The implementation of the settlement is especially high stakes in these territories.

Fieldwork took place between October and December of 2017, about a year after the peace agreement was signed and while its implementation was being discussed in Congress. By the time of data collection, the race for the presidency was beginning and parties were defining their lists for the congressional 2018 national elections. Our survey suggests that the main concerns of residents in the sampled municipalities were issues related to peace and security and the economy. The sample consisted of 1,391 respondents.¹⁶

As discussed in the theory section, the incentives to become informed about the settlement, including its specific provisions, should increase as citizens perceive direct effects on their lives. These citizens are more likely to be eligible for programmes introduced by peace agreements, potentially increasing their support for new rural development, increased political representation, etc., but they may also have been affected by conflict, potentially shaping their views about programmes such as reconciliation. Due to these interests, citizens in our survey may be the most willing to obtain and sort through different information on the settlement to form their attitudes, although our theory still hypothesizes that even they will continue to use political elite cues instead, given the cue’s legibility and the competing circumstances. Our survey sample, therefore, provides a hard test for our theory.¹⁷

Experimental design

To test our theory, we used a vignette survey experiment that tests the effect of this type of cuing by randomizing elites’ cues that endorse provisions coming out of the peace agreement. In our ideal design, we would have randomized assignment of endorsements of the same provision by different political elites eliciting polarized affections (i.e., Santos and Uribe in the case we study). However, we did not use this design as it would have required deception, since it would have involved saying that one of the elites supported a proposal that it did not support. Instead, we examined the effect of cues using two separate vignettes, a design that allowed us to study these effects without deceiving participants about the actual elites’ positions. This design potentially limits the interpretation of eventual differences identified in the results of these two vignettes, and we probe these differences through additional tests below, but this was the ethical choice given the circumstances.¹⁸

The provisions included in the vignettes

We presented each respondent with two provisions from the settlement that were in the works during our survey, either in the legalization process or just becoming law, and that were on topics with generally mixed support, as just discussed. Former president Santos, who negotiated the original peace agreement, supported one provision; his predecessor, Uribe, supported the other (which revised what had been in the original settlement). Although these two provisions provide within-subject variation in terms of the affinity with the political elites, which we discuss later, we analyse them separately.¹⁹

Jurisdicción Especial para la Paz (JEP): This provision was set to establish the Special Tribunal for Transitional Justice to investigate, prosecute and punish crimes committed in the civil conflict. JEP’s judges would determine whether crime perpetrators can receive amnesty or a reduced sentence, in part depending on their confessions. The JEP’s creation was a central provision in the settlement, with support among negotiators, including Santos, but it was also contentious. The tribunal started operating in March 2018, after our survey, but has faced a series of political hurdles, many of them initiated by Centro Democrático, Uribe’s party.

Circunscripciones Especiales de Paz: This provision was set to create a special district in Congress, composed of 16 seats, reserved for the areas of the country most affected by civil conflict. The peace agreement stated that only indigenous communities, social movements and groups of citizens would be able to compete for these special seats, but Uribe, along with other politicians, proposed instead that any party could compete for these seats. We asked respondents about Uribe’s proposed modification. The revised proposal failed to receive the needed majority in a congressional vote in November 2017, just at the end of our survey. In 2021, the 16 seats became law due to a ruling from the Constitutional Court.

Randomization of treatments

We showed each participant the following vignettes for the two different provisions, and for each, we randomized whether a political elite endorsed the proposal (Santos in the JEP proposal and Uribe in the change to the special seats proposal) or not (in which case, the proposal uses a generic endorsement in the form of ‘some people’). This design thus allows us to avoid deception given the polarized context, as described above:²⁰

[Some people/President Santos] have/has supported the creation of the Special Jurisdiction for Peace (JEP).

[Some people/former president Álvaro Uribe Vélez] have/has supported that the political parties that are already in Congress can also compete for the seats that are reserved for the regions of the country that are most affected by the conflict.

We cross-randomized between provisions, such that the treatment received for the second proposal was independent of what was received for the first one. We also randomized the order in which proposals were presented. Table I summarizes the distribution of our sample of subjects across all experimental conditions.²¹

Table I.

