Wartime Experiences and Popular Support for Peace Agreements: Comparative Evidence from Three Cases

Abstract

Peace agreements are negotiated and signed by representatives of the government and the rebels, often after many years of violent conflict, but their ability to transform a war-torn society hinges on the approval of ordinary people. Yet we have little systematic knowledge of what ordinary people think of peace agreements in the long run. This study begins to fill that gap, drawing on a set of comparative public opinion surveys from Guatemala, Nepal, and Northern Ireland, three cases where long civil wars were ended by peace agreements. The peace agreements in these countries have strong popular support, though there is variation across specific provisions. Across these cases, our findings suggest that legacies of violence are not generally associated with long-term support for peace agreements. However, when we look at provisions that grant concessions to the rebels, there is some evidence of lasting legacies.

Keywords

peace agreement civil wars conflict conflict resolution

Introduction

This study examines public support for peace agreements in three post-conflict contexts, 10 to 20 years after the agreements were signed. Though peace agreements are often ambitious documents offering prospects of fundamental restructuring of society through specific reforms, they rarely represent ideal solutions or roadmaps for future development—and many break down (e.g., Bell and Badanjak 2019) or are not implemented in full (e.g., Walter 2002; Derouen, Lea, and Wallensteen 2009; Joshi, Quinn, and Regan 2015). Peace agreements are negotiated and signed by political elites—typically the result of lengthy bargaining processes between warring parties who will not sign unless they believe they have more to gain from collaborating than from continued fighting—sometimes with civil society involvement (e.g., Nilsson 2012). The ability of these elite-level agreements to transform a war-torn society hinges also on what ordinary people think of the agreement (c.f. Nilsson and González Marín 2020)—ordinary people who may have endured years of violence. We argue that wartime experiences have long-term implications for what people think about peace agreements. Indeed, peace agreements represent pivotal historical moments, bringing hope about a peaceful future, but they often remain contentious long after they are signed (c.f. Morgan-Jones, Stefanovic, and Loizides 2021).

A growing number of studies investigate what people think about peace processes and peace agreements, but most examine the immediate post-signing period (e.g., Hayes and McAllister 2001; Ringdal, Simkus, and Listhaug 2007; Dyrstad et al. 2011) or people’s perceptions in the midst of peace negotiations (e.g., Tellez 2019; Liendo and Braithwaite 2018; Matanock and Garbiras-Díaz 2018, Loizides et al. in this Special Feature).¹ While people in conflict-affected societies may show support for efforts to end violence when these efforts are fresh, particularly if they have experienced violence, our study asks if conflict legacies also drive people’s perceptions of peace agreements in the long term? This research question situates us within the lower right quadrant of this Special Feature’s conceptual framework (see Introduction in this Special Feature), in which peace processes have long-lasting effects on citizens’ perceptions and behaviors. We pay special attention to provisions that give concessions to the opposition—the ‘rebels’. Such concessions may be necessary, even crucial, to end the conflict, but may be particularly controversial as they may be seen as rewarding violence. To the degree that peace agreements often entail a shift in power relations in favor of the rebels, we expect support for provisions granting them concessions to follow wartime cleavages. That is, we expect that the ‘side’ on which people participated, or the ‘side’ by which they were victimized by, will color their views on the peace agreement.

Employing a most-different case approach, we examine public support for peace agreements in three post-conflict contexts—Guatemala, Nepal, and Northern Ireland—10 to 20 years after the agreements were signed. The comparative Post-Conflict Attitudes for Peace (PAP) survey was conducted between January and July 2016. Our analysis reveals that overall, the peace agreements in these countries enjoy strong popular support even many years after they were signed, though there is significant variation across provisions. We find that experiences of wartime violence are not generally associated with support (or lack thereof) for peace agreements in the long run, but wartime cleavages, in the sense of which ‘side’ people fought on or were victimized by, shape their views on concessions given to the rebels. The most consistent finding is that compared to non-victims, victims of government-sponsored violence are more in favor of such concessions. Overall, our analysis shows substantial variation from one context to the next, which serves as a warning against drawing general conclusions from single-case studies.

Argument: The Long-Term Legacies of Wartime Violence

A long-standing body of research has shown that experiences of violence shape people’s political behavior and attitudes even long after a conflict ends (e.g., Dyrstad et al. 2011; Brounéus 2014; Gilligan, Pasquale, and Samii 2014; Nussio and Oppenheim 2014; Hall et al. 2018; Hong and Kang 2017; De Juan and Pierskalla 2016)—even across generations (e.g., Balcells 2012; Lupu and Peisakhin 2017; Rozenas and Zhukov 2019; Costalli and Ruggeri 2019; Villamil 2020).² Warfare is likely to generate both psychological and material grievances: trauma or disability from one’s own experiences of violence (either as a victim or perpetrator or both), the loss and victimization of loved ones, and loss of one’s home and income (e.g., Justino 2009; Gates et al. 2012). Though there is growing debate about the conditions under which experiences of violence may have positive effects on certain postwar political behaviors, such as participation in politics and social capital (e.g., Blattman 2009; Bellows and Miguel 2009; Voors et al. 2012; Bauer et al. 2016), research (and practice) on conflict reconciliation has long highlighted the detrimental legacies of violence on people’s identities and attitudes, particularly perceptions of the ‘outgroup’ (e.g., Bar-Tal 2000; Kelman 2008). We argue that experiences of violence are likely to have long-term implications for people’s perceptions of the very initiatives aimed at bringing the conflict to an end (and preventing its recurrence). This is due both to the nature of peace agreements and the lasting legacies of violence.

Peace agreements represent pivotal historical moments, bringing hopes for a peaceful future, and they remain salient long after they are signed. They are typically the result of drawn-out bargaining processes in which the warring parties attempt to ‘correct’ the ‘wrongs’ of both the pre-conflict period and conflict itself, laying out a blueprint for the future. They promise peace, reconciliation, and economic reconstruction to people affected by years of violence, suspicion, and neglect. Peace agreements are, however, also the result of political compromises, with the possibility of dissatisfaction both among those party to the negotiations and those left out, which can create spoiler dynamics that impede implementation (e.g., Stedman 1997; Greenhill and Major 2007; Pearlman 2009). Indeed, many peace agreements are not implemented—or implemented slowly or only partially. Consider the slow implementation of the peace agreement in Nepal, of which Human Rights Watch noted, 10 years after the agreement was signed: “[T]he promises of accountability for abuses and the resolution of thousands of disappearances have been broken by Nepal’s main political parties, all of which have taken turns at leading the government in the last decade” (Human Rights Watch 2016). According to the Peace Accord Matrix (Joshi, Quinn, and Regan 2015), which compares the implementation of individual provisions in 34 comprehensive peace agreements signed between 1989 and 2012, the average implementation record 10 years after signing is 49 percent. Evidently, peace agreements may be contentious political issues—certainly salient—long after they are signed.

