Social Dominance Orientation and exposure to violence as predictors of support for past peace agreements

ABSTRACT

The purpose of this study is to explore individual predictors of support for peace agreements after their implementation. The aim is to test if individual differences in Social Dominance Orientation (SDO) temper the effect of exposure to violence on support for peace agreements. One hundred and eighty participants completed measures surveying their SDO, exposure to violence and voting behavior four years after the peace agreement between the Colombian government and the FARC in 2016. Results show that SDO lowers support for the peace agreement four years after it was signed among those who experienced violence but has a smaller effect on those who did not experience violence. The implications of the interaction between SDO and exposure to violence are discussed.

KEYWORDS:

• Social dominance orientation
• peace referenda
• peace agreements
• Colombia
• exposure to violence

Introduction

Colombia’s newly elected President Gustavo Petro is seeking a paz total, that is, total peace with all remaining armed actors throughout the country following the 2016 peace agreement with the Fuerzas Armadas Revolucionarias de Colombia (FARC). There is considerable evidence showing that previous violent victimization increased the proportion of votes for peace, at least in the case of Colombia’s 2016 referendum (Branton et al., 2019; Pechenkina & Gamboa, 2022; Tellez, 2019). Years later and in light of further negotiations and possible referenda under Petro’s leadership, this article explore to what extent people’s experiences of violence and support for peace is moderated by individual dominance preferences. Taking into account the interaction of people’s experiences of violence and personal preferences could explain support for peace in future peace referenda in Colombia and active conflict zones elsewhere.

Referenda are relatively rare, and outcomes depend on political and economic issues and on the level of support for the incumbent government or opposition. However, peace agreements have increasingly been put to a vote in countries such as Colombia, Northern Ireland, and South Sudan, leading to growing practical and theoretical interest in the factors influencing public opinion vis-à-vis either supporting or rejecting them. Support for peace agreements matters to prevent support for parties who prefer the continuation of violence and a potential conflict relapse. The reasons for peace support include situational conditions, such as economic stability, military capacity, or mutually hurting stalemates, or the involvement of civil society actors in peace negotiations (Orjuela, 2003) and the framing of campaign information (Masullo & Morisi, 2019).

Individuals might also support and reject peace referenda for reasons unconnected to the nature of the state entities and political elites. Despite concerns regarding the influence of the unemployment rate on individuals’ preferences, studies of the Colombian case have described why support for the opposition party and the experience of victimization were ultimately the most important factors determining voting patterns (Dávalos et al., 2018; Weintraub et al., 2015). Indeed, exposure to violence and living in war zones seem to explain much of the variation witnessed in the results of the peace referendum in Colombia (Branton et al., 2019; Pechenkina & Gamboa, 2022; Tellez, 2019).

However, many contributions from psychology highlight that conflict exposure can make citizens feel a ‘hardening’ against the perpetrator or enemy populations, as driven by grievances and hatred (Hirsch-Hoefler et al., 2016; Zeitzoff, 2014). Exposure to violence is associated with psychological stress that leads to rejecting compromise and favoring militarism (Canetti et al., 2013; Juan et al., 2023). The role of Social Dominance Orientation (SDO) has recently received increasing attention in explaining political beliefs and attitudes among citizens that tolerate the continuation of injustices to maintain social hierarchies (Espinosa et al., 2022). Hence, SDO may play an important role in support of peace after protracted conflict.

In the current study, I therefore test whether the diverse effects of experiencing violence is conditioned by preferences for group-based hierarchies in society four years after the peace agreement in Colombia in 2016. The interaction is especially relevant after peace agreements because citizens who experienced violence have to accept that former perpetrators challenge existing social hierarchies in the years following the agreement.

The Colombian context

Colombians suffered in different ways from the conflict because violence against civilians was unevenly distributed across the national territory, thus affecting the population unequally (Kreiman & Masullo, 2020). The ongoing Colombian conflict dates to the 1960s. Over the past five decades, leftist guerrilla groups, right-wing paramilitary organizations, and Colombian state forces have all been involved in violence against the civilian population. Among the various guerrilla groups operating in the country, the FARC quickly became the largest and most powerful and all Colombian presidents since Julio César Turbay in 1978 attempted to conclude a peace agreement with them. The 2016 peace agreement aimed to provide a comprehensive rural reform, give permission for the FARC to participate in elections, offer a strategy to curb coca cultivation and stop drug trafficking, and bring truth and justice for victims. Addressing strong grievances and injustices was at the core of the peace agreement and was particularly relevant for the Colombian context with a very unequal society and a Gini index that makes the country one of the most unequal countries in Latin America (Romero, 2022).