Matrix of treatments for both proposals

Santos’ elite cue treat	Number	Cell %
No cue	478	51.7
Santos’ cue	447	48.3
Total	925	100.0
Uribe’s elite cue treat	Number	Cell %
No cue	439	48.3
Uribe’s cue	470	51.7
Total	909	100.0

Measuring individuals’ affinity with elites

We operationalize individuals’ affinity with the elites representing the political camps in dispute using the following question:

On this card there is a staircase with steps numbered from one to seven, in which 1 is the lowest step and means VERY UNFAVORABLE and 7 is the highest step and means VERY FAVORABLE. If your opinion is between very unfavorable and very favorable choose an intermediate score. I am going to mention the name of various public figures and I would like you to tell me on that scale what is your concept of each one of them.

Based on answers to this question, we classified an individual as having affinity with the elite (Santos or Uribe) if she responded 5, 6, or 7, and non-affinity otherwise.²² Given our focus on the affinity between the respondent and the randomly-assigned endorser, we blocked on the individuals’ pre-treatment affinity with the political elites in our randomization. Thus, we created four blocks defined by respondents’ joint affinity with Santos and Uribe. Based on our sample, 42.42% of citizens in conflict-affected areas have affinity with Santos, and 34.22% have affinity with Uribe (see Table A2 in the Online appendix). We randomly assigned respondents to each of the experimental conditions within each block.

Operationalizing the outcome of interest

Our main hypotheses are related to public support for the specific provisions that emerged from the Colombian settlement. We measure support for each one of the proposals using a 1 to 7 scale (where 1 means ‘No support at all’ and 7 ‘A lot of support’). Our hypothesis about the mechanism is related to the expected beneficiaries of these provisions. So we also ask about the extent to which they think the provision would benefit each of the following actors: (i) the government; (ii) the FARC; and (iii) citizens like them.²³

Analysis

Testing the effects of political elite cues

We can test our main hypotheses by simply comparing individuals’ level of support for each provision across type of endorsers and interacting these treatment indicators with the degree of affinity between the individual and the elite who endorses the proposal. Affinity with the political elites is measured with a dichotomous variable, as described. We therefore estimate the effect of interest with ordinary least squares regression, with robust standard errors, as follows:

\begin{array}{l} y_{i} = & β_{0} + β_{1} P E_{i} + β_{3} A P E_{i} + β_{4} P E_{i} \times A P E_{i} + ε_{i}, \end{array}

where y_i corresponds to support for the provision under consideration; $P E_{i}$ is the treatment indicator corresponding to a political elite endorsement; and $A P E_{i}$ to our pre-treatment measure of affinity to either political elite.²⁴

Results

Elite cues and affinity (H1 and H2)

Testing first for the role of political elite cues alone,²⁵ we find an effect of elite endorsements on citizens’ opinions

Figure 2.

The effect of political elite cues conditioned on affinity with the endorser

about the provisions emerging from the settlement, but it is only statistically significant for one cue. Figure 2 presents the results of the estimation of Equation (1). The first column of Tables IIa and IIb depict the results of the estimation of the average treatment effect of elite endorsements. We see that Santos’ cue does not have a statistically significant effect on support for the creation of the JEP in the sample overall. In contrast, Uribe’s cue has a positive and statistically significant effect on support for existing parties to run for the Special Seats.

However, we expected the effect of elite cues to depend on individual affinity with political leaders,²⁶ and our results support this expectation. The second column of Tables IIa and IIb and Figure 2 show that the direction and magnitude of the effect of both Santos’ and Uribe’s cues are contingent on individual affinity with these elites.

The second column of Table IIa indicates the effect of Santos’ cue conditional on individual affinity: Among those with an affinity for Santos, receiving his cue increases support for the provision by 0.15 points (although this effect is not statistically significant), but, among those without an affinity for Santos, it decreases support by 0.27 points (and this effect is statistically significant at p<0.1). This difference in effect size is statistically significant (p<0.1).

The second column of Table IIb indicates the effect of Uribe’s cue: a positive, statistically significant effect for those with affinity (0.951, p-value <0.001) – in fact, its magnitude is substantial, increasing support for the policy by almost one point on a 1–7 scale – and a negative but, as expected, not statistically significant effect for those without affinity. In other words, we find even stronger, and more nuanced, evidence for the role of political elite cues once we factor in individual affinity with elites.

Information (H3)

To explore the possibility that elite cues work as information shortcuts for citizens – and the mechanism more broadly – we test whether elite cues shape citizens’ perceptions of the provisions’ beneficiaries. When asked to what extent they believed each provision would benefit different actors, we find mixed support for this hypothesis (Figure 3).