Peace agreements are likely to be particularly salient for individuals who experienced violence firsthand, as such experiences leave enduring legacies. Indeed, both wartime victimhood and participation are likely to shape how individuals view peace agreements in the long term. First, although victims of violence may be skeptical of political authorities and, by extension, elite-level bargains,³ and may find it difficult to forgive perpetrators of violence (Bakke, O’Loughlin, and Ward 2009), victims are also acutely aware of the high costs of violence (Tellez 2019)—and, indeed, are likely to experience war-related stress long after the fighting ends (Ringdal, Ringdal, and Simkus 2008). Hence victims of violence may be more likely than the general population to favor any agreement that puts an end to the war, even years after the violence ceased—and perhaps particularly in hindsight, when they have experienced peace.

H1a

Compared to non-victims, victims of wartime violence are more likely to display support of the peace agreement.

Second, just like victims of the war may be relieved it is over, so may armed participants, who bore the costs of combat. As such, they may be likely to hold a positive view of any agreement that brought an end to the fighting, also as time passes.⁴ Alternatively, participants may have entrenched interests in continued warfare, as they enjoy benefits such as status in their community (c.f. Jankowski 1991; Kostelnik and Skarbek 2013 on gangs) and access to wartime selective incentives (c.f. Lichbach 1994), or combat may have hardened their attitudes towards their rivals and to compromise (Grossman, Manekin, and Miodownik 2015). Even if that is the case, former participants not only bore the costs of combat but may also be also members of or affiliated with the very organizations that negotiated and signed the peace agreement(s), and as such, they may be more likely to approve of the agreement than the overall population who, generally, have less of a say in designing the agreement. To the degree that the elites signing the agreement support it and former participants follow their leaders in the calculus that their ‘side’ had more to gain from a peace agreement than from continued fighting (c.f. Greenhill and Major 2007)—which may be a tall assumption—we would expect armed participants to be supportive of the agreement. Thus, while recognizing that the literature suggests expectations that go in different directions, we propose the following hypothesis:

H1b

Compared to non-participants, former participants in armed conflict are more likely to display support of the peace agreement.

The political compromises in peace agreements often involve granting concessions to the opposition—the ‘rebels’. At an aggregate level, research suggests that such concessions are crucial in ensuring an end to fighting and preventing that armed conflict resumes (e.g., Derouen, Lea, and Wallensteen 2009; Hartzell and Hoddie 2007; Jarstad and Nilsson 2008; Ottmann and Vüllers 2015). These concessions represent a shift in power relations between the government and the rebels, by which the government loses some authority and the rebel gains some, for example through power sharing (Svensson 2007). By conceding to some of the rebels’ claims, peace agreements redress grievances that presumably fueled the conflict in the first place and address the opposition’s concerns about their post-war access to power—and grant them a certain level of legitimacy. In Nepal, for example, the Maoist party became the largest party in the first postwar election (Pokharel and Rana 2013). In essence, the civil war and ensuing peace agreement in Nepal allowed the Maoist party to move from a marginal to a large political party. In Northern Ireland, the Good Friday Agreement gave representatives of both ‘sides’ a seat at the negotiation table and entailed reforms that sought to redress the ‘Nationalist’ side’s sense of long-term marginalization and discrimination of the Catholic community (e.g., Wolff 2001), though there were elements on both sides who rejected the agreement—‘Nationalists’ who thought it did not go far enough towards creating a united Ireland and ‘Unionists’ thinking it granted too many concessions to the Nationalists. In Guatemala, in comparison, it is clear that the peace agreement did little to improve the political influence of the insurgent groups, nor the people they claimed to represent, the indigenous rural poor, as URNG, the political party established by the rebels, has experienced limited electoral success.⁵

While wartime experiences may make individuals supportive of any peace agreement that ends violence—as H1a and H1b above state—they may view the specific provisions in these agreements differently, pending on their wartime ‘side’.6 Granting concessions to rebels especially, may have a different standing among civilians who sympathized with their cause compared to those who were harmed by their violence. As such concessions are often crucial for ending conflict and maintaining peace, it is particularly important to understand how people evaluate them.

Recent studies from Colombia reach diverging conclusions. Matanock and Garbiras-Díaz (2018), relying on an endorsement survey experiment, find that Colombians in general were skeptical towards provisions seen as granting concessions to the FARC—especially if they viewed the FARC negatively—even if they supported the peace process overall. In contrast, Tellez (2019) finds that people in conflict-affected areas in the country were more likely to support both the peace process and the granting of concessions to the FARC. Tellez (2019), like Kreiman and Masullo (2020), argues that people are ‘safety-seekers’ who support peace initiatives and provisions they consider to be most suited to ensure their current and future security.⁷ While these studies from Colombia examine people’s perceptions either when the armed conflict was still ongoing or immediately after the 2016 peace agreement, we do not know if people weigh these issues in the same way when the armed conflict has turned into a more distant memory.

As noted earlier, research on the long-term legacies of political violence show that these experiences influence people’s opinions long after the violence ended, even across generations (e.g., Lupu and Peisakhin 2017; Rozenas and Zhukov 2019). Examining the long-term effects of the Spanish civil war, Balcells (2012) finds that the wartime victimization of an individual’s family members led the individual to reject the perpetrator’s political identity in terms of present-day political cleavages. Hong and Kang (2017) find that violence committed by South Korean forces and their allies during the 1950–53 Korean War made people skeptical towards the South Korean government and military more than 50 years later.

These studies do not examine concessions to rebels, but they speak to our assumption that who or which ‘side’ committed the violence an individual experienced during conflict—even a long time ago—shapes how s/he views these actors today. With respect to people’s views on concessions given to the rebels, we believe that while fear of renewed violence may make people willing to accept such concessions to ensuring safety in the immediate post-conflict period, this fear diminishes over time such that deep-rooted grudges towards perpetrators will dominate in forming people’s opinions.

In line with this, we expect that the identity of the perceived perpetrators is important for understanding how experiences of wartime violence affects people’s perceptions of specific peace agreement provisions, particularly provisions that are seen as clear concessions to the rebels. Victims of violence at the hands of the rebels may be skeptical of peace agreement provisions that in principle grant concessions to their (or their family’s) perpetrator(s). Following the same side-taking dynamics, victims of government violence may be more inclined to support provisions that reduce the wartime government’s formal power by granting concessions to rebels. Therefore, we propose the following hypotheses:

H2a

Compared to non-victims, individuals who were victimized by rebels are less likely to favor provisions that give concessions to the rebels.

H2b

Compared to non-victims, individuals who were victimized by the government are more likely to favor provisions that give concessions to rebels.

Similarly, we expect support for provisions that give concessions to rebels to be colored not only by the experience of former participation in general but the ‘side’ on which people participated. We expect former participants on the rebel side to be largely in favor of provisions giving concessions to the rebels, while former participants on the government side, including pro-government militias, are more likely to be opposed to such provisions, which are seen to empower the rebels.

