Keeping electoral peace? Activities of United Nations peacekeeping operations and their effects on election-related violence

Abstract

In war-torn countries, elections are held to support peacebuilding, but they sometimes trigger new violence. While peacekeeping operations (PKOs) regularly accompany electoral periods, we lack systematic knowledge on how they influence election-related violence. I argue that variation in peacekeepers’ activities is fundamentally important: only if PKOs assist with securing and organizing elections can they reduce election-related violence. Using novel data on PKOs’ election-related activities and accounting for endogeneity in both peacekeeping deployment and activities, the analyses of all 630 elections in conflict-affected countries support this expectation. The result implies that the design of PKOs is crucial for effectively managing post-war political transitions.

Keywords

Peacekeeping elections election-related violence peacebuilding democracy

Introduction

Do United Nations (UN) peacekeeping operations (PKOs) contribute to more peaceful elections in conflict-affected countries and, if so, how? Since the end of the Cold War, elections have become a prominent feature of war-to-peace transitions and often a ‘core business’ of multidimensional peacekeeping operations (United Nations Department of Peacekeeping Operations, 2008: 22). According to the Capstone Doctrine, a high-level guideline for contemporary peacekeeping operations, peaceful elections are a benchmark for peacekeeping success and even a pre-condition for handing over peacebuilding responsibilities to post-war governments (United Nations Department of Peacekeeping Operations, 2008: 87–89). Whether and how PKOs help stabilize the electoral environment is also important for the peacebuilding process. Peaceful elections can support the establishment of political order after war by providing legitimacy for post-war governments, establishing an accountability mechanism and offering incentives for good governance (Reilly, 2008; Wantchekon, 2004).

While promoting elections as a stepping-stone to peace, the UN also recognizes its dangers. For instance, the High-Level Panel on United Nations Peacekeeping Operations (2015: 48) notes that ‘peace processes do not end with […] an election. Such events constitute merely a phase, rather than the conclusion, of a peace process. In fact, they may be times of great vulnerability’. In the 1989–2012 period, about one-third of the elections held in conflict-affected countries were marred by severe events of election-related violence, that is, acts of coercion, intimidation, or physical harm perpetrated to affect an electoral process or that arise in the context of electoral competition (Sisk, 2008a).¹ Violent elections hamper peacebuilding by exacerbating divisions in society (Gutiérrez-Romero, 2014), reducing government legitimacy (Collier and Vicente, 2012) or even triggering renewed fighting (Brancati and Snyder, 2013). Thus, whether and how PKOs help prevent election-related violence is an important question.

My argument starts from the premise that election-related violence is distinct from other forms of political violence in terms of its purpose, i.e. to affect the electoral process and outcome, and its targets, i.e. voters, candidates, election workers and materials. Only if PKOs adapt their activities to the distinct nature of election-related violence can they contribute to peaceful elections. Specifically, I argue that peacekeepers’ effectiveness in reducing incentives for coercive electoral tactics hinges on whether they engage in election-related activities, i.e. ‘efforts to organize, monitor and carry out free and fair elections through the provision of security, technical advice, logistical support and other forms of electoral assistance’ (United Nations Department of Peacekeeping Operations, 2008: 28). In so doing, they can augment the costs of electoral coercion and reduce electoral manipulation and, in turn, mitigate stakeholders’ incentives for employing election-related violence. In contrast, if PKOs are not engaged in election-related activities, their attention and resources are likely to shift to other priorities and they fail to mitigate election-related violence.

Case studies on peacekeeping in post-conflict elections support this conditional effect (Kumar, 1998). For instance, Mozambique’s first post-conflict elections in 1994 remained largely peaceful. This outcome is credited to ‘the presence of the United Nations Operations in Mozambique’, its ability to supervise domestic security forces’ electoral security operations and ‘its willingness to investigate any complaint, regardless of its plausibility, [which] ensured transparency and built confidence in the process’ (Turner et al., 1998: 171). Similarly, Sisk (2008b: 216) finds that in Liberia in 2005 ‘both rounds of elections went off peacefully, in no small measure attributable to … the ability of the UNMIL force [the PKO] to provide security’. In contrast, if PKOs do not assist with electoral security or the election organization, they are likely to fail in preventing election-related violence. For instance, the PKO in the 1992 elections in Angola had neither the mandate to secure elections nor the power to postpone them to a less volatile period (Ottaway, 1998: 144). The Angolan elections were marred by violence and led back to war. Likewise, Leininger (2006: 512) concludes for Haiti that ‘the peacekeeping missions’ narrow focus on police reform was rather to the disadvantage of democratization’. Indeed, the scrapping of all election-related functions from the PKOs’ mandate correlated with widespread unrest during the 1997 parliamentary elections in Haiti. In short, the effect of PKOs depends on their engagement in election-related activities.

Using novel data on PKOs’ election-related activities, I provide the first study to systematically test this expectation across a global sample of all 630 elections during or after the end of an armed conflict in the 1989–2012 period and a reduced sample of 92 elections accompanied by a PKO in the same period.² Evaluating the effectiveness of PKOs is generally complicated by selection effects (Gilligan and Stedman, 2003). In the absence of good instruments for PKOs’ presence and election-related activities, I use statistical matching to better approximate PKOs’ effects. Matching pre-processes the original sample of electoral periods to create two balanced samples where influential covariates are similarly distributed across (a) periods with and without a PKO and (b) periods with and without PKOs’ election-related activities, respectively. In both the original sample and the two post-matching samples, the argument receives empirical support. PKOs reduce election-related violence if they are actively involved in assisting with elections.

This article’s contribution is three-fold. First, the peacekeeping literature largely agrees that PKOs reduce war-related violence (Fortna, 2008; Beardsley, 2011; Hultman et al., 2016). It remains debated, however, whether PKOs mitigate violence that is below the threshold of war and that does not directly mimic the wartime master cleavage (Autesserre, 2014; Bara, 2020; Di Salvatore, 2019; Fjelde and Smidt, 2021). This article contributes to this debate on peacekeeping effects beyond war and provides first systematic evidence on whether and how PKOs help reduce violence related to elections (see Fjelde and Smidt, 2021 for the only systematic study). Second, a small but growing body of literature finds that peacekeepers’ activities on the ground matter for peacekeeping effectiveness (e.g. Dorussen and Gizelis, 2013; Heldt, 2011; Hultman, 2010; Murdie and Davis, 2010; Smidt, 2020b). Yet, election-related activities, a core part of the peacekeeping agenda, have been overlooked in cross-national research on peacekeeping (see Mvukiyehe and Samii, 2017; Smidt, 2020a, for quantitative case studies on peacekeepers’ election-related activity). Using original cross-national data on PKOs’ activities, this study fills this gap. Its findings point to one causal pathway through which peacekeepers may help stabilize post-war political transitions. Finally, the study contributes to our growing knowledge on how international organizations shape domestic electoral politics. It thus complements assessments of election monitoring (Daxecker, 2012; Hyde and Marinov, 2014; Little, 2012; Luo and Rozenas, 2018; Simpser and Donno, 2012; von Borzyskowski, 2019a, b) and technical election assistance (Birch and Muchlinski, 2018; Lührmann, 2018; von Borzyskowski, 2016).

After reviewing existing research on international interventions in elections, I develop my argument on how peacekeepers’ election-related activities mitigate election-related violence. The subsequent sections detail the research design and present evidence for the argument. The article concludes by suggesting how its findings inform theory and policy.

What we know about peacekeepers’ effects in elections

An increasing number of contemporary PKOs engage in election-related activities, ranging from reforming electoral law and transporting election equipment to securing polling stations and election rallies (Kumar, 1998; United Nations Department of Peacekeeping Operations, 2008). Peacekeepers actively assisted with about 15% of all elections in conflict-affected countries in the 1989–2012 period. Five currently ongoing PKOs in the Central African Republic, the Democratic Republic of the Congo, Kosovo, Mali and South Sudan are tasked with supporting elections. Yet, we have little systematic knowledge about the contribution that peacekeepers may make to peaceful elections across countries and over time. Given the prominence of elections in peacebuilding strategies, it is timely to study the links between peacekeeping and election-related violence after war.