The referendum was to confirm that Colombians supported a peace agreement between the two parties. It narrowly failed − 50.2% voted against and 49.8% voted in favor. Shortly after the referendum failed, the Colombian government and the FARC signed a revised peace agreement in November 2016 which was ratified by the congress and not put out for a second referendum. The implementation of the peace agreement started immediately but came to a halt when Iván Duque became president of Colombia in 2018 who firmly rejected the agreement. In the years that followed, addressing strong grievances and injustices was pushed to the back and the full implementation of the agreement stagnated. Sentiment against the FARC rose when not all ex-combatants demobilized and popular support for the FARC party (renamed Comunes in 2017) was extremely low in the first postagreement parliamentary elections, with less than 0.5% of the electoral vote in 2018. Since 2019, many victims have been frustrated by FARC’s unwillingness to talk to the Special Jurisdiction for Peace (JEP) regarding forced abductions, among other crimes.

Theoretical discussion

The experience of violence promotes ingroup preferences, outgroup distrust, and also distrust toward the state who failed to protect survivors of violence. There is no consensus as to whether the exposure to violence increases the preference for supporting peace agreements in general and whether other individual characteristics temper the effect of experiencing violence. Peace agreements can represent a form of positive peace, local development, or demobilization of fighters who previously committed violence. If the latter is the main association of citizens, individual characteristics such as a preference for group-based hierarchies in society are particularly relevant to citizens’ support for peace agreements. After peace agreements, societies have to accept former fighters into society or even into politics. Former fighters often receive financial and educational support when they disarm in countries in which there are often no welfare benefits for lower-income citizens .

An explanation the has received little attention in existing research on accepting former combatants as political actors is the role of status quo-legitimizing ideologies such as SDO (Sibley & Duckitt, 2010; Zeigler-Hill et al., 2021). SDO, defined as the individual disposition to accept hierarchy and dominance among groups, is an important difference between persons regarding the extent to which they endorse a system that upholds injustice or not. Individuals high in SDO accept intergroup inequality and are generally in favor of social norms that promote intergroup domination (Pratto et al., 1994). SDO legitimizes hierarchical power structures, and it could be fruitful to understand when individuals prefer to maintain the hierarchy of a certain group through peace agreements.

Seminal work in psychology has found that SDO correlates with attitudes toward war and peace (Bizumic et al., 2013; Blumberg et al., 2017; Heaven et al., 2006; Zeigler-Hill et al., 2021). There is evidence showing that people high in SDO tend to disapprove of dovelike and conciliatory attitudes when these challenge group-based hierarchies (Cohrs et al., 2005; Ho et al., 2015; Kleppestø et al., 2020). Those with higher levels of SDO should not support either war or peace for the sake of conflict but for the sake of maintaining their social hierarchy (Henry et al., 2005; Lucas & Kteily, 2018). SDO depends on the status of the perpetrator(s) and the granted social upward mobility (Henry et al., 2005; Porat et al., 2015).

For the Colombian context, it became clear in the 2018 elections, two years after the FARC signed the peace agreement and disarmed, that they would not receive much public support as political actors. Taking their political failure into consideration, an important aspect in connection with experiences of violence and later support for peace agreements seems to be the question of whether the perpetrators go unpunished or whether they become political actors, for example, by founding a political party (Montoya, 2014). Many Colombians criticized the fact that former FARC fighters were given five seats in Congress and the Senate for the following two legislative periods through the special arrangement of the peace agreement. Moreover, many communities affected by the violence have difficulties reintegrating former combatants into society or even tolerating them, considering that more than 300 demobilized FARC combatants were killed between 2016 and 2020.

Recent studies on SDO in Colombia indicate that it correlates with a set of political beliefs and attitudes among citizens that tolerate and favor dishonest and corrupt actions on the part of political and economic decision-makers (Espinosa et al., 2022). Colombian voters who claimed that they generally prefer a military solution to the conflict with the guerrillas to a negotiated one were more likely to be right-wing and high in SDO compared with counterparts who supported the government of President Juan Manuel Santos (García-Sánchez et al., 2022; Muñoz & Pachón, 2021). Colombians with high levels of SDO also displayed more support for the conflict in general (García-Sánchez et al., 2022) and prejudices against FARC members, indicating that negative attitudes toward ex-combatants are related to perceptions of threat vis-à-vis social order or status (Ramírez, 2018).