H2c

Compared to non-participants, individuals who participated on the rebel side are more likely to favor provisions that give concessions to rebels.

H2d

Compared to non-participants, individuals who participated on the government side are less likely to favor provisions that give concessions to rebels.

We test these propositions about the long-term legacies of violence across three post-conflict cases where relative peace has prevailed for several years after the signing of a peace agreement: Guatemala, Nepal, and Northern Ireland.

Case Selection

Our case selection follows a most-different case approach (e.g., Seawright and Gerring 2008). The logic of this design implies that if a similar pattern is found across the three cases, despite differing on a range of characteristics, this pattern should not be contingent on any of these characteristics, and the findings may apply to other cases within the same population. The subsections below describe the population and the criteria for selecting cases within this population and provide some details about the three cases.

Population and Selection Criteria

When designing the study, we defined our population as post-conflict electoral democracies that have experienced an internal armed conflict ending through a comprehensive peace agreement and with no resumption of violence. As peace agreements are more common in the post-Cold War period (e.g., Kreutz 2010), we focus on agreements signed after 1990.⁸

To maximize variation, we selected cases from Asia, Europe, and Latin America (geography), including cases of different levels of economic development. They also differ on key conflict characteristics (type and duration), and on types of peacebuilding strategies implemented after the violence ended (see Table 1 and Online Supplemental Material A).

Table 1.

Key case characteristics at the time the survey was conducted (2016).

Dimension	Characteristic	Guatemala	Nepal	Northern Ireland
Conflict	Ethnic conflict	Yes	No	Yes
	Duration of conflict (# of years)	36	10	28
	Year of peace agreement	1996	2006	1998
Peace-building strategies	Inclusiveness: Power sharing	No	Yes	Yes
	Inclusiveness: Type of elections	PR	Mixed	PR (STV)
	Inclusiveness: Rebel integration in army	No	Yes	No
	TJ: Amnesty	No	No	No*
	TJ: TRC report	1999	In process	No**
	TJ: Tribunal	No	No	No
	Security sector: Army reduction	Yes	No	Yes
	Security sector: Police reform	Yes	No	Yes

Notes: PR = proportional representation; STV = single transferable vote; TJ = transitional justice; TRC = truth and reconciliation commission; * a release of political prisoners; ** a commission on a specific incident, Bloody Sunday (30 January 1972), has released a report, but no full TRC report has been called for. Sources: Database of Political Institutions (Beck et al. 2001); UCDP Peace Agreement Dataset (Harbom et al. 2006); US Institute of Peace.

Hence, similar findings across the three cases should not be contingent on any of these characteristics.

For a more detailed discussion of the conflicts, see Dyrstad et al. (2021), but we note here that in all three cases, group inequalities played an important role in fueling the conflict in the first place. In Guatemala and Nepal, the rebels enjoyed support from marginalized groups, including rural indigenous people, lower castes, and ethnic minorities—and, indeed, Guatemala’s rural indigenous population was heavily targeted (Commission for Historical Clarification 1999). In the territorial conflict of Northern Ireland, the dominant cleavage was between those preferring to remain within the United Kingdom (on the political side known as ‘Unionists’, while the paramilitary groups tend to be referred to as ‘Loyalist’), predominantly from the Protestant community, and those preferring to join Ireland (on the political side known as ‘Nationalists’, while the paramilitary groups are known as ‘Republicans’), predominantly from the Catholic community, though central to the conflict was economic discrimination of Catholics.

The Peace Agreements

The Guatemalan Agreement of Firm and Durable Peace (1996) ended 36 years of internal armed conflict between a coalition of left-wing guerrilla groups (the Guatemalan National Revolutionary Unity, URNG) and the state, which led to more than 200,000 civilian deaths. The state was responsible for most of the violence, which has been characterized as genocidal (Commission for Historical Clarification 1999). The peace agreement—which consists of 11 different agreements—was ambitious and included the formal recognition of indigenous people and their rights, the promise of a land reform, and several other measures to reduce inequalities and expand political and economic rights (Caumartin and Sánchez-Ancochea 2011). However, the implementation has been slow, and economic and political power remains in the hands of a small elite.

The Comprehensive Peace Agreement (2006) in Nepal was signed after 10 years of intense conflict between Maoist insurgents and the state, a conflict that claimed more than 13,000 lives. A key aspect of the agreement was to hold elections for a Constituent Assembly to draft a new constitution. The agreement also abolished the monarchy, established that Nepal would be a federal state, and called for socioeconomic restructuring, improved minority rights, and the end of gender and caste discrimination.

In Northern Ireland, the Good Friday Agreement (1998) put an end to 30 years of armed conflict. While protracted, the conflict was less intense than in Guatemala and Nepal, with about 3700 people killed. The agreement restored a devolved Northern Ireland legislative assembly within the United Kingdom and included provisions for executive power sharing between parties representing the Protestant and Catholic communities—though the implementation of power sharing has been hampered by political deadlock. A highly contentious issue in the agreement was policing, with the Catholic/‘Nationalist’ community considering the Royal Ulster Constabulary to be a symbol of a sectarian state, partial to the Protestants/‘Unionists’.

Data and Methods

We designed the PAP survey to map and compare citizens’ perceptions of peace agreements, post-war developments, and experiences of violence. The questionnaire is unique for this project, but to facilitate comparative research, we drew on previous post-conflict surveys (in particular, Humphreys and Weinstein 2004, Simkus 2007, and O’Loughlin, Kolossov, and Toal 2014). Table 2 summarizes key aspects of the data collection. As we relied on local practices and recommendations for conducting surveys, the sampling procedure varied from case to case. A closer description of the sampling and the questionnaire can be found in the Online Supplemental Material B.

Table 2.

Key characteristics of the PAP survey.

What	Guatemala	Nepal	Northern Ireland
Number of respondents	1216	1200	813
Time of survey	Jan 2016	March-April 2016	May-July 2016
Representativeness	Nationally	Nationally	Nationally
Sampling frame	2015 electoral roll	2011 census	Postcode address file
Sampling design	1. PSUs^a = 120 segments within 99 (of 340) municipalities 2. Households drawn at random 3. Respondents selected based on ‘last birthday’ rule	1. PSUs = 60 wards from 45 (of 77) districts 2. Households drawn at random 3. Respondents selected with kish grid	1. Households drawn at random from national sampling frame 2. Respondents selected based on ‘first birthday’ rule
Mode	Face-to-face, tablet-assisted	Face-to-face, smartphone-assisted	Face-to-face, tablet-assisted

^aPSU: Primary sampling unit. In Nepal, the survey was conducted before the federal state was fully implemented and new administrative units were established.

We note that during the fieldwork, current events made the armed conflicts, which in each case ended more than a decade ago, salient in different ways. In Guatemala, just before the survey was launched, 14 former military officers were arrested, accused of crimes against humanity. Nepal experienced continued contention related to the implementation of a federal, secular state, while in Northern Ireland, the Brexit referendum—with its potential implications for the Northern Ireland border—took place during the fieldwork.

