SciELO - Scientific Electronic Library Online

 
 issue38The 2018 legislative elections. About the survival of democracy in São Tomé and PríncipeThe effect of electoral violence on electoral participation in Africa author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

  • Have no similar articlesSimilars in SciELO

Share


Cadernos de Estudos Africanos

Print version ISSN 1645-3794

Cadernos de Estudos Africanos  no.38 Lisboa Dec. 2019  Epub June 11, 2021

 

DOSSIER

The Role of the Courts in Mitigating Election Violence in Nigeria

O papel dos tribunais na redução da violência eleitoral na Nigéria

Stephanie M. Burchard1 

Meshack Simati2 

1Institute for Defense Analyses, United States of America, sburchar@ida.org

2California State University San Marcos, United States of America, msimati@csusm.edu


Resumo

A violência ocorre numa ligeira maioria das eleições realizadas na África subsariana. Vários estudos recentes têm analisado a eficácia de diversas intervenções sobre a prevalência de eleições violentas; no entanto, o papel dos tribunais tem sido relativamente menos estudado. Utilizando dados sobre a violência eleitoral durante as eleições nigerianas de 2015 e 2019, examinamos em que condições os tribunais podem moderar a violência associada às eleições. Observamos que contestar com sucesso um resultado eleitoral nos tribunais está relacionado com uma redução da letalidade da violência nas eleições seguintes, mas só se os tribunais forem comummente percebidos como confiáveis. Estes resultados indicam que além de reforçar a capacidade judicial, melhorar a confiança generalizada nas instituições judiciais é fundamental para reduzir a violência eleitoral.

Palavras-chave: eleições; violência eleitoral; poder judicial; resolução de disputas; confiança institucional; Nigéria

Abstract

Violence occurs in a slight majority of elections held in sub-Saharan Africa. Several recent studies have examined the effectiveness of various interventions on the prevalence of violent elections; however, the role of the courts has been relatively less studied. Using data on electoral violence during the 2015 and 2019 Nigerian elections, we examine under what conditions courts can moderate election-related violence. We find that successfully challenging an election outcome in the courts is related to a reduction in the lethality of violence in the next election, but only if the courts are generally perceived as trustworthy. These findings indicate that in addition to building judicial capability, improving generalized trust in judicial institutions is key to reducing electoral violence.

Keywords: elections; electoral violence; judiciary; dispute resolution; institutional trust; Nigeria

African countries reintroduced1 multiparty politics in the early 1990s as the curtains came down on the Cold War era. By 1997, 75% of countries in sub-Saharan Africa had adopted multiparty elections, with the exception of Eritrea, which has not conducted presidential or national assembly elections since its independence referendum in 19932. Many sub-Saharan African countries have made significant strides towards democratic development, but election-related violence has also increasingly become a common feature of the politics of alternation of power in Africa. Thus, while elections in sub-Saharan Africa have become ubiquitous, challenges around election management, particularly related to electoral violence, still remain.

On average, a majority of elections conducted in sub-Saharan Africa result in violence. Conservative3 estimates from the Social Conflict in Africa Dataset (SCAD) (Salehyan et al., 2012) indicate that sub-Saharan African countries conducted 390 elections from1990 to 2012 in which 57 percent of the elections experienced some form of electoral violence. Of the 390 elections, 132 elections experienced pre-election violence while 91 elections also experienced post-election violence. Consistent with the SCAD dataset, Straus and Taylor’s (2012) African Electoral Violence Database (AEVD) also indicate that at least 60 percent of African elections experience electoral violence from 1990 to 2008. Burchard’s (2015) work concludes that at least 55 percent of elections conducted in Africa experienced some form of electoral violence and suggests that whereas post-election violence is less common, it tends to be more intense.

Electoral violence has been associated with various negative consequences ranging from instability to the increased likelihood of civil war (Bekoe, 2012). For instance, electoral violence experienced in Ethiopia 2005, Nigeria 2007, Kenya 2007, Zimbabwe 2008 and Ivory Coast in 2011 was deemed violent enough to destabilize each of these countries4. Similarly, electoral violence has been linked to effects on voter turnout (Bekoe & Burchard, 2017; Collier & Vicente, 2012) and the quality of democracy as assessed by potential voters (Burchard, 2015). Burchard further suggests that voters are less likely to express satisfaction with and support democracy or have trust in governing institutions if they are routinely subjected to electoral violence. Therefore, lessening election violence might help improve voter satisfaction, trust in governing institutions and overall shore up support for democracy. Such sentiments also find support in works such as Diamond (1997), which lend credence to the role of institutions in improving the quality and consolidation of democracy5.

What we know so far in the burgeoning literature on electoral violence are potential reasons for some countries provide fertile grounds for election violence both at the national or macro-level (Fjelde & Höglund, 2016; Hafner-Burton et al., 2014; Norris et al., 2015; Salehyan & Linebarger, 2015) and subnational or micro-level (de Smedt, 2009; Höglund & Piyarathne, 2009; Klaus & Mitchell, 2015 and Söderberg Kovacs & Bjarnesen, 2018). Yet, less examined, is the role of institutions in exacerbating or reducing electoral violence.

Recent research efforts, motivated to identify effective interventions to mitigate electoral violence, have focused on exploring strategies that can be employed to alleviate electoral violence; however, the role of courts as an intervention strategy remains under-researched. To this end, we conduct an empirical analysis on the effect of election petition tribunals (EPTs) on the likelihood of election violence taking place in contemporary Nigerian elections. EPTs are temporary courts specifically established by the Nigerian government to resolve disputes after an election has occurred. Nigerian elections have been particularly violent since the country’s return to civilian governance in 1999. In relation to the rest of Africa, elections in Nigeria tend to be more violent than the average case, with significant numbers of physical injuries and fatalities. Fatalities have occurred before and after elections in 1999, 2003, 2007, 2011, 2015, and 2019.

Elections by their very nature are contentious and it seems reasonable to assume that a dedicated mechanism put in place to adjudicate election-related grievances and suspected violations could reduce incentives to engage in electoral violence. If electoral actors believe that irregularities can be fairly challenged in an impartial venue, they may be less likely to resort to violence to win. In situations where electoral actors believe that there are no viable venues to resolve problematic elections, political actors may be more likely to turn to intimidation, harassment, and physical attacks to win elections. Furthermore, if political actors believe that an independent court system will hold them accountable for electoral infractions, they may be less likely to engage in fraud and violence. If no such judicial avenue exists, the inverse may be true.