Existing research on international intervention in electoral politics focusses on election monitoring and technical election assistance (and not peacekeeping). International election monitors deter violent and fraudulent behaviour by exposing and shaming its perpetrators (Daxecker, 2012; Hyde, 2010; Hyde and Marinov, 2014; Little, 2012; Simpser and Donno, 2012). Although election monitors’ exposure of fraudulent behaviour also increases contentious mobilization by election losers (Daxecker, 2012; Fearon, 2011; von Borzyskowski, 2019b), the net effect of election monitoring is a security-enhancing one (Luo and Rozenas, 2018). Researchers suggest similar positive effects for international technical election assistance. Weak and dysfunctional election management structures can trigger election-related violence. Technical election assistance can strengthen and professionalize election management bodies and thus prevent violence (Birch and Muchlinski, 2018; von Borzyskowski, 2016), at least if governments principally support credible elections (Lührmann, 2018).³ As PKOs’ election-related activities also include monitoring and technical election assistance, evidence for positive effects of these activities done by other international organizations should make us optimistic about the prospects of peacekeepers’ contribution to electoral peace. Yet, beyond monitoring and technical assistance, this article also examines PKOs’ unique ability to augment electoral security (see also Fjelde and Smidt, 2021).

The peacekeeping literature tends to focus on peacekeepers’ effectiveness in reducing the risk of renewed war and related violence. Researchers agree that peacekeeping deployment helps stabilize post-war environments (Beardsley, 2011; Fortna, 2008; Hultman et al., 2016). Some analyses also suggest that the deployment of PKOs during electoral periods lowers the risk of civil war recurrence after elections (e.g. Brancati and Snyder, 2013; Sisk, 2008b). Yet, this finding tells us little about whether PKOs prevent war despite elections or actually mitigate election-related violence. Recent studies have revealed the significant limitations of peacekeepers in curbing post-war violence that does not mimic the wartime master cleavage, such as crime or local power struggles (e.g. Autesserre, 2014; Bara, 2020; Di Salvatore, 2019). Thus, PKOs’ ability to reduce violence motivated by election dynamics could also be questioned. Related, Autesserre (2009) finds that PKOs generally lack the willingness and ability to address violence of lower intensity than war-fighting, including election-related violence (see also Pouligny, 2000). Instead, PKOs often view low-intensity violence as ‘innate and therefore acceptable’ in conflict-affected countries (Autesserre, 2009: 249).⁴ Thus, while the literature on election monitoring and technical election assistance provides a basis for optimism, whether peacekeepers help or hinder electoral security remains an open question (see Fjelde and Smidt, 2021 for the only other study providing cross-national evidence for PKO’s electoral security-enhancing effects).

Incentives of election-related violence

How may PKOs help reduce the risk of election-related violence in conflict-affected countries? The answer to this question starts with understanding why electoral stakeholders organize election-related violence. The main electoral stakeholders in war-torn countries are election candidates competing for local or national offices and armed groups (Daxecker and Jung, 2018) that either directly participate in elections as candidates or indirectly participate in elections by backing their preferred candidate or obstructing the campaign of candidates with incompatible policy positions (Matanock and Staniland, 2018).⁵ Following existing research, I argue that these electoral stakeholders employ election-related violence to maximize electoral gains, i.e. political power after elections (Dunning, 2011). Election-related violence is thus distinct from war-related violence owing to its motives, i.e. to affect the electoral process and outcome, and its targets: before elections, electoral stakeholders may violently target rival candidates to hamper their campaign, intimidate opposition voters to lower their turnout, and forcefully disrupt the work of election officials in competitors’ strongholds. After elections, electoral losers may target rival supporters and state agents with the aim of forcing an election re-run or entering a power-sharing agreement (i.e. von Borzyskowski, 2019a). Various parameters shape electoral stakeholders’ incentives to supplement their electoral strategies with violence (Fjelde and Höglund, 2021). I focus on those that can be tackled by PKOs’ election-related activities: (a) cost of election-related violence; and (b) likelihood of electoral manipulation.

Electoral stakeholders have greater incentives for coercive electoral tactics if violence incurs few costs. War shapes this cost parameter in ways that augment the risk of violent elections (Brancati and Snyder, 2013; Flores and Nooruddin, 2012). First, perpetrators may face low reputational costs for election-related violence because war-induced financial and logistical challenges prevent domestic ‘watchdog’ groups, such as journalists or civil society organizations, from effectively monitoring electoral processes and calling out abuse (Kumar, 1998). Even if domestic groups manage to monitor elections, political divisions in war-torn societies probably reduce the credibility of their reports (Snyder and Mansfield, 2012). Moreover, wars tend to lower the implementation costs of election-related violence. Armed conflicts produce a surplus of violence specialists (Staniland, 2015; Christensen and Utas, 2008). Since wars weaken courts and law enforcement bodies, electoral stakeholders may use these violence specialists with impunity (Höglund, 2009: 421). Wars also erode institutional constraints that could prevent governments from misusing state security forces for coercive electoral strategies (Flores and Nooruddin, 2012; Hafner-Burton et al., 2014). Yet, wars do not only increase the availability of violence-specialists. Given long-standing grievances in war-torn countries, electoral stakeholders may also find it easy to mobilize citizens for violence (Cederman et al., 2011). Overall, the low costs of election-related violence may encourage opportunistic election-related violence. Electoral stakeholders may choose to supplement their electoral strategy with coercive tactics if it is cheap to do so (see Fjelde and Smidt, 2021, for distinguishing reputational and implementation costs).

Second, electoral stakeholders use violence if a high risk of electoral manipulation makes peaceful electoral tactics seem futile for maximizing electoral gains. That is, electoral stakeholders deem violence necessary to compensate for opponents’ electoral misbehaviour and resulting disadvantages in elections (Beaulieu, 2014). Wars tend to increase the likelihood of electoral manipulation. As elaborated above, capacity constraints and divisions in war-torn society hamper effective and credible domestic monitoring of the electoral process (Flores and Nooruddin, 2012). The lack of oversight means that stakeholders might escape reputational costs for manipulating elections. In war-torn countries, constraints on implementing electoral manipulation are likely to be low too. Wars tend to erode the capacity and independence of state institutions in general (Brancati and Snyder, 2013; Flores and Nooruddin, 2012) and election management bodies specifically (Höglund, 2009). If electoral management bodies are weak or biased, governments face few obstacles to manipulating the vote (Opitz et al., 2013). Finally, missing state capacity and infrastructure may make election logistics, such as transport of election materials, vulnerable to political interferences. Overall, electoral manipulation is likely, increasing the incentives for coercive electoral strategies: electoral stakeholders are afraid of being put at a disadvantage owing to competitors’ electoral misbehaviour and they engage in violence to compensate for or protest and retaliate against these real or perceived disadvantages (Beaulieu, 2014). In manipulated elections, violence becomes a rational electoral strategy (Collier and Vicente, 2012).

If PKOs engage in election-related activities, they may augment the costs associated with election-related violence and reduce the likelihood of electoral manipulation, thereby decreasing incentives for coercive electoral strategies. Yet, there are additional determinants of violent elections, e.g. the stakes in elections (Fjelde and Höglund, 2015) or the chances of victory in future elections (Przeworski, 2015), which PKOs may not influence. Therefore, I expect peacekeepers to have a violence-reducing effect at the margins.