All models are estimated using ordinary least square regression (OLS), with standard errors adjusted to account for the stratified sampling in Guatemala and Nepal.⁹

Dependent Variables

The outcomes of interest in this paper are (1) overall support for the peace agreement (corresponding to H1a-b) and (2) support for provisions giving concessions to former rebels (H2a-d).

Table 3 below shows descriptive statistics for all the peace agreement provisions included in the survey, as well as composite measures of support for each agreement (1 = strong disapproval; 5 = strong approval).¹⁰ Provisions that grant concessions to former rebels are highlighted in grey in the table. We use a conservative measure for concessions to the rebels, examining only provisions that clearly grant concessions to armed actors. In Guatemala, we measure concession to a rebel group by a question asking people about their views on the transformation of the former rebel coalition URNG into a political party. In Nepal, we rely on two questions about views on the political and military integration of the Maoist rebels. In Northern Ireland, we use a question about the early release of paramilitary prisoners.¹¹ For the complete wording of the items, see Online Supplemental Material C.

Table 3.

Support for the main provisions in peace agreements in Guatemala, Nepal, and Northern Ireland.

Peace agreement/provision	Mean	St. dev	N
Guatemala: The Agreement of Firm and Durable Peace (1996)	3.739	0.721	962
URNG to political party	2.936	1.451	1115
Recognition of indigenous people	4.328	1.022	1147
Establishment of TRC	3.981	1.165	1104
Land reform	3.774	1.222	1048
Civilian control over army	3.894	1.184	1104
Security sector reform	2.55	1.35	1101
Police reform	2.3	1.25	1137
Nepal: Comprehensive Peace Agreement (2006)	3.691	0.622	787
Abolishment of monarchy	3.003	1.526	1198
New constitution	4.137	0.828	1135
Military integration	3.707	1.128	1185
Establishment of TRC	4.219	0.621	804
Power sharing government	3.908	0.941	1175
Federal state	2.963	1.414	1132
Northern Ireland: The Good Friday Agreement (1998)	3.961	0.552	713
Devolved assembly	4.059	0.864	767
Power sharing executive	3.986	0.907	772
Decommissioning of arms	4.431	0.772	784
Troop reductions	4.239	0.867	778
Police reform	4.042	0.982	771
Early release of prisoners	2.621	1.275	759
Reducing unemployment differences	4.263	0.806	777
Remain part of UK	4.1	0.968	771
Overall	3.788	0.654	2462

Note: All answer categories follow a scale from 1 to 5, where 1 indicates a strong disapproval of the provision, and 5 indicates strong approval/support. Provisions that grant concessions to former rebels in grey.

Independent Variables

A set of dichotomous variables capture wartime experiences, honing in on both participation and victimization. To test H1a, we measure victimization as a dichotomous variable that takes the value of 1 if a respondent reports to have experienced any type of war-related violence.¹² About 41 percent (1328 respondents) report to have experienced one or several types of war-related violence. Similarly, to test H1b, participation equals 1 for respondents who reports to have participated in the armed conflict; otherwise 0. About six percent (208 respondents) reported to have participated.

To test H2a-d, we disaggregate participation and victimization based on the ‘sides’ in the conflict. Wartime victimization is measured based on whether self-reported victims recognized the perpetrators of the acts of violence they had experienced. The answers were grouped into four mutually exclusive groups: non-victims; victims of government-sponsored violence—including pro-government militias or paramilitary groups; victims of rebel violence; and ‘others’, i.e. victims of violence from both sides, other actors including family members, criminal groups, or others, and cases in which the identity of the perpetrator(s) is unknown.¹³ Our measure of wartime participation distinguishes between participation on the government or pro-government paramilitary side;¹⁴ the rebel side; or both sides, unknown or other side(s). As we do not know the exact nature of participation, we refer to the groups as, for example, ‘participants on the rebel side’ rather than ‘rebels’. Ideally, we would have used a more disaggregate measure, for instance distinguishing between participants on the government side or in pro-government-militias, but this would have yielded categories with very few observations.

For descriptive statistics of variables measuring wartime experiences, see Online Supplemental Material D.

We control for sociodemographic characteristics, including age (measured in years), gender, poverty, and level of education.¹⁵ In the pooled sample, we also control for case, using a set of dummy variables with Guatemala as the reference category. This ensures that findings are driven by within-country variation and not due to an uneven distribution of individual-level characteristics.

Finally, as wartime experiences may overlap with other, pre-existing grievances and group identities that contribute to the conflict in the first place (e.g., Horowitz 1985; Gurr and Robert 2000; Stewart 2002; Østby 2008; Cederman, Gleditsch, and Buhaug 2013) and may be associated with different views on the peace agreement, we control for group-based grievances and salient group identities. To control for group-based grievances, we rely on a dichotomous variable of self-reported discrimination due to one’s belonging to an identity group (such as ethnicity, caste, language, or religion). We also use case-specific measures of group identity, reflecting the most salient identity markers in each context: indigenous identity in Guatemala, caste in Nepal, and community background in Northern Ireland.¹⁶

Analysis

The empirical analysis is structured as follows: The first part describes levels of support for peace agreements overall, across cases, and across provisions, paying particular attention to the provisions that give concessions to the rebels. In the second part we test our hypotheses, overall and by country.

Descriptive Statistics: Support for Peace Agreements Across Cases

Despite the elite-led nature of the creation of most peace agreements, the PAP survey documents strong popular support for the agreements in Guatemala, Nepal, and Northern Ireland (Figure 1). The mean score of about 3.8 translates into “approving” of all the peace agreement provisions included in the survey.¹⁷ Overall, the Good Friday Agreement in Northern Ireland enjoys stronger support than the other agreements. While the difference is statistically significant, it is small. There is greater diversity of opinion in Guatemala than in the other cases. The survey questions about the provisions are not directly comparable, as they were designed to reflect the key content of three different peace agreements, and conclusions should be drawn with some caution. In sum, however, our surveys indicate that the peace agreements following these three protracted conflict contexts enjoy widespread support. Provided that our most-different case approach gives some inferential leverage beyond the three cases examined here, peace agreements may enjoy substantial support among individuals in post-conflict societies.

Figure 1.

Distribution of overall support for the peace agreements, by country.

Breaking down overall support into support for specific key provisions, Figure 2 shows that while most provisions are popular, there is substantial variation. In Guatemala and Northern Ireland, the provisions giving concessions to the former rebel groups—the transformation of URNG into a political party and the early release of paramilitary prisoners—are the most controversial. Indeed, the early release of paramilitary prisoners remains the single most controversial of all the provisions across the three cases, which is striking given that the Good Friday Agreement overall is somewhat more popular than the agreements in Guatemala and Nepal. In contrast, in Nepal, provisions giving concessions to the Maoists through political and military integration seem to be more broadly accepted. Here, the institutional reform that the country has been going through remains both the most popular (new constitution) and the most disputed element (federal structure; new regime) of the agreement. In the other two cases, institutional reform and social redress seem less disputed.