In this paper, we test the efficacy of the judiciary in mitigating electoral violence with the caveat that for the courts to have a suppressive effect on electoral violence, political actors must have trust in the system. We argue that if those aggrieved during the electoral process, specifically political elites, lack trust in the dispute resolution mechanism, the expected utility of the courts as an alternative to violence for resolving election disputes will remain low. We argue that, particularly in Nigeria, the courts can provide an alternative to violence in disputed elections and reduce the prevalence of electoral violence, but only if the courts are perceived as trustworthy. The paper proceeds as follows. We start with an overview of electoral violence in sub-Saharan Africa. We follow with the theoretical argument on why courts can reduce electoral violence. We then conduct an empirical test on the effects of the courts on electoral violence in Nigeria’s 2015 and 2019 elections. We analyze and discuss the findings in the section thereafter then conclude with how and why courts matter in reducing electoral violence and consolidating democracy.

Overview of electoral violence in sub-Saharan Africa

Electoral violence is not a new phenomenon nor is it idiosyncratic to Africa. Countries in the early throes of democratic transition often face election-related violence as citizens seek to exercise their freedoms within an extended franchise. Seymour and Frary (1918) present an early example of violent electoral tactics in the 20th century positing that Magyars in Hungary engaged in non-Magyar voter suppression by destroying bridges and quarantining horses in the outlying villages that would be used to get the non-Magyars to polling stations.

More recent research has argued for disaggregating electoral violence based on timing, strategy and actors to better understand the motivation behind violence. On the timing of electoral violence, scholars have argued that election violence essentially falls into three phases with each phase ideally driven by different theoretical motivations and causal processes. Both Höglund (2009) and Daxecker (2014) suggest that the motivation behind pre-election and election day violence is to affect the outcome of the elections or disenfranchise voters that would otherwise help tip the election in favor of an opponent. On the other hand, post-election violence is employed as a response to the outcome of an election deemed fraudulent or in order to force the hand of other actors like courts or election commissions to recount the ballots or cancel the results and call fresh elections. Essentially, whereas pre-election and election day violence is theoretically intended to increase the chances of winning for the violence entrepreneur by suppressing voter turnout for the opposing party election, post-election violence happens as a reaction to a contested election outcome.

To this end, although the timing of elections can be categorized into three distinct phases of pre-electoral, election day and post- electoral violence, the motivations for electoral violence map onto two broad categories of pre-6 and post-election7 violence. As we have suggested above, the motivations for pre-electoral violence are to primarily influence the outcome of the election whereas post-electoral violence is a reaction to the election outcome. Subsequently, we conceptualize electoral violence into two phases of pre-electoral violence and post-electoral violence. Due to the messiness of separating election violence that happens six months before election day and violence that happens on election day, we stay consistent with other researchers (Burchard, 2015; Straus & Taylor, 2012) who conceptualize pre-electoral violence as combining both phases since both phases have similar theoretical motivations for violence. Similarly, electoral violence can be strategically employed to suppress voter turnout or mobilize supporters, or it can occur incidentally due to heightened competition between different political camps (Bekoe & Burchard, 2017). At the same time, both state actors and opposition groups can engage in electoral violence strategically to influence the turnout and outcome of elections, although state actors are much more likely to engage in electoral violence. Taylor et al. (2017) estimate that upwards of 80 percent of electoral violence is perpetrated by the state.

Prevention of electoral violence

The general understanding that electoral violence can lead to prolonged conflict (Bekoe, 2012), lower satisfaction with and support for democracy (Burchard, 2015), and democratic stagnation (Bogaards, 2013) due to poor quality elections has led to a budding research whose efforts are directed at exploring election violence prevention. These efforts are informed by the theoretical argument that political exclusion is the main driver for electoral violence and political inclusivity might lower the likelihood of electoral violence. To this end, electoral violence prevention strategies have been based around interventions such as capacity building (Birch & Muchlinski, 2018; Claes & von Borzyskowski, 2018; Darnolf & Cyllah, 2014), attitude transformation (Birch & Muchlinski, 2018; Finkel, 2014; Fischer, 2017; Höglund & Jarstad, 2011), security planning (Claes, 2016; Claes & von Borzyskowski, 2018), and election monitoring (Asunka et al., 2019; Claes, 2016; Claes & von Borzyskowski, 2018).

Proponents of capacity building suggest that building technical skills in entities entrusted with conducting elections, for instance, election management bodies, to help conduct credible elections can reduce electoral violence by increasing institutional trust, the quality of elections, ensure the acceptability of the election outcome for involved parties and, subsequently, legitimates the winner. Attitude transformation interventions are also deemed to reduce electoral violence by targeting grievances that would otherwise foment and lead to electoral violence. These interventions can include educating voters, peace messaging, mediation, engaging the youth, roundtable discussions, peace pledging, and codes of conduct. Similarly, as Claes and von Borzyskowski (2018) find, security engagement by way of the police having a closer working relationship with local communities is associated with lower levels of electoral violence as evidenced by Kenya and Liberian elections. The evidence regarding the impact of election monitoring interventions, on the other hand, is very mixed. While some have found that the presence of election observers can decrease the likelihood of electoral violence from breaking out under certain circumstances (see Smidt, 2016), recent research suggests that election observation can exacerbate electoral violence instead of mitigating it by incentivizing political actors to strategically shift violence to the pre-election period not observed by election monitors (see Daxecker, 2014; Luo & Rozenas, 2018; von Borzyskowski, 2019).

Although the above research on interventions to mitigate electoral violence is instrumental and groundbreaking, it tends to focus on programmatic efforts that can be undertaken in the pre-election period to avoid electoral violence during or after elections. Subsequently, the gap we find in the literature so far is the lack of scholarly work on the role of domestic courts as key intervention actors by virtue of their dispute resolution role. Therefore, we are examining what exactly is the role of courts in mitigating electoral violence.