We observe that receiving the elite endorsement, in general, does not modify citizens’ expectations about the presumed beneficiaries of the provision. There are two exceptions when Uribe’s cue seems to convey information: among those who like him, receiving his cue

Table II.

The effect of elite cues on average support for specific policies from the peace agreement

(a) Support for the Colombian Special Tribunal for Transitional Justice
	(1)	(2)
	Pure	Affinity
Santos’ cue	-0.0600	-0.271^†
	(0.125)	(0.164)
Affinity w/Santos		1.210***
		(0.159)
Santos × affinity		0.425^†
		(0.226)
Constant	4.817***	4.316***
	(0.0859)	(0.115)
Observations	891	891
(b) Support for parties in Congress running for Special Seats
	(1)	(2)
	Pure	Affinity
Uribe’s cue	0.307*	-0.0381
	(0.131)	(0.162)
Affinity w/Uribe		0.165
		(0.196)
Uribe × affinity		0.951***
		(0.268)
Constant	3.580***	3.522***
	(0.0946)	(0.119)
Observations	860	860

Robust standard errors in parentheses. ^† $p < 0.1$ , * $p < 0.05$ , ** $p < 0.01$ , *** $p < 0.001$ .

made respondents more likely to perceive that the proposal will benefit the government (

\hat{β} = 0.522

, p-value = 0.011) and citizens like them (

\hat{β} = 0.52

, p-value = 0.021).

The evidence from Uribe’s cue lends support to the explanation that individuals learn information from his endorsement, changing their perceptions about beneficiaries. Moreover, that Uribe’s cue is only informative among those who like him is consistent with models of cheap talk where communication is only informative if the sender’s (elite) and receiver’s (citizens) ideal points are not too distant (Crawford & Sobel, 1982).²⁷

However, the differences in the results between Santos’ cue and Uribe’s cue make it less likely that the mechanism is that their endorsements are conveying information to citizens (especially since the data in Figure A3 in the Online appendix show that both Santos and Uribe are political elites who could each theoretically drive a similar effect).²⁸ These elites, and the cues, though, may not be providing the same amount of information as we discuss in the next section.

To further explore the extent to which additional information shapes the process of attitude formation, we cross-randomized the endorsement experiment with the receipt of additional information about the provisions, resulting, for each vignette, in a $2 \times 2$ factorial design. Prior to receiving the elite cues, we randomly presented an authoritative statement that described the purpose of the proposal in more detail.²⁹ In Figure 4, we compare the effect of the cue on those that do and do not receive additional information. We estimate whether the difference in effects is statistically significant using a difference-in-differences estimator (formal tests are include in Appendix E). As expected, additional information attenuates the effect of the cue, but the differences in effects are generally not statistically significant (see column 1 of Tables A5 and A6 in Online appendix E). These null effects hold regardless of respondents’ affinity with the elite. Our results, therefore, do not show strong support for an information conveyance mechanism but do offer some evidence consistent with it.³⁰

Extensions and robustness checks

An important set of extended tests emerged from the differences we found between the effects of the cues from the two political elites. While Santos’ cue reduces support among those who dislike him, his out-group, Uribe’s cue increases support among those who like him, his in-group (which is in line with Haas & Khadka, 2020).

Figure 3.

The effect of elite cues on the perceived beneficiary of the provision

Figure 4.

The joint effects of elite cues and information about the provisions under consideration

So, we sought to explore briefly what might explain these differential effects. Our design restricts our ability to fully answer this question, as it does not allow for a causal explanation of the divergent results; however, in this section, we discuss some exploratory explanations for the differences.

One strong explanation is that Santos has a more diverse ‘camp’ that therefore does not take his cues as readily. During the years of negotiations with the FARC, public opinion and electoral support gravitated around two factions representing divergent stances on the issue of peace (Botero, García-Sánchez & Wills-Otero, 2018). Supporters of the peace process conformed a coalition from the left to the centre of the ideological spectrum that claimed affinity with Santos because they match on this particular issue even though it was otherwise diverse on other issues; the other camp was a more clustered group on the right that claimed affinity with Uribe

Figure 5.

Distribution of support for different provisions by level of elites’ favourability rating and treatment status

because he represented a true ideological alliance on various issues (Rivera, Plata-Caviedes & Rodríguez-Raga, 2019). Uribe supporters can therefore more readily use his views to form opinions, knowing that they stand near him on many issues.