Figure 2.

Support for PA provisions, by country.

A possible explanation for the difference in people’s views on concessions in Guatemala and Northern Ireland on one hand and Nepal on the other, is the relative longevity of peace in the former two cases. As we suggest above, when the memory of the armed conflict becomes more distant and the fear of renewed violence diminishes, people’s assessments of the peace agreement may be more based on principles and moral judgements, and less based on fear and safety-seeking (c.f. Tellez 2019; Kreiman and Masullo 2020).

In the next section, we turn to statistical analyses to test our hypotheses, starting with overall support for the peace agreements before turning to the provisions that gave concessions to the rebel group(s).

Legacies of Violence and Popular Support for Peace Agreements

According to H1, war experiences—whether as a victim (H1a) or a former participant (H1b) —are associated with stronger support for peace agreements.

Figure 3 shows estimated marginal effects from the statistical models (with and without control variables) for the pooled sample and the country subsamples. The full statistical models are reported in Online Supplemental Material E. According to Figure 3, only in Nepal are former participants consistently more in favor of the peace agreement than non-participants. Victims of wartime violence seem more in favor of the peace agreement in Northern Ireland, but this association is not robust to the inclusion of the control variables.

Figure 3.

Estimated marginal effects of OLS regression of wartime experiences on support for peace agreements, pooled sample and by country.

Taken together, these results indicate that wartime experience per se is not a key determinant of attitudes to peace agreements. This probably reflects that as discussed in H1 above, there may be competing mechanisms working in opposite directions. While, as we would expect, war-weariness and safety-seeking may lead to endorsement of the peace agreement—or, indeed, any agreement on the table—it is also plausible that those with experiences of violence have low political trust that leads them to reject the peace agreement, and former combatants may come out of the war with hardened attitudes to compromise. We also recognize that we cannot rule out the possibility that initial differences between the most war-affected individuals and the overall population may have been higher at the time the agreements were signed.

Turning to the control variables (Online Supplemental Material Table E), sociodemographic characteristics are only weakly associated with assessment of the peace agreement, and the associations vary from one context to another. We note that, while group-based cleavages played a role in all three conflicts, perceived group-based discrimination is not associated with support for peace agreements in any of the cases.

These findings are robust to several alternative model specifications. Because a few questions about specific provisions have a low N, and we use listwise deletion for the measure of overall support,¹⁸ the number of valid observations for the dependent variable is quite low. Replacing our dependent variable with one without listwise deletion (but controlling for missingness on one or more items using a dummy variable) yields similar results (Online Supplemental Material H1), but provides some additional evidence in favor of H1b in the case of Guatemala. The findings also hold when replacing perceived group-based discrimination with a measure of membership in a marginalized group (see Online Supplemental Material H2).

Legacies of Violence and Support for Concessions to Rebels

The second part of our analysis is dedicated to investigating the correlates of support for provisions that provide concessions to former rebels. In Guatemala, this is the provision allowing the URNG to transform into a political party; in Nepal, the provisions about political and military integration of the Maoists; and in Northern Ireland, the provision about the early release of paramilitary prisoners. Results for victimization and participation disaggregated by ‘sides’ in the conflicts are shown in Figures 4a-b, while the full analyses are reported in the Online Supplemental Material F.

Figure 4.

(a) Estimated marginal effects of OLS regression of victimization by different groups of perpetrators on support for concessions, with and without control variables. (b) Estimated marginal effects of OLS regression of participation in different sides on support for concessions, with and without control variables.

According to our hypotheses, support for these provisions follow the main wartime cleavages. We expect individuals victimized by the rebel side to be less in favor of provisions that give concessions to the rebels (H2a) and individuals victimized by the government side to be more in favor (H2b). The findings vary across provisions and cases (Figure 4a). In Guatemala, victims who identify their perpetrator(s) as someone from the insurgent side are not significantly more opposed to allowing URNG to form a political party, contradicting H2a. However, individuals who identify their perpetrator(s) as someone from the government side, including pro-government militias (PGM), are more in favor of concessions to the insurgents in the full model, suggesting some evidence in favor of H2b.

In Northern Ireland, the response pattern is in line with both H2a and H2b, with victims of government violence more in favor, and victims of rebel violence more opposed to the early release of paramilitary prisoners. The substantial effects are quite strong, with a predicted difference between a victim of Republican violence and government/Loyalist violence of about 0.8 on a variable ranging from 1 to 5. In Nepal, victims of government violence are significantly more in favor of integrating Maoist fighters into the government army, but not of the political integration through a power sharing government, so the evidence remains inconclusive. Victims of rebel violence, in contrast, are not more opposed to concessions to their former perpetrators.

Turning to the legacy of wartime participation, we expected that, compared to non-participants, former participants on the rebel side were more in favor (H2c) and former participants on the government side, including pro-government militias, were less in favor (H2d) of concessions to rebels. Figure 4b provides some evidence for these hypotheses, but again, the results are not consistent across provisions nor cases. In Guatemala, both H2c and H2d are rejected. Contrary to expectations in H2d, in Nepal, former participants on the government side are not less supportive of concessions to their former enemies, but former rebels are significantly more in favor of political power sharing, so H2c finds some support. Finally, turning to Northern Ireland, the pattern of support for the early release of paramilitary prisoners is in line with both H2c and H2d, i.e., former participants on the government side are significantly less in favor of this provision, while former participants on the rebel side are significantly more supportive of the early release, and the substantive effect is strong (about 1.6 on a five-point scale).

In sum, the most consistent pattern of wartime cleavages cemented into enduring postwar identities is found in Northern Ireland. This is where the ‘master narrative’ of the conflict (Kalyvas 2003) most closely followed pre-existing social identities, as it pitted the Catholic and Protestant communities against each other. It is worth noting that these findings are robust to the inclusion of community background, i.e., these patterns come in addition to the main cleavage between Protestants and Catholics.

Finally, turning to the sociodemographic control variables, reported in Supplemental Material F, the substantial effects of these variables are weak and inconsistent, adding little explanatory power to the models.

Summary

Taken together, our analysis provides mixed evidence. Table 4 provides a summary of hypotheses and findings. Across our cases, wartime experiences are not generally associated with support for peace agreements in the long run (H1a and b). The partial exception here is Nepal, where, at the time of the survey, political turmoil about the implementation of the peace agreement was prevalent (International Crisis Group 2016), and where the more recent conflict experience may, among former combatants, have heightened risk aversion and safety-seeking preferences.

Table 4.

Summary of hypotheses and findings.