Ideally, the courts would alleviate electoral violence by providing a dispute resolution mechanism through which aggrieved political contenders can seek redress. In other words, whereas interventions around capacity building, attitude transformation, and security planning can pacify the eruption of electoral violence, they lack a dispute resolution component that political actors can revert to if they perceive that they have lost an election unfairly. It is against this backdrop that countries like Nigeria have tried to address election-related disputes by creating election petition tribunals (EPTs) at the federal and state level for legislative and gubernatorial election petitions that deal specifically with cases stemming from electoral fraud, irregularities, and other illegalities. Established by the 1999 Constitution, and amended in 2010, the specific statute tasks tribunals with determining whether an individual has been validly elected (Library of Congress, 2011). Consequently, we are interested in understanding if an impartial and independent court, as measured by trust in judiciary and EPT ruling, has any effect on electoral violence. Because in Nigeria there is variation in both the occurrence of electoral violence and public trust in the courts across the country’s 36 states, Nigeria is an ideal candidate to examine how judicial processes could impact the frequency and/or intensity of electoral violence.

Theory: judicial systems and electoral violence

The constitutional function of courts as arbiters of domestic disputes predisposes them to being viewed as one of the first institutions political actors might seek intervention from during the electoral period. Ideally, the courts would provide an alternative for resolving election disputes because political actors would seek redress from the courts instead of resorting to violence. Additionally, a robust court system may act as a deterrent for violence if political actors believe they will be punished for the use of violence during an election campaign, but no research at this point exists to help us understand the effect of courts on electoral violence.

What we know so far is that judicial independence in Africa varies widely and different countries have taken different pathways to achieve judicial independence. Works such as Widner (2001), VonDoepp (2006), VonDoepp and Ellet (2011), Ellet (2013), and Prempeh (2017), although unrelated to the effectiveness of judiciaries in reducing electoral violence, have spoken to the evolution of the independence in African judiciaries. While Widner (2001) attributes the rise of judicial independence since the end of colonialism to the role of judges within the judiciary fighting for the independence of courts, Prempeh (2017) offers cautious optimism that the judicial constitutional revival in Africa has been curtailed by imperial presidencies that are too strong to be checked by weak legislatures. VonDoepp (2006) and VonDoepp and Ellet (2011) examine factors that might affect judicial assertiveness suggesting that the security of tenure by the incumbent, extent of power concentration in one actor, and the level of interest in a given case by political actors will determine whether judges will choose to be assertive or exercise strategic self-restraint. In analysis of Uganda, Tanzania and Malawi, Ellet (2013) finds that judicial independence was higher in countries that had a robust civil society, a strong independent media, and a strong law society.

We also know that, at least for Latin American countries, effectively functioning judiciaries are important for maintaining democratic gains. Larkins (1996) opines that an institutionalized judiciary is integral to democratic consolidation because it is the one institution tasked with enforcing the rule of law both on citizens and governments from exceeding their reach. To this end, various researchers are in agreement (Chavez, 2004; Diamond, 1997; Larkins, 1996; Utter & Lundsgaard, 1993) that strong judiciaries play a critical role in consolidating democracy; however, the effectiveness of judiciaries in Africa still remains under-researched.

To this end, research conducted by Abul-Ethem (2002) and Widner (2001), for example, point to the shortcomings of African judiciaries to safeguard the rule of law, while Kaaba (2015) suggest that African judiciaries have been complicit in the failure to help consolidate democracy in Africa. For instance, Widner (2001) suggests that competing centers of power and the weight of colonial legacies effectively delegitimized the African judicial systems by understaffing them while expecting them to blend customary and common law. Handicapped as such, African judiciaries perennially suffered from a lack of legitimacy by overpromising but failing to deliver justice. The lack of legitimacy by African judiciaries can be seen by the perennially low trust in judiciaries among African citizens. Based on survey data from 2014 and 2015, only a slight majority of respondents across 36 African countries stated that they had confidence in the courts, and around one-third stated that “most” or “all” judges are corrupt (Logan, 2017). The lack of trust in the judiciary is compounded by African court’s proclivity for consistently ruling in favor of the incumbent in presidential election petitions (Kaaba, 2015), with the rare exception being Kenya in 2017.

As poignant as these works are, they fail to interrogate the extent to which judiciaries have been effective in reducing violence or forestalling instability. This is partly due to the lack of data on judicial decisions and the long processes the courts take before making pronouncements on cases. For instance, Simati’s (2018) dissertation work finds that it took an average of two years for the courts to pronounce themselves on election petitions filed against President Moi’s election victories in 1992 and 1997. There is also a consensus that institutions responsible for dispute resolution in unconsolidated democracies have become increasingly important. Widner (2001) posits that when courts adjudicate election-related disputes, they forestall the potential for violence and instability between contestants and negate the attraction of these groups from resorting to pressing their grievances through their social groups. Ostensibly, during elections, those that feel aggrieved by the electoral process but have less trust in the judiciary are more likely to resort to other means of dispute resolution, for instance violence. Essentially, by providing an avenue through which electoral disputes can be addressed, courts would ideally reduce the utility of reverting to electoral violence during elections by political actors.

Nevertheless, the mere presence of courts will not incentivize political actors to choose courts to resolve their election disputes instead of using violence. Political actors are more likely to take their election-related grievances to court if they have trust that the court will be impartial. For instance, while looking at the behavior of the Ugandan courts regarding presidential and parliamentary election petitions, Murison (2013) finds that whereas the Ugandan High Court upheld election petitions and nullified various parliamentary elections, the Court of Appeal overturned most of the high court decisions. Similarly, although the Uganda Supreme Court acknowledged irregularities in the presidential elections won by Museveni in 2001 and 2006, it was unwilling to rule in favor of the opposition candidate and nullify the presidential elections. Under such circumstances, the opposition is less likely to lodge their complaints with the courts due to perceived lack of impartiality and trust. Consequently, they are more likely to resort to electoral violence.