The impact of peacekeepers’ election-related activities

If PKOs engage in election-related activities, they may discourage electoral stakeholders from supplementing their campaigns with coercion. Election-related activities consist of two types: electoral security-related activities that increase the costs of coercion and election organization-related activities that help prevent electoral manipulation. In contrast, if PKOs do not engage in these election-related activities, their attention and resources shift away from electoral processes and they may fail to mitigate electoral stakeholders’ incentives for using coercive tactics.

Augmenting costs of election-related violence

Similar to election observer missions, PKOs’ electoral security activities entail gathering information on the electoral environment and documenting violent incidents. Yet, beyond adopting the usual functions of election observers, PKOs also set up investigation and complaint mechanisms for election-related human rights violations. For instance, during the 2011 parliamentary elections, ‘UNOCI has established an integrated human rights monitoring and investigation task force to document violations and a 24-hour call centre to receive information about alleged violations. The call centre has received more than 9,000 calls’ (United Nations Security Council, 2011). PKOs’ information on the electoral security environment and violent incidents is used for internal planning, but also shared with a wider audience. This information enters PKOs’ regular reports to the UN Secretariat, Security Council members and troop-contributing countries and it is distributed to domestic and international journalists at press conferences and in UN media. For instance, before the 2018 elections in the Democratic Republic of Congo, UN News released detailed information on government responsibility in pre-election violence (United Nations (2018).

The exposure of election-related violence is associated with reputational costs at both the international and domestic levels (see Fjelde and Smidt, 2021, for the distinction between international and domestic reputational costs). International stakeholders in elections in war-torn countries often condition their support on compliance with norms of electoral integrity. Thus, government-affiliated electoral stakeholders risk aid cuts or trade embargoes if PKO reports provide credible information on violent non-compliance. Leaders of opposition parties and armed groups probably face targeted sanctions such as frozen bank accounts. In some instances, the exposure of violence may also serve as ‘tripwire’ for more robust peacekeeping action (Fortna, 2008: 98–99). For example, after the PKO in Côte d’Ivoire called attention to government responsibility in the 2011 post-election violence, the Security Council first issued sanctions and then unanimously passed Resolution 1975, which authorized UNOCI to ‘use all necessary means’ to protect civilians – despite the fact that the defeated government called for the PKO’s departure.

In addition, peacekeepers’ exposure of violent incidents may also trigger domestic reputational costs. While election-related violence is usually intended to de-mobilize targeted voters and candidates, its exposure may provoke a popular backlash (see Fjelde and Smidt, 2021, for this mechanism). Indeed, research shows that voters often defy rather than submit to intimidation by sanctioning violent candidates and voting for more peaceful ones (e.g. Collier and Vicente, 2014; Gutierrez-Romero and LeBas, 2020). In summary, PKOs’ electoral security activities in the form of monitoring, documenting and reporting incidents of violence augment both domestic and international reputational costs of election-related violence.

Beyond mirroring the functions of election observers and imposing reputational costs, PKOs have the unique ability to engage in electoral security activities and augment the implementation costs for organizing election-related violence. Electoral security activities entail patrols at electoral sites, such as polling stations and rallies, and the re-deployment of UN forces to vulnerable electoral districts. These operations augment the capacity of UN forces to militarily retaliate against perpetrators of election-related violence (Hultman, 2013). For instance, after the 2015 elections in Côte d’Ivoire, ‘peacekeepers repelled an attack by armed individuals against the electoral office in Bonon and extricated UNOCI and Ivorian polling staff to another location’ (United Nations Security Council, 2012: 2). Often, UN military is also involved in security planning for election logistics. For instance, the PKOs’ military component in Cambodia ‘relocated a number of polling stations at last minute’ and ‘arranged the transportation of voters to new sites’ to protect the electoral process from attacks by the Khmer Rouge rebel faction (Whalan, 2013, 116). Thus, electoral security operations constrain and deter election-related violence and reduce violence specialists’ effectiveness and possibly their availability. In so doing, peacekeepers augment the implementation costs for electoral stakeholders who command election-related violence.

Moreover, PKOs’ electoral security activities increase the implementation costs of election-related violence by de-escalating ongoing electoral conflict. For example, the PKO in Côte d’Ivoire used inter-positioning and on-the-spot mediation to help restore order during violent clashes among rival partisan youth groups that had ensued after the results of the 2015 presidential elections were announced (United Nations Security Council, 2015: paragraphs 35–38). Inter-positioning and mediation thus constrain electoral stakeholders who try to exacerbate tensions and incite rioting, e.g. to mobilize supporters or intimidate opponents.

Finally, PKOs’ electoral security assistance augments implementation costs of election-related violence since peacekeepers train and supervise domestic security forces in electoral security operations. Research shows that the provision of police and military training by peacekeepers improves the professionalism of host state forces and even diffuses norms of non-violence (Belgioioso et al., 2020). Professional security forces are likely to refuse to implement coercive electoral strategies, e.g. threatening and intimidating voters and opposition candidates. PKOs’ electoral security assistance usually also entails co-location of UN military and police with domestic security forces in ‘integrated’ electoral security operations. Sometimes, PKOs even help integrate domestic non-state armed groups in electoral security arrangements (Stedman, 1997: 31). As UN force commanders supervise and obtain control over domestic forces, they create obstacles and implementation costs for electoral stakeholders that seek to organize election-related violence.

Overall, the reputational and implementation costs imposed by PKOs’ electoral security activities may reduce opportunistic election-related violence (see Fjelde and Smidt, 2021, for this argument). As the cost of coercive electoral strategies increases, their associated expected benefits decrease and peaceful campaigning becomes a relatively more beneficial tactic for maximising electoral gains.

Reducing the likelihood of electoral manipulation

Beyond securing elections, PKOs’ election-related activities also include election organization assistance, e.g. peacekeepers help reform electoral laws, advise electoral commissions and transport election materials. Through assisting the election organization, PKOs can, much like electoral observers, increase the reputational costs of electoral manipulation by augmenting the probability that electoral manipulation is detected, publicized and punished (e.g. Daxecker, 2012; Hyde and Marinov, 2014, on election monitoring). While modern PKOs are seldom tasked with election monitoring as such, their election organization-related activities allow them to gain oversight over crucial steps in the electoral process and collect information on fraudulent behaviour, e.g. during voter registration or tabulation of ballots. Through regular reports, press conferences and UN media, PKOs’ information on non-compliance with electoral rules is channelled to multiple domestic and international audiences and triggers reputational costs for the principals behind electoral manipulation. In this way, PKOs should decrease the likelihood of electoral manipulation.

Unlike election monitoring organizations, PKOs do not only observe but also assist with the election organization and constrain electoral stakeholders’ ability to manipulate the elections. Election organization assistance entails both advice and training of local and national electoral commission personnel. By providing this assistance, peacekeepers help insulate these election management bodies from undue interference by governments and other electoral stakeholders. Indeed, cross-national research suggests that such technical election assistance can improve election quality (Finkel et al., 2007). Related, PKOs tend to be involved in procurement of sensitive election material such as ballot papers and election-related technology. For instance, the PKO in East Timor organized the procurement and delivery of ballots, which had been printed in Australia on special security paper (UNTAET, 2001). In so doing, PKOs shield contentious electoral processes from political interference and manipulation.