Hypothesis	Overall	Guatemala	Nepal	N. Ireland
H1a. Victims $\overset{+}{\to}$ PA	Rejected	Rejected	Rejected	Rejected
H1a. Participants $\overset{+}{\to}$ PA	Rejected	Rejected	Supported	Rejected
H2a. Victims of insurgents $\bar{\to}$ concessions	—	Rejected	Mixed	Supported
H2b. Victims of gov’t $\overset{+}{\to}$ concessions	—	Supported	Mixed	Supported
H2c. Insurgent participants $\overset{+}{\to}$ concessions	—	Rejected	Mixed	Supported
H2d. Pro-gov’t participant $\bar{\to}$ concessions	—	Rejected	Rejected	Supported

We observe more evidence of legacies of the war when we look at support for provisions that grant concessions to the rebels and disaggregate wartime experiences into the different ‘sides’ (H2a-d). In Northern Ireland, we find a clear pattern of side-taking dynamics: while victims of government violence and former participants on the rebel side display stronger support of concessions to the rebel side, victims of rebel violence and former participants on the government are more strongly opposed to such concessions. Note that the fault lines in the conflict in Northern Ireland largely followed pre-existing cleavages, and community background remains an important predictor of support for the peace agreement. Yet, our analysis shows that wartime experiences, both in the shape of participation and victimization, seem to create additional cleavages that help explain post-conflict attitudes long after the conflict ended.

In contrast, the findings in Nepal are mixed, depending on type of experience but also type of concession—military or political integration of the former Maoists. In Guatemala, all but one hypothesis are rejected, the exception being that victims of government violence are more in favor of concessions to the rebels.

Our findings are robust to the inclusion of case-specific identity markers, such as indigenous language (Guatemala), caste (Nepal), and community background (Northern Ireland). These are identity markers that overlap with the side-taking dynamics during the conflicts (see Online Supplemental Material H3-H4) and, thus, goes some way towards trying to single out the role that experiences of violence play. That said, an alternative explanation is that victims were different than non-victims prior to their victimization and were targeted because of this difference, i.e., they were targeted because they had a different ideology or identity than their perpetrator, which signaled (presumed) closer ties to the ‘other side’—thus victims of government violence may have leaned positively towards (concessions to) the rebel group even prior to victimization. We cannot empirically rule this out in our analysis. We do note, though, that, as argued by Kalyvas (2003, 487)—and with examples from both Guatemala and Northern Ireland—the ‘master narrative’ of a conflict is often created after a war has ended to “simplify, streamline, and cover up the war’s ambiguities and contractions.” The targets of violence are not always a result of this master narrative.

To shed light on these highly divergent findings, we estimated the same models reported in Figure 4a-b on overall support for the peace agreements as well as the support for all the specific provisions we have information about (shown in Figure 2 above). The results are reported in Online Supplemental Material G. The findings are mixed, and caution is warranted against drawing strong conclusions. Yet, our interpretation of these findings is that side-taking dynamics due to wartime experiences seem to be associated with enduring ideological positions that follow the conflicting sides. For example, in line with the ‘official’ positions of the conflicting parties in Nepal, former participants on the insurgent side are more in favor of institutional reform, including the abolishment of monarchy and the establishment of a federal state, while former participants on the government side express significantly lower support for the provision about a new federal structure.¹⁹ Similarly, insurgent participants in Guatemala are more in favor of the recognition of indigenous people and their rights. In Northern Ireland, victims of government-sponsored violence are more likely to favor of troop reductions and measures against unemployment differences.

Conclusion

The 2016 referendum in Colombia, in which people rejected the proposed peace agreement with a very small margin, illustrates that popular support for peace agreements should not be taken for granted. As described in this article, people’s wartime experiences of violence may be associated with their approval of the means to end the war even long after a peace agreement is signed. If peace negotiators fail to consider these experiences among ordinary people, elite compromises may have a hard time enabling enduring peace. In Colombia, the negotiators had to return to the table, searching for more acceptable bargains. The seemingly successful peace agreements in Guatemala, Nepal, and Northern Ireland have had 10–20 years to gain wider acceptance, and initial skepticism may have been much higher. The relationship between popular support for peace agreements and durable peace has yet to be properly analyzed by the larger research community. It is possible that agreements are popular when they succeed in ending conflict. However, one might also posit that agreements are able to secure durable peace precisely when and if a large majority of the population supports them.

While these issues remain speculations, our study shows that ‘mature’ peace agreements enjoy widespread popular support across three very different cases, indicating that this may also be the case for other peace agreements. This overall high level of support may help explain why none of the three conflicts has seen major spoiler problems, though none of these post-conflict societies are without postwar challenges—high criminal violence in Guatemala, political protests and turbulence in Nepal, and paramilitary groups exercising social control in certain areas of Northern Ireland. Nonetheless, these are ‘successful’ post-war societies in the sense that the war has not resumed. As Greenhill and Major (2007) note, the most significant determinant of spoiler behavior is the expected utility of continued fighting. Facing an apparently acceptable peace agreement, the incentives to continue or resume a costly war across all three of our cases may be low. As the Colombian referendum showed, popular support for peace agreements delimits the available space for compromise among negotiating elites: a durable peace agreement must give sufficient concessions to eliminate spoiler problems, yet still be palatable to the broader populations (see also Introduction to this Special Feature section).

That said, the overall high support for the peace agreement in our three cases conceals substantial variation among different provisions. In two of our cases of more ‘mature’ peace, Guatemala and Northern Ireland, we observe that provisions giving concessions to former rebel groups remain controversial. In particular, the early release of paramilitary prisoners in Northern Ireland remains the single most controversial provision of all the 21 provisions included in the PAP survey. In Nepal, in contrast, provisions giving concessions to the Maoists rebels are relatively more popular.

We began this paper with an expectation that wartime experience would lead to greater overall support for peace agreements. However, we find little evidence that war exposure—either as a victim or participant—by itself is associated with different views of the peace agreements overall. What we do find is that when we look at specific provisions and disaggregate victimization and participation into the different wartime ‘sides’, in some cases and under certain circumstances, wartime experiences appear to leave long-lasting legacies. Specifically, looking at peace agreement provisions that grant concessions to rebel groups reveals that wartime cleavages, in the sense of which ‘side’ people fought on or were victimized by, can contribute to shaping people’s views on concessions given to the rebels and on the peace agreements overall. However, our findings can also be read as a warning against drawing sweeping conclusions from single case studies. As our diverging results illustrate, findings from one context may not always travel far.

More generally, our findings contribute to the emerging research agenda on perceptions of peace agreements (e.g., Tellez 2019; Matanock and Garbiras-Díaz 2018; Kreiman and Masullo 2020), adding a long-term perspective. In so doing, this study draws on and adds support to studies that have shown that wartime cleavages may have long-term implications for political preferences in the post-war era (e.g., Balcells 2012). Bracketed in our study—and a fruitful avenue for further research—is how past violence may be used by present-day political elites (c.f. Villamil 2020). Indeed, a follow-up to our study would be to delve into each of these cases to examine whether and how, over time, wartime victimization and participation have been memorialized and politicized and with what effect on popular perceptions, including views on peace agreements.