Two different pieces of evidence corroborate this interpretation. First, focusing on the control group, average support for the JEP increases along with Santos’ favourability rating, while support for the revision of the special seats is independent of Uribe’s favourability (see Figure 5). Support for the JEP is highly correlated with support for the peace agreement overall, so this suggests the expected correlation between the peace process and affinity with Santos. Second, we re-ran the analyses from Table II including a battery of covariates that may be correlated with elite affinity (see Table A16 in the Online appendix).³¹ While the effects of Uribe’s cue conditional on affinity hold, Santos’ cue’s negative effect among those who dislike him is no longer statistically significant. Thus, Santos seems to be bringing together a diverse group of citizens in favour of the agreement, which, may make his cues less potent for his camp.

Another explanation of the results is that Santos’ endorsement was less surprising than Uribe’s (Nicholson, 2011). As the main promoter of the peace negotiations, Santos was likely to support the provisions. In contrast, Uribe’s endorsement of a peace agreement policy – albeit conditioned on modification of what was originally signed in Havana – was more unexpected and therefore persuasive. While we cannot directly test this explanation, and, given the context it likely has some merit, some evidence does counter it.

To assess different levels of prior knowledge about the provisions in the two vignettes, we compared the overall support for the proposal and the percentage of respondents answering ‘Don’t know’ or not answering in the control group without information and the control group with additional information. We do not find statistically significant differences in either case, thus differences in underlying knowledge are unlikely to be behind the divergent results (see Tables A17 and A18 in Online appendix K). However, this test does not get at the specific question of whether the Uribe endorsement is more surprising, so that is a question for future work.

We also ran several robustness checks on our results, including some already noted in the endnotes. For example, to further explore when citizens rely on cues, we provided endorsements from professors, elites considered more technical – defined as those whose perceived source of authority does not derive from the exercise of power or public office – but not polarized in the Colombian context. Yet, we found that they have little effect on support for the provisions (see Online appendix H). Moreover, we ruled out the possibility that respondents react to any ‘expert cue’ effect by using the technical elites’ experimental condition as a control group for political elites (see Online appendix I).³² Again, the results for Uribe’s cue are robust to this specification, while the effect of Santos’ cue loses statistical significance. We also examined the robustness of our results with respect to different definitions of our affinity measure to ensure that the results are not driven by a particular way of operationalizing this variable,³³ and we find that they hold across different specifications (see Online appendix J). Finally, data from our sample show that 16.4% have an affinity for both leaders (see Table A2 in the Online appendix), and so we checked that our results are not driven by this special group; our results hold even excluding this group (see Table A13 in the Online appendix that replicates the estimations of Table II).

Conclusions

We analyse attitude formation toward the complex policies that compose peace agreements to end civil war, examining the case of Colombia. We find that cues from political elites have strong effects on individuals’ support for specific provisions that emerge from the Colombian settlement, and that these effects are conditioned on individual affinity with elites. These findings support our expectations that peace processes are politics as usual. We also find some evidence that these cues may be providing information about the beneficiaries of these policies to the respondents, but additional information about the provisions does not similarly shape public opinion.

Our work contributes to a growing body of experimental literature on public opinion about peace processes – although it focuses on the implementation rather than the negotiation period – and, on the whole, it suggests that attitude formation even in post-conflict contexts does not work differently from other political processes, unlike most of the existing post-conflict models, perhaps because of the high stakes of conflict (see Walter, 1997). Our findings imply that citizens use cognitive shortcuts even when evaluating these important policies. Specifically, they take cues from political elites who represent polarized factions, depending on their affinity for that elites’ camp.

By showing that citizens split along partisan lines or political camps as implementation takes place, we suggest that a broad elite deal remains the crucial component of a stable agreement even when citizens weigh in. Citizens strongly react to cues from political elites, so their input is not likely to counterweight elite divisions, and, instead, stems from it. Even when these policies are likely to benefit citizens, they may still reject the provisions.

A growing number of countries with civil conflict hold either direct referendums on peace agreements or elections of officials negotiating them. Thus, our results likely travel to other countries that, like Colombia, try to resolve a civil conflict via settlement, and, at the same time, have a competitive political system. In these cases, the possibility of a division of political elites along partisan lines, even if they represent the political establishment, may prove detrimental to peace.