Finally, logistical election support is a crucial part of election organization-related activities by PKOs in countries where infrastructure is wrought by war. For example, the PKO in Sierra Leone, UNAMSIL, assisted the first post-war election with ‘the transport of electoral materials and personnel, including the use of the air assets of UNAMSIL to reach areas inaccessible by road, the storage and distribution of election materials prior to the elections, the movement of ballot papers after the elections, logistical assistance to international election observers, and the use of the civilian communications facilities of UNAMSIL in the provinces’ (Atuobi, 2009: 23). In so doing, PKOs make it harder for electoral stakeholders to manipulate election logistics to their advantage, e.g. by disrupting ballot delivery or election monitoring in opponents’ strongholds. A report of the elections in Sierra Leone concludes that ‘UNAMSIL exercised such overwhelming control over the electoral process that it appeared to some citizens of Sierra Leone that the NEC [National Election Commission] had been sidelined’ (Atuobi, 2009: 23). Logistical support reduces the influence of domestic agencies and actors and may thus insulate the electoral process from political interference. For instance, Nelson (1998) argues for the elections in Haiti in 1995 that the PKO ‘UNMIH made certain that technical pieces were in place and that politicians could not use implementation problems to delay or subvert the elections for purely political reasons’.⁶

In summary, PKOs’ election organization-related activities impose constraints on fraud in the electoral process through monitoring and reporting misbehaviour and taking over crucial parts of the electoral preparation. Thus, electoral stakeholders do not have to fear that fraud puts them at a disadvantage that lowers their incentives for violent compensatory or retaliatory action. Furthermore, because PKOs constrain electoral misbehaviour, electoral stakeholders committed to clean and peaceful elections can credibly signal their intention to refrain from fraudulent activities. These signals should reduce all stakeholders’ fears of being cheated on by electoral competitors. Overall, PKOs’ election organization assistance can thus reassure electoral stakeholders and eliminate their incentives for coercive campaign strategies to protest against or compensate for real or perceived disadvantages in the electoral process (Fortna, 2008: 98–102).

PKOs without election-related functions

If PKOs do not engage in election-related activities, they may fail to shape electoral stakeholders’ incentives for election-related violence. Modern PKOs are engaged in a wide variety of activities, ranging from disarming and demobilizing ex-combatants, over monitoring borders to protecting vulnerable civilians and delivering humanitarian aid (Paris, 2004). Yet, resources of PKOs are limited. If PKOs are not involved in assisting with electoral security or organizing elections, then their attention and efforts probably shift to their other priority tasks. More specifically, electoral security activities direct peacekeepers’ attention to those actors and districts that are vulnerable to election-related violence. Without this focussed attention, PKOs probably fail to detect incidents of election-related violence and trigger reputational costs. Moreover, if UN forces are not present in hotspots of election-related violence, they are less efficient in raising the implementation costs associated with coercive electoral strategies. Furthermore, PKOs’ involvement in the election organizations is crucial for imposing costs for electoral manipulation. If PKOs remain outsiders to election management processes, they cannot trigger reputational costs for fraud. Likewise, when PKOs do not help organize the elections, they do not insulate election management bodies, procurement and electoral logistics from political interference and fail to prevent manipulation.

Overall, PKOs may only help secure elections if they engage in election-related activities. Election-related activities augment the costs of election-related violence and prevent manipulation, thereby encouraging electoral stakeholders’ peaceful electoral behaviour. Figure 1 illustrates the mechanisms underpinning this effect. While electoral security activities and election organization activities may work through analytically and operationally distinct mechanisms, studying their differential effects is difficult. This is because PKOs often support the security and the organization of elections simultaneously. Nevertheless, the analyses in this article provide exploratory evidence for both mechanisms. The main observable implication of the argument, however, pertains to both election-related activities together.

Hypothesis: If a PKO engages in election-related activities, then its presence reduces the risk of election-related violence.

Figure 1.

Mechanisms of peacekeeping in elections.

Beyond PKOs, a variety of other international actors – UN entities, regional organizations and NGOs – engage in election-related assistance and they shape electoral dynamics in important ways (Birch and Muchlinski, 2018; Matanock, 2017; von Borzyskowski, 2016, 2019a). The empirical analyses account for these efforts. In contrast to all other international actors, however, PKOs do not only provide technical assistance to the organization of elections, but also include military and police personnel and they assist with electoral security arrangements. This study explores the previously underexplored mechanism underpinning peaceful elections in challenging environments.

Research design

I examine the influence of PKOs’ election-related activities on election-related violence across (a) 630 electoral periods in conflict-affected countries holding elections in the 1989–2012 period and (b) those 92 electoral periods that are accompanied by a PKO in the same period. The full sample includes all national-level legislative and presidential elections held during or at the most 20 years after minor or major intra-state armed conflict. Two elections on the same day are counted as one. Online Appendix E shows that results are substantively the same for post-conflict elections within 10 years after the end of armed conflict. Election dates are recorded in the National Elections Across Democracies and Autocracies (NELDA) dataset version 4 (Hyde and Marinov, 2011). Information on PKOs’ deployment is taken from Kathman (2013). Conflict episodes are recorded in the UCDP-PRIO Armed Conflict Dataset version 4 (Gleditsch et al., 2002; Themner, 2013; Themner and Wallensteen, 2013).

The dependent variable is a binary indicator for whether there was significant violence involving civilian deaths immediately before, during, or after the election. Information comes from the NELDA dataset (NELDA33). For instance, deadly election-related violence recorded in NELDA occurred in the 1993 elections in Cambodia, where the Khmer Rouge murdered members of rival political parties. As the dependent variable is a binary indicator, I employ logistic regression. Cluster-robust standard errors account for potential non-independence of errors across elections within the same country.

Recently, scholars also collected events data on election-related violence, including the Electoral Violence And Contention dataset by Daxecker et al. (2019), the Countries at Risk of Electoral Violence dataset by Birch and Muchlinski (2020) and the Deadly Electoral Conflict Dataset (DECO) dataset by Fjelde and Höglund (2021). I prefer the binary measure from NELDA because it captures significant and lethal election-related violence that severely hampers the electoral process. This is the type of violence that peacekeepers ought to address. Moreover, events datasets include different types of violent events and counts of these different events may lack comparability. For example, three violent demonstrations in country A might not be similarly severe to three murders of electoral candidates in country B. The NELDA measure draws on detailed qualitative information to assess the severity of election-related violence and thus guarantees comparability at a more aggregate level.

Measuring peacekeeping activities

The analyses make use of my novel dataset on PKOs’ activities during electoral periods. The dataset draws on information from UN Secretary-General reports that describe the activities that peacekeepers implement in their host countries. I rely on these Secretary-General reports rather than PKOs’ mandates (e.g. the UN Security Council resolutions authorizing a PKO) for two reasons. First, relying on mandates alone would lead to a failure to capture peacekeepers’ activities on the ground because the activities that PKOs do can diverge from the activities that are authorized in their mandates (Blair et al., 2019). Second, reports are published more regularly than mandates, usually four times a year and sometimes weekly, for example, in the case of UNPROFOR in 1992 (UNPROFOR in Bosnia, 1992). Coding the reports before and after election day allows me to capture the election-related activities of peacekeepers during the electoral period.

The Secretary-General reports are not without limitations. They may emphasize those activities that best reflect the PKO’s achievements (Clayton et al., 2017), although there is no systematic evidence for such a bias. Reports may exclude activities that are considered minor or routine. The bureaucratic language of the reports may obscure the nature or intensity of activities that a PKO pursued in the electoral period. Notwithstanding these limitations, the Secretary-General reports usually contain specific and detailed information on election-related activities. They are the most comprehensive, publicly available source of information on PKOs’ activities in their host countries.

For example, the Secretary-General report before the 2006 elections in the Democratic Republic of the Congo reports on activities pertaining to electoral security assistance: the PKO ‘will assist in security measures for the elections using its military and police assets. The military component’s main effort during this period will be to deter armed challenges to the electoral process, contain spoilers and provide a safe environment for electoral workers and voters’ (United Nations Security Council, 2006b). An example of reported activities related to election organization assistance can be found in the Secretary-General report before the 2006 elections in Haiti. The report states that the PKO ‘will continue to provide operational assistance and policy advice to the Provisional Electoral Council to complete the current electoral cycle’ (United Nations Security Council, 2006a).