Supplemental Material

Supplemental Material - Wartime Experiences and Popular Support for Peace Agreements: Comparative Evidence from Three Cases

Supplemental Material for Wartime Experiences and Popular Support for Peace Agreements: Comparative Evidence from Three Cases by Karin Dyrstad, Helga M. Binningsbø, Kristin M. Bakke in Journal of Conflict Resolution

Footnotes

Acknowledgments

We would also like to thank the Editor and two anonymous reviewers for their valuable suggestions. Complete replication files can be found along with the online supplemental material on the JCR SAGE webpage.

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Norges Forskningsråd (240446).

ORCID iDs

Karin Dyrstad

Helga M. Binningsbø

Kristin M. Bakke

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Bakke

Kristin

O’Loughlin

John

Ward

Michael

2009. “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

Laia

. 2012. “The Consequences of Victimization on Political Identities: Evidence from Spain.” Politics & Society 40 (3): 311-347. DOI: 10.1177/0032329211424721.

Bar-Tal

Daniel

. 2000. “From Intractable Conflict Through Conflict Resolution To Reconciliation: Psychological Analysis.” Political Psychology 21 (2): 351-365. DOI: 10.1111/0162-895X.00192.

. 2016. “Can War Foster Cooperation?” Journal of Economic Perspectives 30 (3): 249-274. DOI: 10.1257/jep.30.3.249.

. 2001. New Tools in Comparative Political Economy: The Database of Political Institutions. The World Bank Economic Review 15(1): 165–176. DOI: 10.1093/wber/15.1.165.

. 2019. “Introducing PA-X: A New Peace Agreement Database and Dataset.” Journal of Peace Research 56 (3): 452-466. DOI: 10.1177/0022343318819123.

. 2009. “War and Local Collective Action in Sierra Leone.” Journal of Public Economics 93 (11–12): 1144-1157. DOI: 10.1016/j.jpubeco.2009.07.012.

Blattman

Christopher

. 2009. “From Violence to Voting: War and Political Participation in Uganda.” American Political Science Review 103 (02): 231–247. DOI: 10.1017/S0003055409090212.

Brounéus

Karen

. 2014. “The Women and Peace Hypothesis in Peacebuilding Settings: Attitudes of Women in the Wake of the Rwandan Genocide.” Signs 40 (1): 125-151.

. 2017. “Exposure to Violence, Ethos of Conflict, and Support for Compromise: Surveys in Israel, East Jerusalem, West Bank, and Gaza.” Journal of Conflict Resolution 61 (1): 84-113. DOI: 10.1177/0022002715569771.

. 2011. “Explaining a Contradictory Record: The Case of Guatemala.” In Horizontal Inequalities and Post-Conflict Development, edited by Langer

. London: Palgrave Macmillan.

. 2013. Inequality, Grievances and Civil War. New York, NY: Cambridge University Press.

13.

Commission for Historical Clarification . 1999. “Guatemala: Memory of Silence,” 13–87. Guatemala City, Guatemala: Commission for Historical Clarification.

. 2019. “The Long-Term Electoral Legacies of Civil War in Young Democracies: Italy, 1946-1968.” Comparative Political Studies 52 (6): 927-961. DOI: 10.1177/0010414018784057.

. 2012. Shirts today, skins tomorrow: Dual contests and the effects of fragmentation in self-determination disputes. Journal of Conflict Resolution, 56(1): 67-93. DOI: 10.1177/0022002711429697.

. 2016. “Civil War Violence and Political Trust: Microlevel Evidence from Nepal.” Conflict Management and Peace Science 33 (1): 67–88. DOI: 10.1177/0738894214544612.

. 2009. “The Duration of Civil War Peace Agreements.” Conflict Management and Peace Science 26 (4): 367-387. DOI: 10.1177/0738894209106481.

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

. 2011. “Microfoundations of Civil Conflict Reconciliation: Ethnicity and Context.” International Interactions 37 (February 4): 363-387. DOI: 10.1080/03050629.2011.622627

. 2019. “A Persuasive Peace: Syrian Refugees’ Attitudes towards Compromise and Civil War Termination.” Journal of Peace Research 56 (1): 103-117. DOI: 10.1177/0022343318814114.

Håvard Strand . 2012. “Development Consequences of Armed Conflict.” World Development 40 (9): 1713-1722. DOI: 10.1016/j.worlddev.2012.04.031.

. 2014. “Civil War and Social Cohesion: Lab-in-the-Field Evidence from Nepal.” American Journal of Political Science 58 (3): 604-619.

. 2007. “The Perils of Profiling. Civil War Spoilers and the Collapse of Intrastate Peace Accords” International Security 31 (3): 7-40.

24.

Grodsky

Brian

. 2009. “International Prosecutions and Domestic Politics: The Use of Truth Commissions as Compromise Justice in Serbia and Croatia.” International Studies Review 11 (4): 687-706.

25.

Grosjean

. 2014. Conflict and Social and Political Preferences: Evidence from World War II and Civil Conflicts in 35 European Countries. Comparative Economic Studies 56(3): 424-451. DOI: 10.1057/ces.2014.2.

. 2015. “The Political Legacies of Combat: Attitudes Toward War and Peace Among Israeli Ex-Combatants.” International Organization 69 (4): 981-1009. DOI: 10.1017/S002081831500020X.

27.

Gurr Robert

Ted

. 2000. Peoples versus States: Minorities at Risk in the New Century. Washington, DC: United States Institute of Peace Press.

Djordje Stefanovic Loizides

Neophytos

. 2018. “Exposure to Violence and Attitudes Towards Transitional Justice.” Political Psychology 39 (2): 345-363. DOI: 10.1111/pops.12412.

. 2006. Armed conflict and peace agreements. Journal of Peace Research 43(5), 617–631. DOI: 10.1177/0022343306067613.

. 2007. Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars. University Park: Pennsylvania State University Press.

. 2001. “Sowing Dragon’s Teeth: Public Support for Political Violence and Paramilitarism in Northern Ireland.” Political Studies 49 (5): 901-922. DOI: 10.1111/1467-9248.00346.

. 2017. “Trauma and Stigma: The Long-Term Effects of Wartime Violence on Political Attitudes.” Conflict Management and Peace Science 34 (4): 264-286. DOI: 10.1177/0738894215593683.

33.

Horowitz

Donald L

. 1985. Ethnic Groups in Conflict. 2nd ed. Berkeley, CA: University of California Press.

. 2017. “How Civil Wars End: The International System, Norms, and the Role of External Actors.” International Security 42 (3): 127-171. DOI: 10.1162/ISEC

35.

Human Rights Watch . 2016. Nepal: Decade after Peace, Scant Progress on Justice. New York, NY: Human Rights Watch.

36.

International Crisis Group . 2016. Nepal’s Divisive New Constitution: An Existential Crisis. Brussels, Belgium: International Crisis Group.

37.