Similar research looking at the effect of courts on electoral violence from a dissertation by Simati (2018) conducted cross nationally on African countries between 1990-2012 finds that variation in judicial independence can influence both state actors and the opposition’s choice to employ post-election violence. The author finds that both the incumbent and the opposition are less likely to employ post-election violence in African countries that have independent or dependent judiciaries. Under independent judiciaries, both actors are more likely to bring their election related grievance to court due to trust in the impartiality of the court system. Similarly, under dependent judiciaries controlled by the incumbent, the opposition is less likely to strategically employ electoral violence to influence the judges because the judges are less likely to strategically defect and rule in favor of the opposition. The author argues that post-election violence is particularly prevalent under semi-independent judiciaries precisely because opposition candidates can strategically use electoral violence to create levels of political uncertainty in order to influence judges to strategically defect or be assertive. Strategic use of violence is especially useful for the opposition if they deem the incumbent weak either due to a minority in the legislature8 or a small win margin.

Since courts can mitigate both pre-election and post-election violence by providing an alternative to election violence, we theorize that a trusted dispute resolution mechanism can alleviate electoral violence. However, in the absence of a trusted court system, political actors are more likely to revert to electoral violence either to influence the outcome of an election or in reaction to a concluded election. We therefore argue that increased trust in courts will be correlated with lower levels of electoral violence while low trust in courts will be associated with higher levels of electoral violence. Similarly, we argue that successful nullification of election results by the EPTs will have a constraining effect on electoral violence. Due to data limitations, we only conduct our analysis on pre-election violence9, not post-election violence.

Data and variables

We seek to test this theory of court effects on electoral violence on the pre-election period in Nigeria during the 2015 and 2019 elections based on the independent variable operationalized as the EPT decisions from the 2011 and 2015 elections, respectively. We conduct this analysis on Nigeria because the country’s history with the systematic violence that routinely breaks out during elections makes it a good candidate for this type of empirical analysis. More importantly, Nigeria is the only African country with specific EPTs set up precisely to deal with election petitions and pre- and post-electoral disputes. A similar court exists in South Africa, but as a standing body that is not reconstituted every election period. Other African countries adjudicate electoral related disputes within the existing judicial framework, which takes longer for the petitions to be resolved. We also examine changes over time in Nigeria as we believe that experience with the courts in one election can impact behavior in subsequent elections.

After independence in 1960, Nigeria was largely ruled by successive military governments, with brief interruptions for episodic civilian rule. In 1999, Nigeria transitioned to a civilian-led government and conducted its first multiparty elections of its Fourth Republic. Violence has been a common feature of every election since 1999, although the intensity, timing and type of violence have varied across different elections. For instance, Nigeria experienced significant levels of pre-electoral violence with an estimated 300 fatalities and 100 fatalities before the 2007 and 2015 elections respectively (Bekoe, 2012). However, the 2011 elections experienced extreme post-election violence with around 800 fatalities taking place after the results were announced (Bekoe, 2012).

Electoral violence in Nigeria has varied across both time and space. Nigeria is a federal republic with 36 states and the Federal Capital Territory, located in Abuja. Certain states such as Lagos, Rivers, and Delta State have seen more of the brunt of electoral violence. During the 2015 election, violence was concentrated in the southwest of the country, with some fatalities also occurring in the Middle Belt (Harwood, 2019). During the 2019 elections, much of the violence took place in southern Nigeria and the Middle Belt again (Harwood, 2019).

For the purposes of our analysis, we use data on violent events from the Armed Conflict Location and Event Dataset (ACLED) to code the dependent variable, pre-election violence events. ACLED collects information on political violence by type, time, intensity, and location. We identify election-related events and the number of associated fatalities, if any, and aggregate them at the state level for elections in 2015 and 2019. Since Nigeria typically holds general elections for all elected offices over a two-week period, we collected information on pre-electoral violence associated with the presidential, assembly, senate, state house, and gubernatorial races. Similar to other researchers, we confine our analysis of pre- electoral violence to the six-month period prior to the elections (including election day) (Burchard, 2015; Straus & Taylor, 2012). After a six-week postponement to allow the government to address insecurity in the North East related to the Boko Haram insurgency, elections in 2015 were held on March 28th for President and National Assembly and April 11th for governor and the state houses. Elections in 2019 were delayed one week due to logistical challenges and ultimately held on February 23rd for President and National Assembly and March 9 for governor and state houses.

Based on data from ACLED, pre-election violence took place in most Nigerian states prior to the 2015 and 2019 general elections at roughly similar rates. Across Nigerian states10, only five states (Abia, Ekiti, Kogi, Niger, and Zamfara) did not have a recorded instance of pre-election violence prior to the 2015 election and only two states (Gombe and Kebbi) did not have a recorded instance prior to the 2019 election. Ahead of the 2015 elections, there was an average of 4.8 violent events per state. Post-election violence was significantly lower in 2015 than in 2011. After the 2015 elections, there was an average of 0.79 violent events per state and an estimated 6 fatalities. Prior to the 2019 elections, there was an average of 4.6 violent events per state. There were at least 98 fatalities across 20 states in 2015 and 91 fatalities across 19 states in 201911. Eighteen states experienced fewer violent incidents in 2019 compared to 2015 while 14 states experienced more violent incidents. In terms of fatalities, 12 states experienced fewer election-related fatalities in 2019 compared to 2015 where 12 states experienced more fatalities.

According to Sani (2015), the Nigerian government established election petition tribunals in the 1999 Constitution to resolve electoral disputes in a timely fashion due to delays experienced through the regular court system in previous petitions. Through the 2000s, however, EPTs were still not known for their speed. For example, it took 34 months to resolve a 2003 gubernatorial dispute. Similarly, the success rate of the courts ability to mitigate election petitions has been low. Omenma et al. (2017) find that the courts in Nigeria have done little to build confidence among voters on the courts ability to overturn elections even in the context of overwhelming evidence of fraud. The authors find that between 1999 and 2011, 160 gubernatorial elections were challenged in court, but the court overturned only 10 elections, which translates to a 6.3 percent success rate.

As such, the 2006 electoral act also introduced a raft of changes aimed at improving efficiency in adjudication of election petitions by empowering any state or federal High Court to address all pre-election disputes. The 2010 Amendment to the Constitution requires that all election petitions be submitted within 21 days after an election has been decided; that EPTs deliver their final written verdict within 180 days; and that all appeals are heard and adjudicated within 60 days of the initial judgment.