Using this dataset on PKOs’ activities, I construct two binary variables – one for the presence of a PKO without election-related activities and another one for the presence of a PKO with election-related activities. Thus, the first binary variable is 1 if a PKO without election-related activities is present during an electoral period, and 0 otherwise. The second categorical variable is 1 if a PKO with election-related activities is present during an electoral period, and 0 otherwise. Both variables are included in the analyses of all 630 electoral periods in conflict-affected countries. If both variables are 0, no PKO is present during an electoral period.

A second set of models differentiates between the two types of election-related activities – one model includes a binary variable for electoral security assistance and another model includes a binary variable for election organization assistance. Finally, a last set of models is estimated in the sample of only those 92 elections accompanied by a PKO. Thus, these models solely include the binary variable for the presence of a PKO with election-related activities, using the electoral periods accompanied by a PKO without these activities as baseline category.

Figure 2 depicts the over-time variation in elections (grey bars), in which a PKO is present (black bars) and assists with electoral security (blue bars). Figure 3 reveals a similar over-time distribution for election organization assistance. Both figures tentatively reveal an over-time increase in election-related activity by peacekeepers. Figure 4 depicts clearer time trends: the five-year mean proportion of elections accompanied by a PKO has remained fairly stable over time (black line), while the five-year mean proportion of elections with a PKO engaged in electoral security assistance (solid blue line) and/or election organization assistance (dashed blue line) has increased since 2000. Figure 5 reveals a similar upward trend in PKO troop size in this period. Indeed, large PKOs with more troops are more likely to engage in election-related activities. Yet, there are also PKOs with many troops without election-related engagement, such as UNPROFOR in Croatia (1992) or UNIFIL in Lebanon (2009). The analyses control for PKO military troops using data from Kathman (2013).⁷

Figure 2.

Electoral security assistance.

Figure 3.

Election organization assistance.

Figure 4.

Trend in election-related activity.

Figure 5.

Trend in UN troop numbers.

PKOs are not the only provider of election assistance. Other UN agencies (e.g. UNDP), international organizations, foreign states and non-governmental organizations also provide election assistance in PKO host countries and elsewhere. If these other forms of election assistance are systematically correlated with PKOs’ election-related activities, omitting them from the models may bias the effect of election-related activities on election-related violence. Therefore, all analyses account for international election monitoring and democracy aid (Matanock, 2017). An indicator records the presence of at least one credible election monitoring mission (Hyde and Marinov, 2011). A continuous variable captures committed democracy aid in constant US dollars (Tierney et al., 2011). The variable is logged owing to skewness. In Online Appendix D, the models also include a control variable for UN election assistance, using data from von Borzyskowski (2016) for the period 1990–2003 and Birch and Muchlinski (2018) for the period 2004–2012. The results for peacekeepers’ election-specific activities remain substantively the same.

Strategy for identifying peacekeeping effects

Peacekeepers and their election-related activities are not randomly assigned to countries and electoral periods. The risk of election-related violence probably influences (a) where a PKO is sent and (b) whether it assists with elections. Previous research shows that PKOs invest more resources in severe and intractable conflicts (Beardsley and Schmidt, 2012; Fortna, 2004; Gilligan and Stedman, 2003; Gilligan and Sergenti, 2008). Thus, PKOs might be more likely to deploy and conduct election-related activities in violence-prone elections in countries where conflict has been most severe and intractable. If this is the case, my analyses would underestimate the violence-reducing impact of peacekeeping deployment and election-related activities. However, the UN could also decide to refrain from assisting elections that are prone to violence, which would make it easier to find a violence-reducing effect of peacekeeping.

Therefore, all models include a wide variety of potentially confounding factors in order to account for important differences between the electoral periods in the sample. I first include armed conflict characteristics because they are likely to influence PKOs’ deployment, their level of engagement (Beardsley and Schmidt, 2012; Fortna, 2008) and election-related violence (Daxecker and Jung, 2018). These characteristics are the number of battle deaths and the number of war-related civilian casualties in the year before elections from the UCDP Battle-Related Deaths Dataset (Allansson et al., 2017) and the UCDP One-Sided Violence Dataset version 1.4 (Eck and Hultman, 2007), respectively. The models also include the duration of the previous conflict in days. All three variables are logged owing to high skewness. Furthermore, the models incorporate indicators for whether elections take place in major armed conflict (war) or after armed conflict (with minor armed conflict as baseline). Finally, days after the end of armed conflict is included to capture the time for recovery before elections (Brancati and Snyder, 2013). All of these variables are constructed from the UCDP-PRIO Armed Conflict Dataset version 4 (Gleditsch et al., 2002; Themner, 2013; Themner and Wallensteen, 2013).

Variation in election characteristics may also confound the estimated effect of PKOs’ involvement in elections. Specifically, PKOs may extend their stay and engage in election-related activities if elections pose a greater risk of violence. Using information from the NELDA dataset, the models include indicators for whether past elections experienced violence and whether elections were expected to be fraudulent (Hyde and Marinov, 2011). I also include an indicator for whether the incumbent was confident of victory as a measure for the lack of electoral competitiveness. Finally, the models incorporate an indicator for founding elections, i.e. first multi-party elections after independence or a significant period of non-democratic rule, which tend to be both violence-prone and supported by PKOs. Regime characteristics and the quality of electoral institutions similarly determine the nature of elections and, thus, peacekeeping presence and characteristics. Therefore, the models include the Polity IV score (Marshall et al., 2014) and the capacity of electoral management bodies from the V-dem project version 7.1 (Coppedge et al., 2017), respectively. To measure opportunity costs for violent behaviour, the models include the country’s GDP per capita measured in constant US dollars. Finally, the analyses include population size. More populous countries are likely to exhibit more violence and receive more assistance from PKOs. GDP and population data are logged owing to skewness and data come from the World Bank (2014). Summary statistics for all variables are in Online Appendix A.

To further address the endogenity concern, I use matching. In the absence of convincing instruments for peacekeeping deployment and election-related activities, matching is the best solution (Gilligan and Sergenti, 2008). Matching pre-processes the original sample with weights and by pruning observations to create a quasi-experimental sample where elections with the ‘treatment’ (e.g. the presence of a PKO active in election-related activities) are similar in terms of influential observable factors (matching covariates) compared with elections without the ‘treatment’. I create two post-matching samples – (a) one sample in which matching covariates are similarly distributed across electoral periods with and without a PKO and (b) another one in which matching covariates are similarly distributed across periods with and without PKOs’ election-related activities. In the second post-matching sample, the ‘control group’ thus includes electoral periods with a PKO not engaged in election-related activities and electoral periods without a PKO. Across all models, I expect the coefficient on PKOs’ election-related activities to be consistently negative. If PKOs indeed choose to deploy and assist with elections in the violence-prone electoral periods, then the negative coefficient should be larger in the more balanced, post-matching samples.

Given the moderate number of electoral periods in the analyses, I only choose matching covariates that were significant predictors of election-related violence in the pre-matching models using the original sample. These are civilian deaths, past election-related violence, fraud, competitiveness, founding elections, the capacity of electoral management institutions, democracy aid and election monitoring. Matching makes the post-matching analyses less sensitive to modelling assumptions. From the many available matching methods, I choose coarsened exact matching (CEM) and present the results after ‘optimal’ propensity score (PS) matching in Online Appendix F. CEM is used in the main analyses because it provides the best balance between ‘treated’ and ‘control’ electoral periods. Moreover, CEM does not require assumptions about the functional form of the data-generating process underlying PKOs’ election-related activities and it reduces not only the overall imbalance but also the imbalances on specific covariates. The balance statistics for the two post-matching samples have improved considerably on most matching covariates (see Tables A2 and A3 in Online Appendix A) (Ho et al., 2011: 7). Yet, even after matching, ‘treated’ electoral periods may still differ from ‘control’ electoral periods. Therefore, matching better approximates but does not recover the true effect of peacekeeping on election-related violence.