Jankowski

Martin Sanchez

. 1991. Islands in the Street: Gangs and American Urban Society. Berkeley, CA: University of California Press.

. 2008. “From Words to Deeds: The Implementation of Power-Sharing Pacts in Peace Accords.” Conflict Management and Peace Science 25 (3): 206-223. DOI: 10.1080/07388940802218945.

. 2015. “Annualized Implementation Data on Comprehensive Intrastate Peace Accords, 1989–2012.” Journal of Peace Research. 52, 551-562. DOI: 10.1177/0022343314567486.

40.

Justino

Patricia

. 2009. “Poverty and Violent Conflict: A Micro-Level Perspective on the Causes and Duration of Warfare.” Journal of Peace Research 46 (3): 315-333. DOI: 10.1177/0022343309102655.

41.

Kalyvas

Stathis N.

2003. “The Ontology of ‘Political Violence’: Action and Identity in Civil Wars.” Perspectives on Politics 1 (3): 475-494. DOI: 10.2307/3688707.

42.

Kardov

Kruno

. 2007. “Remember Vukovar.” In Democratic Transition in Croatia: Value Transformation, Education & Media, edited by Ramet

Sabrina P

Matić

Davorka

, 63-89. College Station, TX: Texas A&M University Press.

43.

Kelman

Herbert C.

2008. “Reconciliation from a Social-Psychological Perspective.” In Social Psychology of Intergroup Reconciliation, edited by Nadler

, 15-32. Oxford; NY: Oxford University Press.

. 2013. “The Governance Institutions of a Drug Trafficking Organization.” Public Choice 156 (1-2): 95-103. DOI: 10.1007/s11127-012-0050-x.

. 2020. “Who Shot the Bullets? Exposure to Violence and Attitudes towards Peace: Evidence from the 2016 Colombian Referendum.” Latin American Politics and Society 62 (4): 24-49.

46.

Kreutz

Joakim

. 2010. “How and When Armed Conflicts End: Introducing the UCDP Conflict Termination Dataset.” Journal of Peace Research 47 (2): 243-250. DOI: 10.1177/0022343309353108.

47.

Lichbach

Mark I.

1994. “What Makes Rational Peasants Revolutionary? Dilemma, Paradox, and Irony in Peasant Collective Action.” World Politics 46 (3): 383-418.

. 2018. “Determinants of Colombian Attitudes toward the Peace Process.” Conflict Management and Peace Science 35 (6): 622-636. DOI: 10.1177/0738894218787783.

. 2017. “The Legacy of Political Violence across Generations.” American Journal of Political Science 61 (4): 836-851. DOI: 10.1111/ajps.12327.

. 2018. “Considering Concessions: A Survey Experiment on the Colombian Peace Process.” Conflict Management and Peace Science 35 (6): 637-655. DOI: 10.1177/0738894218787784.

. 2021. Citizen endorsement of contested peace settlements: public opinion in post-Dayton Bosnia. Democratization. 28(2): 434-452. DOI: 10.1080/13510347.2020.1828356.

52.

Nilsson

Desirée

. 2012. “Anchoring the Peace: Civil Society Actors in Peace Accords and Durable Peace.” International Interactions 38 (2): 243-266. DOI: 10.1080/03050629.2012.659139.

. 2020. “Violent Peace: Local Perceptions of Threat and Insecurity in Post-Conflict Colombia.” International Peacekeeping 27 (2): 238-262. DOI: 10.1080/13533312.2019.1677159.

. 2014. “Anti-Social Capital in Former Members of Non-State Armed Groups: A Case Study of Colombia.” Studies in Conflict & Terrorism 37 (12): 999-1023.

. 2010. “The Justice Balance: When Transitional Justice Improves Human Rights and Democracy.” Human Rights Quarterly 32 (4): 980-1007. DOI: 10.1353/hrq.2010.0021.

56.

Østby

Gudrun

. 2008. “Polarization, Horizontal Inequalities and Violent Civil Conflict.” Journal of Peace Research 45 (2): 143-162. DOI: 10.1177/0022343307087169.

. 2014. “Inside the post-Soviet De Facto States: A Comparison of Attitudes in Abkhazia, Nagorny Karabakh, South Ossetia and Transnistria.” Eurasian Geography and Economics 55 (5): 423-456.

. 2015. “The Power-Sharing Event Dataset (PSED): A New Dataset on the Promises and Practices of Power-Sharing in Post-Conflict Countries.” Conflict Management and Peace Science 32 (3): 327-350. DOI: 10.1177/0738894214542753.

59.

Pearlman

Wendy

. 2009. “Spoiling Inside and Out.” International Security 33 (3): 79-109.

. 2013. Nepal Votes for Peace. Delhi: Cambridge University Press India Pvt. Ltd.

. 2007. “Disaggregating Public Opinion on the Ethnic Conflict in Macedonia.” International Journal of Sociology 37 (3): 75-95.

. 2008. “War Experiences and War-related Distress in Bosnia and Herzegovina Eight Years after War.” Croatian Medical Journal 49 (1): 75-86.

. 2019. “Mass Repression and Political Loyalty: Evidence from Stalin’s ‘Terror by Hunger.’” American Political Science Review 113 (2): 569-583. DOI: 10.1017/S0003055419000066.

. 2008. “Case Selection Techniques in A Menu of Qualitative and Quantitative Options.” Political Research Quarterly 61 (2): 294-308. DOI: 10.1177/1065912907313077.

65.

Stedman

Stephen John

. 1997. “Spoiler Problems in Peace Processes.” International Security 22 (2): 5-53. DOI: 10.1097/EDE.0b013e3181.

66.

Stewart

Frances

. 2002. “Horizontal Inequalities: A Neglected Dimension of Development.” QEH Working Paper Series 81 (81): 1-40. DOI: 10.1057/9780230501850_5.

67.

Svensson

Isak

. 2007. “Bargaining, Bias and Peace Brokers: How Rebels Commit to Peace.” Journal of Peace Research 44 (2): 177-194. DOI: 10.1177/0022343307075121.

68.

Tellez

Juan Fernando

. 2019. “Worlds Apart: Conflict Exposure and Preferences for Peace.” Journal of Conflict Resolution 63 (4): 1053-1076. DOI: 10.1177/0022002718775825.

69.

Villamil

Francisco

. 2020. “Mobilizing Memories: The Social Conditions of the Long-Term Impact of Victimization.” Journal of Peace Research, 1 (3), 399-416. DOI: 10.1177/0022343320912816.

. 2012. “Violent Conflict and Behavior: A Field Experiment in Burundi.” American Economic Review 102 (2): 941-964. DOI: 10.1257/aer.102.2.941.

71.

Walter

Barbara

2002. Committing to Peace: The Successful Settlement of Civil Wars. Princeton: Princeton University Press.

72.

Wolff

Stefan

. 2001. “The Road to Peace: The Good Friday Agreement and the Conflict in Northern Ireland.” World Affairs 163 (4): 163-170.

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

0.99 MB