Previous research has shown that individuals with a higher SDO are generally less likely to approve or take political action whereas individuals who have experienced violence are more likely to support changes (Lemieux & Asal, 2010). It is to be expected that individuals with a high SDO will be unwilling to accept a peace agreement that changes intergroup hierarchies and allows those lower-status individuals who have committed violence to become political actors or even common citizens. When peace agreements are negotiated, some former perpetrators are reintegrated into the regions from which they came and committed most of the violence and others aspire to enter politics. Individuals who have not been subjected to violence should be unconcerned regarding former perpetrators returning to their regions or entering politics independent of their SDO.

In the case of Colombia and other armed conflicts, people living in rural areas are often disproportionately affected by violence. Most of the perpetrators also come from rural areas and often belong to the lower strata of society. I argue that citizens who have suffered violence at the hands of these ex-combatants, and who associate their demobilization with peace agreements, are less likely to support previous peace agreements if they favor existing hierarchies. Existing hierarchies generally do not allow ex-combatants to reintegrate into society and even to advance socially. People who have experienced violence may be strongly opposed to giving ex-combatants the opportunity for social advancement through a peace agreement if they prefer the maintenance of these hierarchies. Those who have not experienced violence are unlikely to be concerned about the social advancement or rural reintegration of ex-combatants.

Pulling together the relevance of exposure to conflict from studies in Colombia and psychological studies on perceptions of inequality, I hypothesize that SDO and the experience of violence should interact in influencing support for the 2016 peace agreement in Colombia. I aim to test how preferences for social hierarchies and exposure to violence interact in future peace referenda in Colombia and active warzones elsewhere. Given that peace processes are closely related to a change in hierarchies, it is tested if SDO is related to the willingness to support such processes. Additionally, exposure to violence is expected to influence support for peace in the Colombian context because the most conflict-affected regions voted overwhelmingly in favor of the 2016 peace agreement.

Method and materials

Participants and design

One hundred eighty Colombian students (98 females and 82 males) participated in this study between October and December 2020. Participants’ ages ranged from 16 to 44 years (M = 22.16; SD = 3.77). Students were recruited from six universities in Bogotá and Medellín (Universidad Nacional de Bogotá, Universidad de los Andes, Universidad Externado, Universidad Nacional de Medellín, Universidad de EAFIT, and Universidad de Antioquia). They were from different regions in Colombia to make sure that participants had lived both within and outside of a conflict zone.¹ Participants’ scores on scales measuring SDO and exposure to violence were tested as predictors of support for the 2016 peace agreement.

Materials

Social dominance orientation was measured using the SDO 7(s) 8-item short scale (Ho et al., 2015), employing statements regarding social dominance and antiegalitarianism (see full text in the Appendix). Although SDO consists of the subdimensions SDO-dominance (SDO-D), which is the desire that people have for some groups to be actively oppressed by others, and SDO-egalitarianism (SDO-E), which is a preference for intergroup inequality, this short scale is used to reflect an overall tendency of accepting hierarchy and dominance among groups. The eight statements in the scale reflect a general preference for group hierarchy (e.g., ‘group equality should not be our primary goal’). A five-point Likert scale was used (1: strongly disagree; 5: strongly agree), yielding an internal reliability of α = 0.86. Higher SDO scores indicate greater levels of support for intergroup status hierarchies. A low SDO has to be lower than 1.5, whereas a high SDO should be above 3.5 (with medium SDO in between).

Exposure to violence was measured by asking if the person had experienced a form of violence due to the armed conflict, with ‘Yes,’ ‘No,’ and ‘Prefer not to answer’ as the options for the answer. Subjects could additionally describe the situation or could make a general comment. Of the 21 participants who made a comment, 12 stated that they had experienced physical violence at the hands of unspecified armed actors; 6, at the hands of guerrillas; 2, at the hands of paramilitaries; and 1, at the hands of the police. In total, 71 participants reported no exposure to violence and 109 claimed having experienced some form of violence. This was a high number but not unusual according to the survey firm Cifras y Conceptos, which collected the data. Because participants were given the opportunity to self-identify as someone who had experienced violence according to their own definition of violence, it appeared that they were more receptive to broader definitions of the phenomenon. Although not all experiences of violence are equally traumatic, the question was kept as open-ended as possible to allow more participants to identify themselves as having experienced it.