Results and discussion

The analyses of all 630 elections in war-torn countries summarized in Table 1 yield results consistent with the argument: the presence of a PKO is associated with a lower risk of election-related violence only if peacekeepers engage in election-related activities (for regression tables with all controls, see Online Appendix B). The coefficient on the presence of a PKO with election-related activities is consistently negative and statistically significant. In contrast, the coefficient on the presence of a PKO without election-related activities is positive and not consistently significant.

Table 1.

Logit regression of election violence on election-related activity.

	Model 1	Model 2	Model 3
Matching on:	No matching	PKO presence	PKO activity
PKO with election-related activities	−2.088***	−2.368**	−2.490***
	(0.700)	(1.087)	(0.932)
PKO without election-related activities	0.913*	0.894*	0.940
	(0.479)	(0.516)	(0.603)
Constant	−4.845***	−13.853***	−9.436**
	(1.516)	(3.363)	(4.008)
Observations	630	363	280
Log likelihood	−286.181	−149.726	−105.868
Akaike information criterion	612.362	339.451	251.737

Note: Cluster-robust standard errors in parentheses; *p < 0.1; **p < 0.05; ***p < 0.01.

Full model includes all controls and is summarized in Online Appendix B.

Figure 6 illustrates the predicted effects of peacekeepers with and without election-related functions on the risk of election-related violence. Based on Model 1, election-related activities are associated with a 32.3 percentage points-decrease in the average risk of election-related violence compared with electoral periods without a PKO. Models 2 and 3 are estimated using the post-matching samples. The predicted effects of election-related activities remain consistently negative. Their larger effect size implies that PKOs tend to engage in election-related activities in more violence-prone elections and that matching reduces the bias due to non-random selection. Interestingly, the presence of a PKO without election-related activities is associated with an increased risk of election-related violence. However, this change is not consistently significant and probably due to the fact that the mere presence of a PKO cannot offset the selection bias resulting from deployment to more violence-prone elections.

Figure 6.

Predicted effects (Models 1–3).

In terms of predictive performance, the addition of the variables for PKO presence with and without election-related activities makes a substantive difference. False negatives – violent electoral periods predicted as peaceful ones – are reduced from 99 to 86 cases in the pre-matching Model 1, from 29 to 20 in the post-matching Model 2 and from 21 to 17 in the post-matching Model 3. Moreover, the addition of PKO-related variables reduces the number of false positives – peaceful electoral periods predicted as violent ones – from 50 to 48 in Model 1, from 55 to 50 in Model 2 and from 37 to 30 in Model 3. Online Appendix G provides separation plots for the graphical examination of predictive performance.

PKOs’ election-related activities include assistance with electoral security arrangements and the election organization. Are both activities equally important for reducing election-related violence? While existing studies support the effectiveness of peacekeepers’ security-related functions more generally (Beardsley, 2011; Hultman et al., 2013, 2016), peacekeepers’ civilian activities (here: election organization assistance) deserve more attention in cross-national empirical research. Yet, only nine elections solely receive electoral security assistance and only seven elections solely receive assistance with the election organization in the 1989–2012 period. Thus, the results for the estimated differential effects should be interpreted with caution. Figure 7 illustrates the predicted reduction in the risk of election-related violence associated with PKOs’ electoral security assistance and election organization assistance (for regression tables, see Online Appendix B). All effects are negative and significant.⁸ What these analyses show is that peacekeepers help reduce the risk of election-related violence through both security-related and civilian functions.

Figure 7.

Predicted effects (Models 4–9).

Elections accompanied by a PKO may differ in unobserved ways from those without a PKO, for example, because peacekeepers influenced the trajectory of the previous armed conflict. Therefore, the last set of regressions tests whether the negative effect of PKOs’ election-related activities holds in a reduced sample which only includes the 92 PKO-accompanied elections. As Figure 8 illustrates, the predicted effects of election-related activities are still consistently negative and statistically significant within a 90% confidence interval.

Figure 8.

Predicted effects (Models 10-12).

Tests for alternative explanations

Election-related activities tend to be part of PKOs’ multidimensional mandates (Paris, 2004). Thus, peacekeepers’ multidimensional responsibilities rather than their election-specific tasks may explain the negative effect coefficient. Online Appendix D examines this possibility by controlling for whether a PKO has a multidimensional mandate. Yet, the effect coefficient on election-related activities remains negative and significant. Moreover, Online Appendix E shows that the impact of PKOs’ election-related activities is not driven by other UN agencies’ technical election assistance. The analyses include a binary measure for any UN election assistance, using data from von Borzyskowski (2016) and Birch and Muchlinski (2018). Nevertheless, PKOs’ involvement in securing and organizing elections remains a significant predictor of electoral peace.

Conclusion

The article shows that peacekeepers’ activities are fundamentally important for whether PKOs help keep electoral peace. PKOs’ election-related activities, their assistance with electoral security arrangements and the organization of elections are associated with a reduction in the risk of election-related violence. If PKOs do not engage in election-related activities, their presence does not affect electoral security.

Complementing existing research on peacekeeping policies (Dorussen and Gizelis, 2013; Heldt, 2011; Hultman, 2010; Murdie and Davis, 2010), my findings underline the importance of differentiating between different activities in complex multidimensional PKOs. Specifically, investigating election-related activities helps us to understand how PKOs can secure political transitions after war and nuances general critiques of liberal peacekeeping (for an overview, see Paris, 2010). Moreover, the findings add to the growing research on the impact of PKOs on violence that is often below the threshold of war and may not directly mimic the wartime cleavage (Autesserre, 2014; Bara, 2020; Di Salvatore, 2019; Fjelde and Smidt, 2021).

The optimistic findings should not deflect from the fact that PKO deployment is expensive and thus limited to a relatively short time period (Matanock, 2017). Future research should evaluate whether PKOs’ election-related activities have long-lasting benefits even after the exit of peacekeepers (Gledhill, 2020). Furthermore, it would be interesting to investigate whether the effects extend to elections at the subnational level or whether PKOs’ concurrent non-election-related activities make a difference for electoral security, too.

A final word on the scope of the findings: UN peacekeepers are deployed to countries affected by armed conflict. The findings and their policy implications pertain to these contexts. Policy-makers designing international election violence prevention strategies for conflict-affected environments should not expect all PKOs to help secure elections, but they may add PKOs with electoral security and election organization-related functions to their tool box for designing a context-specific electoral security strategy.

Supplemental Material

sj-pdf-1-cmp-10.1177_0738894220960041 – Supplemental material for Keeping electoral peace? Activities of United Nations peacekeeping operations and their effects on election-related violence

Supplemental material, sj-pdf-1-cmp-10.1177_0738894220960041 for Keeping electoral peace? Activities of United Nations peacekeeping operations and their effects on election-related violence by Hannah Smidt in Conflict Management and Peace Science

Footnotes

Acknowledgements

I am grateful to Kristin M. Bakke, Neil J. Mitchell, Andrea Ruggeri and Julian Wucherpfennig for extensive feedback on a previous version of this article. Previous versions of this article and the data collection were presented at the annual conferences of the European Political Science Association in 2020, the American Political Science Association in 2019, the International Studies Association in 2017 and 2019 and the Peace Science Society in 2015 and 2016. I would also like to thank the participants of panels at these conference, especially Ursula E. Daxecker, Hanne Fjelde, Lisa Hultman, Desirée Nilsson, Gudlaug Olafsdottir and Jun Koga Sudduth, for their helpful comments. Finally, I would like to thank three anonymous reviewers as well as the CMPS editors for their valuable feedback.

Funding

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

ORCID iD

Hannah Smidt

Supplemental material

All data, replication materials, and instructions regarding analytical materials upon which published claims rely are available online through the SAGE CMPS website:

Notes

References

Allansson

Melander

Themnér

(2017) Organized violence, 1989–2016. Journal of Peace Research 54(4): 574–587.