Our work points to some additional research needed on the role of information. In one of our robustness checks, we also examine elites viewed as more technical, and we do not find that they shape attitudes. The theory also suggests that these political elites are most legible. Another recent study shows that religious leaders can shape citizen attitudes towards reintegration (Blair et al., 2021). Together our research suggests that factionalization may matter a good deal for cue uptake in these contexts, but future work could further explore this by using cues from more leaders. In addition, given the mixed findings presented on the mechanism, we are cautious with the interpretation of how elite cues influence attitude formation. Nevertheless, both these results and the null results of providing information examined in the extensions help point the path forward on this topic. They indicate the importance of more exploration on the mechanism of attitude formation, especially the role of information in cue-taking, and they suggest that citizens may rely on types of information that they can easily absorb to the exclusion of other information.

In addition, considering that the implementation of a settlement takes several years, a key question that emerges from this analysis and merits further study is the extent to which the effect of elite cues changes over time. We think that elite cues remain a strong influence on public opinions on peace as long as the peace agreement remains a salient issue and key public figures continue debating around it. The Colombian experience suggests that such conditions hold for a relatively short period of time. In 2016 conflict/peace issues were the main concern of Colombians, but, by 2020 such issues dropped fourth place. Additionally, recent evidence indicates that the opinion gap regarding peace issues, between political camps, has been closing, most likely because political elites are debating less about these topics (Ávila García et al., 2023). Future work that uses similar experiments later in the process, and in other contexts, will therefore be useful to understanding these politics that can prove divisive at least in this moment in Colombia.

Footnotes

Acknowledgements

This project was approved by University of California, Berkeley’s Institutional Review Board protocol #2017-02-9601 and by Universidad de los Andes’ Committee on Research Ethics protocol #791-2017. A pre-analysis plan for the survey experiment was registered in Evidence in Governance and Politics #20171104AB. We thank Dan Corstange, Alexander de Juan, Caroline Hartzell, Chad Hazlett, Gabe Lenz, Katerina Linos, Enzo Nussio, Francesca Parente, Alison Post, Julia Raven, Jake Shapiro, Laura Stoker, Carlos Ávila, and participants at American Political Science Association 2018, EGAP 22, a 2018 Folke Bernadotte Academy workshop, Peace Science Society workshop, Paris School of Economics Conflict Workshop, Brown, Columbia, and New York University’s speakers’ series, and Berkeley’s MIRTH and VIO Lab for excellent comments and suggestions. All errors or omissions are our own. We are grateful to USAID (Colombia) for their generous financial support to conduct the survey in which we embedded our experiment.

Replication data

The dataset, codebook, and do-files for the empirical analysis in this article can be found at .

ORCID iDs

Natalia Garbiras-Díaz

Miguel García-Sánchez

Aila M. Matanock

Notes

References

Ahler

Douglas J

(2014) Self-fulfilling misperceptions of public polarization. Journal of Politics 76(3): 607–620.

Arceneaux

Kevin

Kolodny

Robin

(2009) Educating the least informed: Group endorsements in a grassroots campaign. American Journal of Political Science 53(4): 755–770.

Assouline

Philippe

Trager

Robert F

(2017) Concessions for Concessions Sake: Injustice, Indignation and the Construction of Intractable Conflict in Israel–Palestine. Los Angeles, CA: UCLA.

Assouline

Philippe

Trager

Robert F

(2021) Concessions for concession’s sake: Injustice, indignation, and the construction of intractable conflict in Israel -Palestine. Journal of Conflict Resolution 65(9): 1489–1520. https://journals.sagepub.com/doi/abs/10.1177/0022002721997543

Ávila García

Carlos

Plata

Juan Camilo

García-Sánchez

Miguel

Rodríguez-Raga

Juan Carlos

(2023) Opinión pública y acuerdo de paz, 2016–2020 [Public opinion and peace agreement, 2016–2020]. In: Rettberg

Angelika

Betancur

Laura

(eds) Después del Acuerdo:¿Cómo va la paz en Colombia?[After the Agreement: How is peace going in Colombia?] Bogotá, Colombia: Ediciones Uniandes. [In Spanish.]

Barabas

Jason

Jerit

Jennifer

(2010) Are survey experiments externally valid? American Political Science Review 104(2): 226–242.

Bar-Tal

Daniel

(2009) Reconciliation as a foundation of culture of peace. In: Joseph

Rivera

(ed.) Handbook on Building Cultures of Peace. Berlin: Springer, 363–377.