To allow participants to rate how much they supported the peace agreement with the FARC, I used a scale from 0 to 10 where 0 means ‘I totally reject the peace agreement of 2016’ and 10 means ‘I totally support the peace agreement of 2016’. The scale specifically mentioned the year of the peace agreement to avoid participants who voted against the one negotiated by Santos claiming to support such agreements in general.

Procedure and ethics

To boost participation, the study design was kept short and simple – needing only 15 min to complete it. All measurements were validated in Spanish versions. Data were collected by the survey firm Cifras y Conceptos. The latter recruited participants who studied in Bogotá and Medellín but who came from different regions of Colombia. Participants took the survey online in their own homes. The team administered the survey and made follow-up calls to confirm that the survey was answered.

Participants were provided with information and consent forms prior to the study and a debriefing afterward. They were free to withdraw their data at any time during the study and up to a week afterward. Each participant received 20,000 Colombian pesos for participating (approximately 5 US dollars). Participants were promised full anonymity. The proposed study underwent an ethical peer review prior to data collection at the authors’ department. The statistical procedure was designed first to investigate support for the 2016 peace agreement, and then, it measured SDO and descriptive statistics on exposure to violence, general opinions on the use of the latter, and demographic statistics on age, gender, study subject, marital status, degree, military service, and socioeconomic stratum. The study protocol can be found in the Appendix.

Analysis

Data were analyzed using the PROCESS macro for SPSS (Hayes, 2013) to test the moderation model, in which exposure to violence was considered the predictor variable; support for the 2016 peace agreement, the outcome variable; and SDO, the moderator. I added age, gender, military service, socioeconomic stratum, and general attitudes toward violence as control variables. The results were bootstrapped to account for the small sample size and to perform hypothesis testing.

Results

First, the two predictor variables and control variables were tested for correlation with support for the 2016 peace agreement. Table 1 illustrates the results of this analysis.

Table 1. Correlation matrix for each predictor variable regarding support for the 2016 peace agreement bootstrapped.

Variable	Mean (SD)	1	2	3	4	5	6	7	8
(1) Support for the 2016 peace agreement	4.97 (3.36)	1
(2) Exposure to violence		0.42**	1
(3) SDO	2.31 (0.98)	−0.28**	0.25*	1
(4) Socio-economic stratum	3.35 (1.01)	0.13	0.05	0.01	1
(5) Gender (male)		0.11	0.32**	0.11	−0.02	1
(6) Age	22.16 (3.77)	−0.08	0.00	0.18*	−0.15	−0.14	1
(7) Military service		−0.23**	−0.45**	−0.09	−0.14	−0.60**	0.19*	1
(8) Violence general	2.09 (0.83)	0.05	−0.05	−0.10	−0.04	−0.09	0.01	0.10	1

**p < 0.01, *p < 0.05. SDO = Social Dominance Orientation. Pearson correlations were used for support for the 2016 peace agreement, exposure of violence, and age; Spearman Correlations for socio-economic stratum, violence general, and SDO; Tetrachoric correlations for gender and military service.

The two predictors significantly correlated with support for the 2016 peace agreement in question. As expected, SDO correlated negatively and exposure to violence correlated positively with support for the peace agreement. Exposure to violence also correlated positively with SDO.² 64.1% reported they had experienced violence, while 35.9% did not. 62.6% of the participants were men and 37.4% were women. Military service correlated negatively with the dependent variable. 58.8% had not served in the military, compared to 40.5% who had served and 0.8% who gave no answer. It should be noted here that Colombia has a compulsory military service for men, which is often circumvented by paying a so-called ‘tax’ to avoid the service. The average support for the 2016 peace agreement was 4.97 (SD = 3.36).