Atuobi

(2009) Peace support operations and post-conflict elections: The case of Sierra Leone. Kaiptc Monograph, December 2009.

Autesserre

(2009) Hobbes and the Congo: Frames, local violence, and international intervention. International Organization 63(2): 249–280.

Autesserre

(2014) Peaceland. Conflict Resolution and the Everyday Politics of International Intervention. Cambridge: Cambridge University Press.

Bara

(2020) Shifting targets: The effect of peacekeeping on postwar violence. European Journal of International Relations [online first].

Beardsley

(2011) Peacekeeping and the contagion of armed conflict. The Journal of Politics 73(4): 1051–1064.

Beardsley

Schmidt

(2012) Following the flag or following the charter? Examining the determinants of UN Involvement in international crises, 1945–2002. International Studies Quarterly 56(1): 33–49.

Beaulieu

(2014) Electoral Protest and Democracy in the Developing World. New York: Cambridge University Press.

Belgioioso

Di Salvatore

Pinckney

(2020) Tangled up in blue: The effect of UN peacekeeping on nonviolent protests in post-civil war countries. International Studies Quarterly [online first].

10.

Birch

Muchlinski

(2018) Electoral violence prevention: What works? Democratization 25(3): 385–403.

11.

Birch

Muchlinski

(2020) The dataset of countries at risk of electoral violence. Terrorism and Political Violence 32(2): 217–236.

12.

Blair

Di Salvatore

Smidt

(2019) When do UN peacekeeping operations implement their mandates? Working Paper. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3529177.

13.

Brancati

Snyder

(2013) Time to kill: The impact of election timing on postconflict stability. Journal of Conflict Resolution 57(5): 822–853.

14.

Cederman

Weidmann

Gleditsch

(2011) Horizontal inequalities and ethnonationalist civil war: A global comparison. American Political Science Review 105(3): 478–495.

15.

Christensen

Utas

(2008) Mercenaries of democracy: The ‘politricks’ of remobilized combatants in the 2007 general elections, Sierra Leone. African Affairs 107(429): 515–539.

16.

Clayton

Kathman

Beardsley

Gizelis

Olsson

Bove

Ruggeri

Zwetsloot

van der Lijn

Smit

Hultman

Dorussen

Diehl

Bosco

Goodness

(2017) The known knowns and known unknowns of peacekeeping data. International Peacekeeping 24(1): 1–62.

17.

Collier

Vicente

(2012) Violence, bribery, and fraud: The political economy of elections in Sub-Saharan Africa. Public Choice 153(1–2): 117–147.

18.

Collier

Vicente

(2014) Votes and violence: Evidence from a field experiment in Nigeria. The Economic Journal 124(574): F327–F355.

19.

Coppedge

Gerring

Lindberg

SI Skaaning

Svend-Erik Teorell

Altman

Bernhard

Fish

Glynn

Hicken

Knutsen

Marquardt

McMann

Miri

Paxton

Pemstein

Staton

Tzelgov

Wang

Zimmerman

(2017) V-Dem [all countries, 1989–2012] Dataset Version 7.1. Varieties of Democracy (V-Dem) Project.

20.

Daxecker

(2012) The cost of exposing cheating: International election monitoring, fraud, and post-election violence in Africa. Journal of Peace Research 49(4): 503–516.

21.

Daxecker

Jung

(2018) Mixing votes with violence: Election violence around the world. SAIS Review of International Affairs 38(1): 53–64.

22.

Daxecker

Amicarelli

Jung

(2019) Electoral Contention And Violence (ECAV): A new dataset. Journal of Peace Research ONLINE FIRST 56(5): 714–723.

23.

Di Salvatore

(2019) Peacekeepers against criminal violence – Unintended effects of peacekeeping operations? American Journal of Political Science 63(4): 840–858.

24.

Dorussen

Gizelis

(2013) Into the lion’s den: Local responses to UN peacekeeping. Journal of Peace Research 50(6): 691–706.

25.

Dunning

(2011) Fighting and voting: Violent conflict and electoral politics. Journal of Conflict Resolution 55(3): 327–339.

26.

Eck

Hultman

(2007) One-sided violence against civilians in war: Insights from new fatality data. Journal of Peace Research 44(2): 233–246.

27.

Fearon

(2011) Self-enforcing democracy. The Quarterly Journal of Economics 126(4): 1661–1708.

28.

Finkel

Pérez-Liñán

Seligson

(2007) The effects of U.S. foreign assistance on democracy building, 1990–2003. World Politics 59(3): 404–439.

29.

Fjelde

Höglund

(2015) Electoral institutions and electoral violence in Sub-Saharan Africa. British Journal of Political Science 46(2): 297–320.

30.

Fjelde

Höglund

(2021) Introducing the Deadly Electoral Conict Dataset (DECO). Journal of Conflict Resolution, Forthcoming.

31.

Fjelde

Smidt

(2021) Protecting the vote? Peacekeeping presence and the risk of electoral violence. British Journal of Political Science, Forthcoming.

32.

Flores

Nooruddin

(2012) The effect of elections on postconflict peace and reconstruction. The Journal of Politics 74(2): 558–570.

33.

Fortna

(2004) Does peacekeeping keep peace? International intervention and the duration of peace after civil war. International Studies Quarterly 48(2): 269–292.

34.

Fortna

(2008) Does Peacekeeping Work? Shaping Belligerents’ Choices After Civil War. Princeton, NJ: Princeton University Press.

35.

Gilligan

Sergenti

(2008) Do UN interventions cause peace? Using matching to improve causal inference. Quarterly Journal of Political Science 3(2): 89–122.

36.

Gilligan

Stedman

(2003) Where do the peacekeepers go? International Studies Review 5(4): 37–54.

37.

Gledhill

(2020) The pieces kept after peace is kept: Assessing the (post-exit) legacies of peace operations. International Peacekeeping 27(1): 1–11.

38.

Gleditsch

Wallensteen

Eriksson

Sollenberg

Strand

(2002) Armed conflict 1946–2001: A new dataset. Journal of Peace Research 39(5): 615–637.

39.

Gutiérrez-Romero

(2014) An inquiry into the use of illegal electoral practices and effects of political violence and vote-buying. Journal of Conflict Resolution 58(8): 1500–1527.

40.

Gutierrez-Romero

LeBas

(2020) Does electoral violence affect voting choice and willingness to vote? Evidence from a vignette experiment. Journal of Peace 57(1): 77–92.

41.

Hafner-Burton

Hyde

Jablonski

(2014) When do governments resort to election violence? British Journal of Political Science 44(1): 149–179.

42.

Heldt

(2011) Peacekeeping and transitions to democracy. In: Fjelde

Höglund

(eds) Building Peace, Creating Conflict? Conflictual Dimensions of Local and International Peacebuilding. Lund: Nordic Academic Press, pp. 47–71.

43.

High-Level Panel on United Nations Peacekeeping Operations (2015) Report of the High-Level Independent Panel on Peace Operations on Uniting Our Strengths for Peace: Politics, Partnership and People A/70/95 (17 June 2015).

44.

Imai

King

Stuart

(2011) MatchIt: Nonparametric preprocessing for parametric causal inference. Journal of Statistical Software 42(8): 1–43.

45.

Höglund

(2009) Electoral violence in conflict-ridden societies: concepts, causes, and consequences. Terrorism and Political Violence 21(3): 412–427.

46.

Hultman

(2010) Keeping peace or spurring violence? Unintended effects of peace operations on violence against civilians. Civil Wars 12(1–2): 29–46.

47.

Hultman

(2013) UN peace operations and protection of civilians: Cheap talk or norm implementation? Journal of Peace Research 50(1): 59–73.

48.