Beck

Paul Allen

Kent Jennings

(1991) Family traditions, political periods, and the development of partisan orientations. Journal of Politics 53(3): 742–763.

Blair

Graeme

Littman

Rebecca

Nugent

Elizabeth R

Wolfe

Rebecca

Bukar

Mohammed

Crisman

Benjamin

Etim

Anthony

Hazlett

Chad

Kim

Jiyoung

(2021) Trusted authorities can change minds and shift norms during conflict. Proceedings of the National Academy of Sciences 118(42): e2105570118.

10.

Botero

Felipe

García-Sánchez

Miguel

Wills-Otero

Laura

(2018) Polarización y posconflicto: Las elecciones nacionales y locales en Colombia, 2014–2017 [Polarization and post-conflict: National and local elections in Colombia, 2014–2017] (1 ed.). Colombia: Universidad de los Andes. [In Spanish.]

11.

Brader

Ted

Sio

Lorenzo De

Paparo

Aldo

Tucker

Joshua A

(2020) “Where you lead, I will follow:” Partisan cueing on high-salience issues in a turbulent multiparty system. Political Psychology 41(4): 795–812.

12.

Branton

Demeritt

Pulido

Amalia

Meernik (2019) Violence, voting and peace: Explaining public support for the peace referendum in Colombia. Electoral Studies 61(2): 102067.

13.

Broockman

David E

Butler

Daniel M

(2017) The causal effects of elite position-taking on voter attitudes: Field experiments with elite communication. American Journal of Political Science 61(1): 208–221.

14.

Bullock

John G

(2011) Elite influence on public opinion in an informed electorate. American Political Science Review 105(3): 496–515.

15.

Carlin

Ryan E

McCoy

Jennifer L

Subotic

Jelena

(2019) Leading the public to peace: Trust in elites, the legitimacy of negotiated peace, and support for transitional justice. In: James Meernik, Jacqueline HR DeMeritt, & Mauricio Uribe-Loepez (eds) As War Ends: What Colombia Can Tell us About the Sustainability of Peace and Transitional Justice. Cambridge: Cambridge University Press, 282–303.

16.

Cohen

Geoffrey L

(2003) Party over policy: The dominating impact of group influence on political beliefs. Journal of Personality and Social Psychology 85(5): 808–822.

17.

Collier

Paul

Elliott

Hegre

Håvard

Hoeffler

Anke

Reynal-Querol

Marta

Sambanis

Nicholas

(2003) Breaking the Conflict Trap: Civil War and Development Policy. Washington, DC: World Bank.

18.

Collin

Katherine

(2020) Peacemaking referendums: The use of direct democracy in peace processes. Democratization 27(5): 717–736.

19.

Crawford

Vincent P

Sobel

Joel

(1982) Strategic information transmission. Econometrica: Journal of the Econometric Society 50(6): 1431–1451.

20.

Druckman

James N

Leeper

Thomas J

(2012) Learning more from political communication experiments: Pretreatment and its effects. American Journal of Political Science 56(4): 875–896.

21.

Druckman

James N

Peterson

Erik

Slothuus

Rune

(2013) How elite partisan polarization affects public opinion formation. American Political Science Review 107(1): 57–79.

22.

Dyrstad

Karin

Bakke

Kristin M

Binningsbø

Helga M

(2021) Perceptions of peace agreements and political trust in post-war Guatemala, Nepal, and Northern Ireland. International Peacekeeping 28(4): 606–631.

23.

Fabbe

Kristin

Hazlett

Chad

Snmazdemir

Tolga

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

24.

Fearon

James D

(2011) Governance and Civil War Onset. Washington, DC: World Bank.

25.

Fearon

James D

Laitin

David D

(2003) Ethnicity, insurgency, and civil war. American political Science Review 97(1): 75–90.

26.

Feldmann

Andreas E.

(2019) Colombia’s polarizing peace efforts. In: Press

B I

(ed.) Democracies Divided: The Global Challenge of Political Polarization. Washington, DC: Brookings Institution Press.

27.

Finkel

Eli J

Bail

Christopher A

Cikara

Mina

Ditto

Peter H

Iyengar

Shanto

Klar

Samara

Mason

Lilliana

McGrath

Mary C

Nyhan

Brendan

Rand

David G

Skitka

Linda J

Tucker

Joshua A

Bavel

Jay J Van

Wangand

Cynthia S

Druckman

James N

(2020) Political sectarianism in America. Science 370(6516): 533–536.