Then, the following model tested the effect of exposure to violence on support for the 2016 peace agreement with SDO as a moderator of peace support using Model 1 in PROCESS with 5,000 iterations with bootstrapping. The overall model was significant (F (8,122) = 17.69, p < .001) and accounted for 46% of the variance (R² = 0.46). The interaction between exposure to violence and SDO significantly predicted support for the 2016 peace agreement; R² change = 0.11, F (8,122) = 29.13, p < .001; b = −2.77, t (122) = −5.40, p < .001, confidence intervals between −3.78 and −1.75. Figure 1 illustrates how increasing levels of SDO moderate the effect of exposure to violence on the support for the 2016 peace agreement. It can be confirmed that SDO lowers support for the treaty among those who experienced violence but has a smaller effect on those who did not experience violence. The graph shows that at levels of SDO higher than 3.24, there is no difference in support for the treaty between those who experienced violence and those who did not experience violence. A summary table of the bootstrap results for this model can be found in Table A1 in the Appendix, including socioeconomic stratum, age, gender, military service, and general attitudes toward violence.³ The effect of exposure to violence on support for the 2016 peace agreement, conditional on SDO, shows that the effect of exposure to violence shrinks as SDO increases (see Figure 1). In summary, the aim of the current study was to test if exposure to violence, preference for group-based hierarchies, and the interaction of the two (continue to) predict support for the 2016 peace agreement 4 years after the peace referendum in Colombia.

Figure 1. Moderation of the effect of exposure to violence on support for the 2016 peace referendum at values of the moderator Social Dominance Orientation (SDO).

As predicted for the Colombian context, experience of violence was positively correlated with support for the 2016 peace agreement. In line with past research on SDO and attitudes toward peace, SDO was negatively correlated with support for this agreement. The moderation analysis allowed better understanding of how these factors interacted with one another, revealing that SDO moderated the effect of exposure to violence. Participants who experienced violence and who have higher levels of SDO display a lower support for the 2016 peace agreement than those with no experience of violence. At higher levels of SDO (>3.24), there is little difference in support for the treaty between those who experienced violence and those did not experience violence.

Discussion

The results suggest that SDO lowers support for the 2016 peace agreement if the person has an experience of violence. These individuals might prefer the continuation of the conflict without offering reintegration or social advancement to former combatants because of what they personally experienced. For those who did not experience violence, SDO has no impact on support for the 2016 peace agreement four years after its implementation. Perceptions and preferences for hierarchical dominance of certain groups over others may be the reason that experiences of violence have differential effects on support for the 2016 peace agreement. These results indicate how a preference for group-based hierarchies could affect support for peace agreements in active warzones.

At this point, the specifics of support for peace agreements are worth reflecting. It can be argued that negative attitudes toward peace as a phenomenon are caused by social inequality as many people take up arms when their social and economic grievances become intolerable. However, the biggest challenge in getting public (rather than combatant) support for peace agreements is selling people on deals that offer social mobility to those who took up arms (Mattes & Weeks, 2019; Schultz, 2005). Most solutions to armed conflict depend on governments offering concessions, for example, by offering former combatants impunity for past crimes, pensions, and pathways to social change and political participation. More so, these former combatants may show no remorse for their actions, reinforcing the impression that a peace agreement exempting them from facing the consequences of committing crimes is not fair. Understanding that allowing combatants to reintegrate into society seems unfair to some is key to understanding the rejection of peace referenda.

In Colombia, the ‘No’ campaign against the 2016 peace agreement succeeded in leveraging the fear of losing personal benefits such as pensions to the benefit of former combatants. Many Colombians considered the concessions to the FARC unfair as people who have worked all their lives without taking up arms and committing crimes often do not receive pensions (Matanock & Garbiras-Díaz, 2018). Interestingly, the ‘No’ campaign not only claimed that the concessions were unmerited but also promoted the idea that a peace agreement with the FARC would turn Colombia into a communist state, which would not be fair to the people who had worked their way up the social ladder for years (Cardona Zuleta & Londoño Álvarez, 2018). Simultaneously, the ‘No’ campaign claimed to protect the ‘traditional’ Colombian family from a peace agreement that would promote ‘unhinged’ sex education and fluid gender identities (Matanock & Garbiras-Díaz, 2018).

It is useful to reflect on the significance of these fears for the results in this study four years after the signing of the peace agreement. In those four years, the opponents of the peace process have brought to power a president who has neither reversed the entire agreement nor actually implemented it. Attitudes toward the 2016 vote continued to be debated, with the supporters and opponents of the peace agreement blaming each other for what they perceived to be current political and economic shortcomings. However, the results suggest that even if the social mobility of former combatants did not occur in the manner feared by peace opponents in 2016, a preference for group-based hierarchies continued to influence past rejection of the peace agreement.