Hultman

Kathman

Shannon

(2013) United Nations peacekeeping and civilian protection in civil war. American Journal of Political Science 57(4): 875–891.

49.

Hultman

Kathman

Shannon

(2016) United Nations peacekeeping dynamics and the duration of post-civil conflict peace. Conflict Management and Peace Science 33(3): 231–249.

50.

Hyde

(2010) Experimenting in democracy promotion: International observers and the 2004 presidential elections in Indonesia. Perspectives on Politics 8(2): 511–527.

51.

Hyde

Marinov

(2011) Which elections can be lost? Political Analysis 20(2): 191–210.

52.

Hyde

Marinov

(2014) Information and self-enforcing democracy: The role of international election observation. International Organization 68(2): 329–359.

53.

Kathman

(2013) United Nations peacekeeping personnel commitments, 1990–2011. Conflict Management and Peace Science 30(5): 532–549.

54.

Kumar

(1998) Postconflict Elections, Democratization and International Assistance. Boulder, CO: Lynne Rienner.

55.

Leininger

(2006) Democracy and UN peace-keeping conflict resolution through state-building and democracy promotion in Haiti. Max Planck Yearbook of United Nations Law Online 10(1): 465–530. DOI:10.1163/187574106x00119.

56.

Little

(2012) Elections, fraud, and election monitoring in the shadow of revolution. Quarterly Journal of Political Science 7(3): 249–283.

57.

Lührmann

(2019) United Nations electoral assistance: More than a fig leaf? International Political Science Review 40(2): 181–196.

58.

Luo

Rozenas

(2018) The election monitor’s curse. American Journal of Political Science 62(1): 148–160.

59.

Marshall

Gurr

Jaggers

(2014) Polity IV Project. Political regime characteristics and transitions, 1800–2013. Center for Systemic Peace, 6 May.

60.

Matanock

(2017) Electing Peace. From Civil Conflict to Political Participation. Cambridge: Cambridge University Press.

61.

Matanock

Staniland

(2018) How and why armed groups participate in elections. Perspective of Politics 16(3): 710–727.

62.

Murdie

Davis

(2010) Problematic potential: The human rights consequences of peacekeeping interventions in civil wars. Human Rights Quarterly 32(1): 49–72.

63.

Mvukiyehe

Samii

(2017) Promoting democracy in fragile states: Field experimental evidence from Liberia. World Development 95: 254–267.

64.

Nelson

(1998) Haitian elections and the aftermath. In: Kumar

(ed.) Postconflict Elections, Democratization and International Assistance. Boulder, CO: Lynne Rienner, pp. 71–86.

65.

Opitz

Fjelde

Höglund

(2013) Including peace: The influence of electoral management bodies on electoral violence. Journal of Eastern African Studies 7(February): 713–731.

66.

Ottaway

(1998) Angola’s failed elections. In: Kumar

(ed.) Postconflict Elections, Democratization and International Assistance. Boulder, CO: Lynne Rienner, pp. 133–151.

67.

Paris

(2004) At War’s End: Building Peace After Civil Conflict. Cambridge: Cambridge University Press.

68.

Paris

(2010) Saving liberal peacebuilding. Review of International Studies 36(2): 337.

69.

Pouligny

(2000) Promoting democratic institutions in post-conflict societies: Giving diversity a chance. International Peacekeeping 7(3): 17–35.

70.

Przeworski

(2015) Acquiring the habit of changing governments through elections. Comparative Political Studies 48(1): 101–129.

71.

Reilly

(2008) Post-war elections: Uncertain turning points of transitions. In: Jarstad

Sisk

(eds) From War to Democracy: Dilemmas of Peacebuilding. Cambridge, MA: Cambridge University Press, pp. 157–182.

72.

Simpser

Donno

(2012) Can international election monitoring harm governance? The Journal of Politics 74(2): 501–513.

73.

Sisk

(2008a) Elections in fragile states: Between voice and violence. Paper prepared for the ISA Annual Meeting, San Francisco, CA, 24–28March 2008.

74.

Sisk

(2008b) Pathways of the political. Electoral processes after civil conflict. In: Paris

Sisk

(eds) The Dilemmas of Statebuilding: Confronting the Contradictions of Postwar Peace Operations. Abingdon: Routledge, pp. 196–223.

75.

Smidt

(2020a) Keeping electoral peace: The impact of UN peacekeeping activities on election-related violence. Journal of Peace Research 57(1): 199–216.

76.

Smidt

(2020b) United Nations peacekeeping locally: Enabling conflict resolution, reducing communal violence. Journal of Conflict Resolution 64(2–3): 344–372.

77.

Snyder

Mansfield

(2012) Democratization and civil war. In: Snyder

(ed.) Power and Progress: International Politics in Transition. New York: Routledge, pp. 175–200.

78.

Staniland

(2015) Armed groups and militarized elections. International Studies Quarterly 59(4): 694–705.

79.

Stedman

(1997) Spoiler problems in peace processes. International Security 22(2): 5–53.

80.

Themner

(2013) UCDP/PRIO Armed conflict dataset codebook version 4. Uppsala Conflict Data Program (UCDP) Centre for the Study of Civil Wars, International Peace Research Institute.

81.

Themner

Wallensteen

(2013) Armed conflicts, 1946–2012. Journal of Peace Research 50(4): 509–521.

82.

Tierney

Nielson

Hawkins

Roberts

Findley

Powers

Parks

Wilson

Hicks

(2011) More dollars than sense: Refining our knowledge of development finance using AidData. World Development 39(11): 1891–1906.

83.

Turner

Nelson

Mahling-Clark

(1998) Mozambique’s vote for democratic governance. In: Kumar

(ed.) Postconflict Elections, Democratization and International Assistance. Bouler, CO: Lynne Rienner, pp. 153–175.

84.

United Nations (2018) DR Congo elections: ‘Excessive use of force’ in campaign must be avoided, says Bachelet. Available at: https://news.un.org/en/story/2018/12/1028561.

85.

United Nations Department of Peacekeeping Operations (2008) Capstone Doctrine – United Nations Peacekeeping Operations: Principles and Guidelines. New York: United Nations.

86.

United Nations Security Council (2006a) Report of the Secretary-General on the United Nations Stabilization Mission in Haiti. S/2006/592, 28 July.

87.

United Nations Security Council (2006b) Twenty-first report of the Secretary General on the United Nations Organization Mission in the Democratic Republic of Congo. S/2006/390, 13 June.

88.

United Nations Security Council (2011) Twenty-seventh progress report of the Secretary-General on the United Nations Operation in Cote d’Ivoire. S/2011/211, 30 March.

89.

United Nations Security Council (2012) Special Report of the Secretary-General on the United Nations Operation in Côte d’Ivoire. S/2012/186, 29 March.

90.

United Nations Security Council (2015) Thirty-seventh Progress Report of the Secretary-General on the United Nations Operation in Côte d’Ivoire. S/2015/940, 8 December.

91.

UNTAET (2001) Ballot papers for upcoming elections arrive in East Timor, UN says, 20 August. Available fronm: https://peacekeeping.un.org/mission/past/etimor/news/01aug20.htm.

92.

von Borzyskowski

(2016) Resisting democracy assistance: Who seeks and receives technical election assistance? Review of International Organizations 11(2): 247–282.

93.

von Borzyskowski

(2019a) The Credibility Challenge. How Democracy Aid Influences Election Violence. Ithaca, NY: Cornell University Press.

94.

von Borzyskowski

(2019b) The risks of election observation: International condemnation and post-election violence. International Studies Quarterly 63(3): 654–667.

95.

Wantchekon

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

96.

Whalan

(2013) How Peacekeeping Operations Work. Oxford: Oxford Univeristy Press.

97.

World Bank (2014) World DataBank – World Development Indicators. Available at: http://databank.worldbank.org/data/home.aspx.

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

0.23 MB

0.19 MB