EPTs vary by level of election and institution. EPTs at the state level hear cases from the National Assembly petitions whose appeal terminates at the Court of Appeals. The presidential and gubernatorial EPTs appeals terminate at the Supreme Court. Only candidates or parties directly participating in an election can file a petition. A petition must be based on at least one of the following four criteria: a candidate was not qualified to contest the election; the election was marred by corruption or non-compliance with the Electoral Act (as amended in 2010); the winner did not actually receive a majority of the votes; and/or the petitioner was unlawfully prevented from participating in the election. Temporary EPT appointments are managed by the President of the Court of Appeals, National Judicial Commission, and Federal Judicial Service Commission. EPTs comprise of three to five justices from the High Court, Sharia Court of Appeal, Customary Court of Appeal, or other judiciary members at the rank of chief magistrate or higher. See Table 1 for an overview of the number of petitions filed per election since 2003 (PLACNG, 2017).

Table 1 Election petitions by year 

Election year Petitions filed
2003 560
2007 1290
2011 732
2015 ~600

To code the independent variable (election petitions filed at the state level), we collected data from a number of sources including the National Democratic Institution, a non-profit organization located in Washington D.C., and the Policy and Legal Advocacy Center, a non-governmental organization located in Abuja. For the 2011 and 2015 elections, in particular, we were able to track how election petitions were ultimately resolved. From this data, we created our independent variables on EPT decisions for the 2011 and 2015 elections. Since elite-level choice to use the court can be informed by previous court decisions, we create the tribunal election nullification variable by coding the petitions that were successfully overturned by the EPTs in 2011 as a 1, representing a nullified election, and 0, which represents an upheld election. We do the same for 2015 EPT decisions and create a dummy variable where 1 represents at least one successful EPT challenge at the state level and 0 represents no EPT successes. Across all elective offices in 2015, there were a total of around 600 election petition challenges filed in 2015 (PLACNG, 2017)12. The range was from 2 to 49 election petitions challenges per state. There were a total of 333 appeals filed, with a range from 1 to 46 per state. Approximately 15 percent (84) of the total challenges at either the tribunal, appeals, or High Court stage were successful in overturning the election result or requiring either a partial or complete re-run of the election. Delta State in particular was forced to re-run several elections for both state and federal offices in March 2016 after the 2015 results were annulled due to corruption and violence. This marks a substantial increase in the success rates of cases before tribunals. In 2007, only around 2 percent of all cases were successful in annulling the results or forcing new elections (Sani, 2015). This figure increased to 4 percent after the 2011 elections (Sani, 2015).

To calculate trust in the courts, we used Afrobarometer survey data. We created an index measuring trust in the courts for respondents using rounds 6 and 7 of the Afrobarometer survey, conducted in 2014 and 2017 before the 2015 and 2019 elections, respectively. The index ranges from low to high trust, captured numerically between 1 and 3, where 1 indicates low trust, 2 some trust, and 3 high trust13.

We control for the total population of the state at the last census (logged due to its right skew) as some have found that localities with higher populations have higher incidences of electoral violence (see Daxecker, 2012, 2014) and trust in the national elections commission (NEC)14. We also control for the intensity of pre-election violence by including a variable for number of pre-election fatalities that took place in a given state in the six-month period prior to an election15. We believe that more violent elections could have a negative influence on the decision of political actors and dilute the role of trust in encouraging actors to turn to the courts16. Our primary independent variable is informed by our theory that courts give election political players a less violent avenue through which they can adjudicate election-related disputes. We use the independent variable, trust in courts, to gauge the likelihood of political players’ choice to take their case to the courts as opposed to the streets. Subsequently, high trust in court indicates the political players’ belief in the impartiality in the court and vice versa. We include the above control variables because previous research suggests they help explain the occurrence of electoral violence. We include them in the model in order to rule out that these control variables are the ones driving the relationship between our dependent and independent variables. Since our analysis only covers three election cycles in the same country, we do not control for electoral system17 because there is no variation in electoral rules that would influence electoral violence.

Discussion and analysis

For the empirical analysis, we use a negative binomial regression to test whether the nullification of elections results by EPTs and trust in court have a constraining effect on pre-election violence at the state level.18 We test this hypothesis for the 2015 elections in model 1 and model 2 for 2019 elections while controlling for logged population, trust in the national election commission, and pre-election fatalities (See table 2).

Table 2 Dependent variable: Number of pre-election violence events 

Model 1: 2015 Model 2: 2019
Election overturned -0.504** (0.266) -0.057 (0.228)
Trust in court
Some trust -0.491 (0.477) -0.750*** (0.272)
High trust -0.775* (0.434) -0.878*** (0.264)
Trust in NEC -0.010 (0.009)
Pre-election fatalities 0.186*** (0.034) 0.094*** (0.019)
Logged population -0.049 (0.357) -0.137 (0.252)
Constant 5.241 (0.357) 3.616 (3.809)
Observations 33 33
Model 1: 2015 Model 2: 2019
Election overturned -0.504** (0.266) -0.057 (0.228)
Trust in court
Some trust -0.491 (0.477) -0.750*** (0.272)
High trust -0.775* (0.434) -0.878*** (0.264)
Trust in NEC -0.010 (0.009)
Pre-election fatalities 0.186*** (0.034) 0.094*** (0.019)
Logged population -0.049 (0.357) -0.137 (0.252)
Constant 5.241 (0.357) 3.616 (3.809)
Observations 33 33

Notes: *p≤0.10, *p≤0.05, ***p≤0.01: Coefficients listed with standard errors in parentheses. Negative binomial regression used.

For model 1, we find support that overturning an election outcome by an election tribunal and high trust in courts reduce the likelihood of pre-election violence in the next election. The election overturn by tribunal variable has a negative coefficient (-0.504) and is statistically significant with a (0.050) p-value. The high trust in court variable is also statistically significant with a negative coefficient (-0.775) and p-value (0.07). The control variable for pre-election fatalities is also statistically significant with a positive coefficient (0.186) and p-value (0.00) indicating that intensity of violence can lead to more violence, possibly retributive. Both controls for population and trust in the national election commission are not statistically significant. Our findings suggest that all else being equal, favorable rulings for the challenger by the EPT in the 2011 elections and trust in court prior to 2015 were important factors in reducing the possibility of pre-election violence during the 2015 elections in Nigeria.