Peace agreements disrupt group-based hierarchies regardless of the concessions and provisions that they contain. They may be more legitimate for one group (the previously excluded one) than another (the previously dominant group), making measures such as SDO particularly pertinent in these contexts. Support for peace agreements may depend on whether individuals can identify with the suffering of others, for example, by experiencing violence themselves. Individuals low in SDO might be more capable of perceiving injustices based on their experience of violence and allowing the possibility of social changes through inter alia peace agreements. It could also be that those who experienced violence have nothing to lose compared to those at the top of society – who potentially fear that their power and standing will be challenged by any change in the social order. Individuals who experienced violence and prefer group-based hierarchies may be fearful or skeptical of social change and uncertain outcomes following a peace agreement being signed. These individuals may want to maintain those hierarchies as long as there is no safe and peaceful alternative to war.

Limitations and possible avenues of future research

This hypothesis-testing study tested whether SDO interacts with the effect of experiencing violence on supporting past peace agreement. The data of this study are consistent with the hypothesis and indicate that SDO lowers support for the treaty among those who experienced violence but has a smaller effect on those who did not experience violence. The study had its limitations, which were primarily connected to the sample: First, a student sample lacks representability because it does not reflect the heterogeneity of the individuals affected by the conflict. The sample is younger and more educated relative to the general population and cannot reflect the diversity of experiences with violence during the Colombian conflict. Additionally, the study was conducted with a modest size and the controlled conditions involving university students were obviously far removed from real-time government policymaking and of mass-opinion formation. Nevertheless, these students are part of the population and display varying levels of exposure to violence and levels of SDO.

Second, using SDO as a moderator has its limits because there is a debate regarding which scores are ‘truly’ low or high (Kleppestø et al., 2020). SDO is often skewed toward the lower end of the scale – raising questions regarding how to best categorize these scores (see Appendix Figure A1). However, even if most people reject strong intergroup inequality, the variance between strong and more modest rejections of intergroup hierarchy, as captured by the SDO scale, has been shown to be one of the most robust predictors of intergroup phenomena in political psychology.

Third, the independent variable, experience of violence, could also be expanded to understand what types of violence elicit distinct preferences for peace among different individuals. As such, sexual violence and violence as a part of military service might affect individuals in very distinct ways. Recent studies on exposure to violence in the military found that soldiers who experienced combat are more supportive of punitive peace in the context of foreign policy (Blair & Horowitz, 2021) and less so of negotiated settlements in general (Grossman et al., 2015). However, there are many ethical concerns associated with studying experiences of violence, and careful consideration should be given to ensure that participants are not retraumatized by surveys that include questions regarding such matters.

Future studies should seek confirmation for the claims made here regarding why SDO lowers the support for peace among those who experienced violence. For instance, they could replicate the study on a sample of current political leaders. The inclusion of a ‘political elites’ sample could contribute to better understanding of how support for peace processes hinges on who exactly is in power. Although elites are influenced by the public, they can also forgo offers of peace so that negotiations depend on their willingness to choose settlement over conflict (Tappe Ortiz, 2020). For example, a policy vignette could make it easier to separate personal resentment of the Santos government from actual low support for a peace agreement with the FARC. Future research should compare the results with data on the severity of attacks in the regions where each study participant lives/lived. Other potential moderators should also be considered. Research by Bizumic et al. (2013) has shown that a right-wing authoritarian (RWA) ideology is another crucial individual difference to be considered vis-à-vis attitudes toward war. RWA can also function as a system justifier in moderating the effect of exposure to violence on support for peace agreements.

Another avenue could be investigating why women often account for most civilian victims in conflict and often suffer from sexual violence that targets specifically their sexuality and bodies (Cohen & Nordås, 2014). Consequently, women might mobilize for peace agreements as a response to the collective threat of victimization (Shesterinina, 2016). Studies on gendered civilian agency are still nascent. However, there are some contributions on women’s empowerment in Colombia indicating that exposure to violence can indeed be a push factor for mobilization (Kreft, 2018; Zulver, 2022). Future studies could take a closer look at how women’s exposure to violence is tempered by SDO in situations of armed conflict.