28.

Gaviria

Adriana

García

Carlos Ávila

García-Sánchez

Miguel

(2019) Barómetro de las Américas Colombia 2018. Paz, posconflicto y reconciliación [Americas Barometer Colombia 2018. Peace, post-conflict and reconciliation] . Uniandes, USAID: Observatorio de la Democracia.

29.

Haas

Nicholas

Khadka

Prabin

(2020) If they endorse it, I can’t trust it: How out-group leader endorsements undercut public support for civil war peace settlements. American Journal of Political Science 64(4): 982–1000.

30.

Hartzell

Caroline A

Hoddie

Matthew

(2007) Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars. Pennsylvania: Penn State Press.

31.

Hazlett

Chad

Parente

Francesca

(2020) https://www.researchgate.net/publication/330008672_Who_supports_peace_with_the_FARC_A_sensitivity-based_approach_under_imperfect_identification Working paper .

32.

Herzog

Stephen

Baron

Jonathon

Gibbons

Rebecca Davis

(2022) Antinormative messaging, group cues, and the nuclear ban treaty. Journal of Politics 84(1): 591–596.

33.

Iyengar

Shanto

Lelkes

Yphtach

Levendusky

Matthew

Malhotra

Neil

Westwood

Sean J

(2019) The origins and consequences of affective polarization in the United States. Annual Review of Political Science 22(1): 129–146.

34.

Jarland

Julie

Nygård

Håvard Mokleiv

Gates

Scott

Hermansen

Emilie

Larsen

Vilde Bergstad

(2020) https://www.prio.org/publications/12303 Oslo: Conflict Trends.

35.

Kahneman

Daniel

(2011) Thinking, Fast and Slow volume 1. New York: Farrar, Straus and Giroux New York.

36.

Kalyvas

Stathis N

(2006) The Logic of Violence in Civil War. Cambridge: Cambridge University Press.

37.

Kao

Kristen

Revkin

Mara Redlich

(2018) To punish or to pardon?: Reintegrating rebel collaborators after conflict in Iraq. Reintegrating Rebel Collaborators After Conflict in Iraq (November 1, 2018). The Program on Governance and Local Development Working Paper Series at the University of Gothenburg (17) (https://gld.gu.se/media/1503/gld-working-paper-17-final.pdf).

38.

Kreiman

Guillermo

Masullo

Juan

(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.

39.

Kroc Institute (2019) Estado efectivo de implementación del acuerdo de paz de colombia 2 años de implementación [Effective status of implementation of the Colombian peace agreement 2 years of implementation]. Technical report. Kroc Institute for International Peace Studies (https://kroc.nd.edu/assets/321729/190523_informe_3_final_final.pdf). [In Spanish.]

40.

Kuklinski

James H

Quirk

Paul J

(2000) Reconsidering the rational public: Cognition, heuristics, and mass opinion. In: Lupia

Arthur

McCubbins

Mathew D

Popkin

Samuel L

(eds) Elements of Reason: Cognition, Choice, and the Bounds of Rationality. Cambridge: Cambridge University Press, 153–182.

41.

Lenz

Gabriel S

(2012) Follow the Leader. How Voters Respond to Politicians’ Policies and Performance. Chicago, IL: University of Chicago Press.

42.

Levendusky

Matthew S

(2010) Clearer cues, more consistent voters: A benefit of elite polarization. Political Behavior 32(1): 111–131.

43.

Liendo

Nicolás

Braithwaite

Jessica Maves

(2018) Determinants of Colombian attitudes toward the peace process. Conflict Management and Peace Science 35(6): 622–636.

44.

López-Pintor

Rafael

(2005) Postconflict elections and democratization: An experience review. USAID, Issue Paper (8) (https://pdf.usaid.gov/pdf_docs/PNADB897.pdf).

45.

Lupia

Arthur

(1994) Shortcuts versus encyclopedias: Information and voting behavior in California insurance reform elections. American Political Science Review 88:(1): 63–76.

46.

Masullo

Morisi

(2019) Risks and opportunities of direct democracy: The effect of information in Colombia’s peace referendum. Politics and Governance 7(2): 242–267.

47.

Matanock

Aila M

García-Sánchez

Miguel

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

48.

Matanock

Aila M

Garbiras-Díaz

Natalia

(2018) Considering concessions: A survey experiment on the Colombian peace process. Conflict Processes and Peace Science 35(5): 637–655.