In model 2, we run the same analysis for the 2019 elections, modeling the effect of EPT decisions and trust in the courts on pre-election violence at the state level. We use successful nullification of an election in 2015 and Afrobarometer’s trust in courts from round 7 (conducted in 2017 before the 2019 elections) as our main independent variables. We control for population and 2019 pre-election fatalities while excluding trust in the national election commission because the variable was not available for 2019. We find that having trust in the court significantly reduced the likelihood for pre-election violence in the 2019 elections. While controlling for population and pre-election fatalities, we find that having some trust in the court was statistically significant with a coefficient of 0.750 and a p-value of 0.006. Similarly, having high trust in the court was also statistically significant with a coefficient of 0.878 and a p-value of 0.001. However, the successful nullification of elections by EPTs in 2015 did not have any significant effect on the pre-election violence experienced during the 2019 elections. Consistent with 2015 elections, pre-election fatalities experienced prior to the 2019 elections also influenced the prevalence of pre-election violence.

Unlike an ordinary least squares model where we can interpret the substantive effects of the independent variable on the dependent variable based on coefficients, the coefficients on the negative binomial regression, as a maximum likelihood model, help indicate the direction of the effect but less so in interpreting the substantive effect of the EPT decisions and trust in courts on pre-election violence. To this end, we use incident rates ratio (IRR) to explain the substantive effect of EPT decisions and trust in courts on electoral violence for models 1 and 2. The results are presented in Table 3.

Table 3 Incidence rates ratio for pre-election violence for 2015 & 2019 by state 

2015 2019
IRR Std. error IRR Std. error
Election overturned 0.604 0.160 - -
Some trust in court - - 0.472 0.129
High trust in court 0.460 0.199 0.415 0.109
Pre-election fatalities 1.204 0.040 1.099 0.022

Notes: We estimate the incidence ratio for pre-election violence for the statistically significant variables found in Table 2. Some trust in the courts (2015) and election overturned (in 2019) are missing incidence ratio estimates in Table 3 because they were not statistically significant in the models reported in Table 2.

Based on the IRR estimated for 2015, every election result that was overturned by an EPT in 2011 reduced the incidence of pre-election violence by a factor of 0.60, holding other variables constant. Comparing low trust and high trust in courts, we find that having high trust in courts reduced the incidence of pre-election violence by a factor of 0.46 in comparison with the reference category of low trust. Pre-election fatalities also increased the incidence of pre-election violence by 1.20. In 2019, having some trust and a lot of trust in the courts reduced the incidence of pre-election violence, by 0.47 and 0.42 respectively, while pre-election fatalities increased incidences of pre-election violence by factor of 1.09.

These findings, although limited to elections in Nigeria, suggest that the role of courts and trust in dispute resolution mechanisms can be instrumental in alleviating pre-election violence. We find that if an EPT overturns an election in favor of the petitioner, the likelihood of pre-election violence in a subsequent election reduces by 60 percent. Similarly, the number of election petitions filed in 2015 was significantly related to the disposition of petitions associated with the previous 2011 election. The number of petitions filed was significantly higher in states where a tribunal overturned previous election results (average: 21.5, n=14 states) in 2011, than in states where no elections were overturned in 2011 (average: 12.9, n=19 states). This suggests that a successful experience with the tribunal system might encourage others to turn to tribunals to resolve disputes in future elections. Most importantly, trust in courts is a good indicator of whether election related disputes will be resolved through the court or through violence. We find that on average, incidences of pre-election violence reduce by 45 percent if citizens have some trust or a lot of trust in courts.

Conclusion

Recent efforts targeted at interventions for preventing electoral violence have emphasized election monitoring, security planning, attitude transformation and capacity building, but the courts as a key domestic component responsible for dispute resolution still remains under-researched. This work contributes to the ongoing research on electoral violence prevention by evaluating the role election petition tribunals and citizens trust in courts play in mitigating electoral violence. We find that all else being equal, rulings that nullify election results and trust in the judicial system constrain the prevalence of electoral violence.

Although these findings may or may not be generalizable to sub-Saharan Africa and the rest of the democratizing countries around the world due to the fact that Nigerian elections are more violent than average, we hope that at the very least they should stimulate future research efforts regarding the role of the judiciary in mitigating electoral violence. From a policy implication perspective, there might be value setting up ad hoc election petition tribunals and building capacity in domestic judiciaries to be efficient in adjudicating election-related disputes in a shorter time frame. Work could also be undertaken to build trust in the courts by citizens. Both factors are important in reducing incentives for election-related violence. Due to data constrains, the focus of this research was on the pre-election period. Future work can look at whether the same relationship holds for the post-election period.

References

Abul-Ethem, F. (2002). The role of the judiciary in the protection of human rights and development: A Middle Eastern perspective. Fordham International Law Journal, 26(3), 761-770. [ Links ]

Asunka, J., Brierley, S., Golden, M. A., Kramon, E., & Ofosu, G. (2019). Electoral fraud or violence: The effect of observers on party manipulation strategies. British Journal of Political Science, 49(1), 2019, 129-151 (published online in 2017 by Cambridge University Press). https://doi.org/10.1017/S0007123416000491 [ Links ]

Bekoe, D. (2012). Voting in fear: Electoral violence in sub-Saharan Africa. United States Institute of Peace Press. [ Links ]

Bekoe, D., & Burchard, S. (2017). The contradictions of pre-election violence: The effects of violence on voter turnout in sub-Saharan Africa. African Studies Review, 60(2), 73-92. [ Links ]

Bellows, J., & Miguel, E. (2006). War and institutions: New evidence from Sierra Leone. The American Economic Review, 96(2), 394-399. [ Links ]

Birch, S., & Muchlinski, D. (2018). Electoral violence prevention: What works? Democratization, 25(3), 385-403. https://doi.org/10.1080/13510347.2017.1365841 [ Links ]

Bogaards, M. (2013). Exchange: Reexamining African elections. Journal of Democracy, 24(4), 151-160. [ Links ]

Borzyskowski, I. von. (2019). The credibility challenge: How democracy aid influence election violence. Cornell University Press. [ Links ]