In conclusion, this study lays a possible foundation for a future study that will test the hypothesis around exposure to violence and social hierarchy preferences in active conflict zones elsewhere that might be set to sign peace agreements in the future, like Afghanistan or Ethiopia. It has shown that years after the peace referenda, the effect of exposure to violence on support for the 2016 peace agreement between the Colombian government and the FARC is diminished by SDO. By focusing on experience of violence, this research broadens the scope of this relationship while underlining the importance of addressing existing social hierarchies if public support for negotiated solutions to conflict is going to increase – whether in Colombia or elsewhere.

Public significance statement

Exposure to violence was associated with support for a peace agreement in 2016 between the Colombian government and an armed group. Four years later, the associations between exposure to violence and support for the 2016 peace agreement are reduced by ideological attitudes such as Social Dominance Orientation.

Acknowledgments

I would like to thank Matthias Basedau, Elisabeth Höhne, Juan Masullo, Julia Köbrich, the editor and the two reviewers for the detailed comments. I received funding from the German Institute for Global Affairs and Area Studies and from the Stiftung der deutschen Wirtschaft (sdw).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are publicly available.

Notes

1. A student sample was used because students were the agents who pushed for the peace agreement when the plebiscite failed. The study would benefit from a replication with a more representative sample.

2. This correlation could indicate that exposure to violence is not independent of SDO if people are high in SDO because they experienced violence, but the results remain robust.

3. A simple regression of the main effects, the interaction effects, and the control variables can be found in Appendix Table A2.

 

Appendices

Study protocol in English

How much do you agree or disagree with the 2016 peace agreement between the Colombian government and the FARC?

Please rate your opinion on a scale from 0 to 10, where the value 0 means “I totally reject the 2016 peace agreement” and the value 10 means “I totally support the 2016 peace agreement.”

Show how much you favor or oppose each idea below by selecting a number from 1 to 5 on the scale given:

• 1. An ideal society requires some groups to be on top and others to be on the bottom.

• 2. Some groups of people are simply inferior to other groups.

• 3. No one group should dominate in society.

• 4. Groups at the bottom are just as deserving as groups at the top.

• 5. Group equality should not be our primary goal.

• 6. It is unjust to try to make groups equal.

• 7. We should do what we can to equalize conditions for different groups.

• 8. We should work to give all groups an equal chance to succeed.

All scaled from 1 (“strongly oppose”) to 5 (“strongly favor”). Items 3, 4, 7, and 8 are reverse scored.

Have you experienced any form of violence due to the armed conflict?

• Yes (could you please specify in the comment box)

• No

Have you served in the military?

• No

• Yes without combat experience

• Yes with combat experience

What do you think about the general use of violence?

Violence is…

• Always justified

• Often justified

• Sometimes justified

• Rarely justified

• Never justified

Summary appendix tables and figures

Table A1. Summarizing bootstrap results for exposure to violence as a predictor of support for the 2016 peace agreement moderated by SDO (including control variables).

	Coefficient	Bootstrap mean	Bootstrap SE	Lower confidence interval	Upper confidence interval
Constant	−0.00	0.05	2.56	−4.71	5.30
Experience of violence	9.49	9.63	1.25	7.18	12.10
SDO	0.42	0.50	0.43	−0.20	1.54
Exposure to violence × SDO	−2.77	−2.84	0.51	−3.94	−1.92
Socioeconomic stratum	0.49	0.49	0.24	−0.00	0.96
Age	0.02	0.01	0.06	−1.11	0.11
Gender (male)	0.21	0.19	0.68	−1.15	1.53
Military service	−0.03	−0.03	0.37	−0.76	0.68
General use of violence	0.07	0.06	0.29	−0.50	0.66

Table A2. Simple (bivariate) regressions on support for the peace agreement 2016.

Regression	Model 1	Model 2	Model 3	Model 4
Experience of violence	2.79** (.44)		3.76** (.63)	9.49** (1.26)
SDO		−1.09** (.21)	−1.50** (.29)	0.42 (.46)
Exposure to violence × SDO				−2.77** (.55)
Socio-economic stratum			0.35 (.27)	0.49 (.24)
Age			0.10 (.06)	0.02 (.06)
Gender (male)			−0.14 (.70)	0.21 (.61)
Military service			−0.12 (.40)	−0.03 (.32)
General use of violence			0.16 (.31)	0.07 (.28)
R^2	0.17	0.10	0.35	0.47

**p<0.01, *p<0.05. Bootstrap results are based on 1000 bootstrap samples.

Figure A1. Simple boxplot of social dominance orientation by support for the 2016 peace referendum.