49.

Mondak

Jeffery J

(1993) Public opinion and heuristic processing of source cues. Political Behavior 15(2): 167–192.

50.

Nicholson

Stephen P

(2011) Dominating cues and the limits of elite influence. Journal of Politics 73(4): 1165–1177.

51.

Nilsson

Desirée

(2012) Anchoring the peace: Civil society actors in peace accords and durable peace. International Interactions 38(2): 243–266.

52.

Office of the United Nations High Commissioner for Human Rights (2009) Rule-of-law tools for post-conflict states: National consultations on transitional justice. Technical report. Office of the United Nations High Commissioner for Human Rights (https://www.ohchr.org/sites/default/files/Documents/Publications/NationalConsultationsTJ_EN.pdf).

53.

Ortiz-Ayala

Alejandra

García-Sánchez

Miguel

(2014) Porque te quiero te apoyo: Estilo de gobierno y aprobación presidencial en américa latina [Because I love you, I support you: Government style and presidential approval in Latin America]. Revista de Ciencia Política (Santiago) 34(2): 373–398. [In Spanish.]

54.

Pettersson

Therése

Högbladh

Stina

Öberg

Magnus

(2019) Organized violence, 1989–2018 and peace agreements. Journal of Peace Research 56(4): 589–603.

55.

Przeworski

Adam

(2005) Democracy as an equilibrium. Public Choice 123(3-4): 253–273.

56.

Rivera

Alejandra

Plata-Caviedes

Juan Camilo

Rodríguez-Raga

Juan Carlos

(2019) Barómetro de las Américas Colombia 2018. Democracia e instituciones [Americas Barometer Colombia 2018. Democracy and Institutions] . Uniandes, USAID: Observatorio de la Democracia. [In Spanish.]

57.

Slovic

Paul

Finucane

Melissa L

Peters

Ellen

MacGregor

Donald G

(2007) The affect heuristic. European Journal of Operational Research 177(3): 1333–1352.

58.

Sniderman

Paul M

Brody

Richard A

Tetlock

Phillip E

(1993) Reasoning and Choice: Explorations in Political Psychology. Cambridge: Cambridge University Press.

59.

Tellez

Juan Fernando

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

60.

Tversky

Amos

Kahneman

Daniel

(1974) Judgment under uncertainty: Heuristics and biases. Science 185(4157): 1124–1131.

61.

Voors

Maarten J

Nillesen

Eleonora EM

Verwimp

Philip

Bulte

Erwin H

Lensink

Robert

Soest

Daan P Van

(2012) Violent conflict and behavior: A field experiment in Burundi. American Economic Review 102(2): 941–64.

62.

Walter

Barbara F

(1997) The critical barrier to civil war settlement. International Organization 51(3): 335–364.

63.

Wantchekon

Leonard

(2004) The paradox of “warlord” democracy: A theoretical investigation. American Political Science Review 98(1): 17–33.

64.

Weingast

Barry R

(1997) The political foundations of democracy and the rule of the law. American political science review 91(2): 245–263.

65.

Weintraub

Michael

Vargas

Juan F

Flores

Thomas E

(2015) Vote choice and legacies of violence: Evidence from the 2014 Colombian presidential elections. Research & Politics 2(2): 2053168015573348.

66.

Wood

Elisabeth Jean

(2003) Insurgent Collective Action and Civil War in El Salvador. Cambridge: Cambridge University Press.

67.

Zaller

John

(1992) The Nature and Origins of Mass Opinion. Cambridge: Cambridge university press.

68.

Zartman

I William

(1995) Elusive Peace: Negotiating an End to Civil Wars. Washington, D.C: Brookings Institution Press.

Political elite cues and attitude formation in post-conflict contexts

Abstract

Keywords

Introduction

How do citizens form their attitudes about settlement provisions?

Elite cues

Affinity

Information

The context of the Colombian peace agreement

Variation in support for provisions emerging from the settlement

Research design

The sample

Experimental design

The provisions included in the vignettes

Randomization of treatments

Measuring individuals’ affinity with elites

Operationalizing the outcome of interest

Analysis

Testing the effects of political elite cues

Results

Elite cues and affinity (H1 and H2)

Information (H3)

Extensions and robustness checks

Conclusions

Footnotes

Acknowledgements

Replication data

ORCID iDs

Notes

References