Burchard, S. M. (2015). Electoral violence in sub-Saharan Africa: Causes and consequences. First Forum Press, Lynne Rienner. [ Links ]

Chavez, R. B. (2004). The evolution of judicial autonomy in Argentina: Establishing the rule of law in an ultrapresidential system. Journal of Latin American Studies, 36(3), 451-478. [ Links ]

Claes, J. (Ed.). (2016). Electing peace: Violence prevention and impact at the polls. United States Institute of Peace. [ Links ]

Claes, J., & Borzyskowski, I. von. (2018). What works in preventing election violence: Evidence from Liberia and Kenya. United States Institute of Peace. [ Links ]

Collier, P., & Vicente, P. (2012). Violence, bribery, and fraud: The political economy of elections in sub-Saharan Africa. Public Choice, 153(1), 117-147. [ Links ]

Darnolf, S., & Cyllah, A. (2014). EMBs and electoral violence. In A. Cyllah (Ed.), Elections worth dying for? A selection of case studies from Africa (pp. 11-27). International Foundation for Electoral Systems. [ Links ]

Daxecker, U. (2012). The cost of exposing cheating: International election monitoring, fraud, and postelection violence in Africa. Journal of Peace Research, 49(4), 503-516. [ Links ]

Daxecker, U. (2014). All quiet on election day? International election observation and incentives for pre-election violence in African elections. Electoral Studies, 34, pp. 232-243. [ Links ]

Demarest, L., & Langer, A. (2018). The study of violence and social unrest in Africa: A comparative analysis of three conflict event datasets. African Affairs, 117(467), 310-325. [ Links ]

Diamond, L. (1997). Consolidating democracy in the Americas. Annals of the American Academy of Political and Social Sciences, 550(34), 12-41. [ Links ]

Dunaiski, M. R. (2015). Institutions and electoral violence. Unpublished thesis. https://www.sciencespo.fr/psia/sites/sciencespo.fr.psia/files/Dunaiski%20-%20Master%20Thesis%20FINAL.pdf Links ]

Ellett, R. (2013). Pathways to judicial power in transitional states: Perspectives from African courts. Routledge. [ Links ]

Englebert, P., & Dunn, K. C. (2013). Inside African politics. Lynne Rienner. [ Links ]

Finkel, S. (2014). The impact of adult civic education programmes in developing democracies. Public Administration and Development, 34(3), 169-181. [ Links ]

Fischer, J. (2017). Social enforcement of electoral security. Working Paper 2. http://www.elec toralviolenceproject.com/pubviolence/Links ]

Fjelde, H., & Höglund, K. (2016). Electoral institutions and electoral violence in sub-Saharan Africa. British Journal of Political Science, 46(2), 297-320. [ Links ]

Hafner-Burton, E., Hyde, S., & Jablonski, R. (2014). When do governments resort to election violence? British Journal of Political Science, 44(1), 149-179. [ Links ]

Harwood, A. (2019). Tracking election violence in Nigeria. Council on Foreign Relations. https://www.cfr.org/blog/tracking-election-violence-nigeria Links ]

Höglund, K. (2009). Electoral violence in conflict-ridden societies: Concepts, causes, and consequences. Terrorism and Political Violence, 21(3), 412-427. [ Links ]

Höglund, K., & Jarstad, K. (2011). Toward electoral security: Experiences from KwaZulu-Natal. Africa Spectrum, 46(1), 33-59. [ Links ]

Höglund, K., & Piyarathne, A. (2009). Paying the price for patronage: Electoral violence in Sri Lanka. Commonwealth & Comparative Politics, 47(3), 287-307. [ Links ]

Kaaba, O. B. (2015). The challenges of adjudicating presidential election disputes in domestic courts in Africa. African Human Rights Law Journal, 15(2), 329-354. [ Links ]

Klaus, K., & Mitchell, M. I. (2015). Land grievances and the mobilization of electoral violence: Evidence from Côte d’Ivoire and Kenya. Journal of Peace Research, 52(5), 622-635. [ Links ]

Larkins, C. M. (1996). Judicial independence and democratization: A theoretical and conceptual analysis. American Journal of Comparative Law, 44(4), 605-626. [ Links ]

Library of Congress. (2011). Nigeria: Election laws. Author. https://www.loc.gov/law/help/nigeria-election-law/index.php#election Links ]

Lindberg, S. (2006). The surprising significance of African elections. Journal of Democracy 17(1), 139-151. [ Links ]

Logan, C. (2017). Ambitious SDG goal confronts challenging realities: Access to justice is still elusive for many Africans. Afrobarometer Policy Paper 39. [ Links ]

Luo, Z., & Rozenas, A. (2018). The election monitor’s curse. American Journal of Political Science, 62(1), 148-160. [ Links ]

Murison, J. (2013). Judicial politics: Election petitions and electoral fraud in Uganda. Journal of Eastern African Studies, 7(3), 492-508. [ Links ]

Norris, P., Frank, R. W., & Martínez i Coma, F. (Eds.). (2015). Contentious elections: From ballots to barricades. Routledge. [ Links ]

Omenma, J. T., Ibeanu, O. O., & Onyishi, I. E. (2017). Disputed elections and the role of court in emerging democracies in Africa: The Nigerian example. Journal of Politics and Democratization, 2(1), 28-55. [ Links ]

PLACNG (Policy and Legal Advocacy Center). (2017). https://placng.org/wp/category/publications/Links ]

Prempeh, H. K. (2017). Africa’s “constitutionalism revival”: False start or new dawn? In E. N. Sahle (Ed.), Democracy, constitutionalism, and politics in Africa (pp. 13-59). Palgrave Macmillan. [ Links ]

Salehyan, I., Cullen, S. H, Hamner, J., Case, C., Linebarger, C., Stull, E., & Williams, J. (2012). Social conflict in Africa: A new database. International Interactions, 38(4), 503-511. [ Links ]

Salehyan, I., & Linebarger, C. (2015). Elections and social conflict in Africa, 1990-2009. Studies in Comparative International Development, 50(1), 23-49. [ Links ]

Sani, I. (2015). Electoral governance: Understanding the democratic quality of elections in Nigeria. Unpublished dissertation, The University of Edinburgh, Scotland. [ Links ]

Seymour, C., & Frary, D. P. (1918). How the world votes: The story of democratic development in elections. C. A. Nichols Company. [ Links ]

Simati, M. (2018). Semi-independent judiciaries and the danger of election violence: An analysis of African elections. Unpublished dissertation, Georgia State University. [ Links ]

Smedt, J. de. (2009). ‘No Raila, no peace!’ Big man politics and election violence at the Kibera grassroots. African Affairs, 108(433), 581-598. [ Links ]

Smidt, H. (2016). From a perpetrator’s perspective: International election observers and post-electoral violence. Journal of Peace Research, 53(2), 226-241. [ Links ]

Söderberg Kovacs, M., & Bjarnesen, J. (Eds.). (2018). Violence in African elections: Between democracy and Big Man politics. Zed Books. [ Links ]

Straus, S., & Taylor, C. (2012). Democratization and electoral violence in sub-Saharan Africa, 1990-2007. In D. Bekoe (Ed.), Voting in fear: Electoral violence in sub-Saharan Africa (pp. 15-38). United States Institute of Peace. [ Links ]

Taylor, C. F., Pevehouse, J. C., & Straus, S. (2017). Perils of pluralism: Electoral violence and incumbency in sub-Saharan Africa. Journal of Peace Research, 54(3), 397-411. [ Links ]

Utter, R. F., & Lundsgaard, D. C. (1993). Judicial review in the new nations of central and eastern Europe: Some thoughts from a comparative perspective. Ohio State Law Journal, 54(3), 559-606. [ Links ]

VonDoepp, P. (2006). Politics and judicial assertiveness in emerging democracies: High Court behavior in Malawi and Zambia. Political Research Quarterly, 59(3), 389-399. [ Links ]

VonDoepp, P., & Ellet, R. (2011). Reworking strategic models of executive-judicial relations: Insights from new African democracies. Comparative Politics, 43(2), 47-165. [ Links ]

Widner, J. (2001). Building the rule of law: Francis Nyalali and the road to judicial independence in Africa. W. W. Norton & Company. [ Links ]

1After the end of colonialism, most African countries emerged with at least two political parties, many informed by either a Western or Soviet Union ideology. Countries such as Botswana, Kenya, Senegal, Tanzania, and Zambia, for example, all held multiparty elections in the 1960s. These countries, however, reverted to a one-party system shortly thereafter as many founding fathers sacrificed democracy for political stability, which they deemed favorable for managing such diverse social groups. See [xref ref-type="bibr" rid="r21"]Englebert and Dunn (2013)[/xref].

2For more on elections in Eritrea, see http://africanelections.tripod.com/er.html

3Demarest and Langer look at conflict patterns in Nigeria as recorded by Social Conflict in Africa Dataset (SCAD), Armed Conflict Locations and Events Dataset (ACLED) and a dataset developed by the authors using Nigerian sources and suggest that ACLED tends to under-report conflict events because it relies on international sources which are more likely to only report high profile events. See Demarest and Langer (2018).

4For more, see Bekoe (2012).

5See Lindberg (2006) field work in Ghana which suggests that the quality of democracy improves as countries engage in regular elections.

6Conceptualized as violence related specifically to elections, starting six months before elections to include election day.

7We conceptualize and measure post-election violent events as election related violent events that happen right after polling stations close to three months after polling.

8For presidential elections.

9At the time of writing this paper, data on pre-election violence for the 2019 elections was available but the data on post-election violence and EPT rulings for 2019 election petitions was not available yet. Subsequently, we are able to test the effect of the 2011 and 2015 EPT rulings on the 2015 and 2019 pre-election violence respectively. In other words, how did the 2011 EPT rulings affect the 2015 pre-election violence? Similarly, how did the 2015 EPT rulings affect the 2019 pre-election violence?

10Due to insecurity in the North East, elections were not held in Adamawa, Borno, or Yobe states. We do not include data for these three states in our analysis.

11Data for the 2019 post-election period was not available at the time of writing.

12PLACNG only provides disposition data for 560 of the petitions filed in 2015.

13For robustness checks, we took the average percentage of respondents who stated they had “no trust” in the courts over a 12-year period at the state level as our alternative measure of trust/lack of trust. We did this for two reasons. First, the total number of respondents per survey per state ranges from about 15 to 80, so aggregating surveys together over time increased our total number of respondents per state, increasing our ability to generalize from the sample. Additionally, because public opinion can be temporarily influenced by short term shocks or specific events close to the timing of the survey (i.e., a corruption scandal or the like), we chose to take the average of the past several surveys (round 3 fielded in 2005 through round 7 fielded in 2017) as an alternative measure of trust and the findings still stay consistent and significant compared with using round 6 and 7 trust indices for 2015 and 2019 elections, respectively.

14The role of competent electoral commissions in reducing the likelihood of election violence taking place has been extensively examined by works such as [xref ref-type="bibr" rid="r6"]Birch and Muchlinski (2018)[/xref]; [xref ref-type="bibr" rid="r14"]Darnolf and Cyllah (2014)[/xref]; Claes and von Borzyskowski (2018).

15While we acknowledge that election violence has many consequences other than death, ranging from physical, psychological, displacement, destruction, etc., our paper focuses on fatality because it is a consequence that is both measurable and readily accessible in available datasets.

16There is some evidence to suggest that the intensity of violence and/or conflict has an impact on institutional trust (see [xref ref-type="bibr" rid="r5"]Bellows & Miguel, 2006[/xref]).

17Works by [xref ref-type="bibr" rid="r27"]Höglund (2009)[/xref], [xref ref-type="bibr" rid="r24"]Fjelde and Höglund (2016)[/xref] and [xref ref-type="bibr" rid="r19"]Dunaiski (2015)[/xref] argue that electoral rules can influence the likelihood of pre-election violence taking place.

18The Nigerian state is our primary unit of analysis. The 33 observations are the number of Nigerian states in which elections were held in both 2011 and 2015. General elections were suspended in three states in 2015 (Adamawa, Borno, & Yobe) due to insecurity caused by Boko Haram.

Recebido: 06 de Junho de 2019; Aceito: 16 de Outubro de 2019

Creative Commons License Este é um artigo publicado em acesso aberto sob uma licença Creative Commons