The Institutional Roots of Electoral Violence
Multiparty elections are fundamental components in the transition from authoritarian to democratic rule. But experiences from around the globe, for instance in Afghanistan, Zimbabwe and Colombia, show that elections can degenerate into violence. Violent elections hinder individuals from exercising their political rights and undermine the legitimacy of democratic institutions. Research on the causes of election-induced violence has primarily focused on the immediate factors that shape the dynamics of electoral contests, such as the uncertainty of the electoral outcome or the presence of election observers. This study shifts focus from the short-term to the long-term processes of institutional development during political transitions: How is the risk of election-induced violence influenced by institutional developments before, during and after the transition from authoritarian to democratic rule? We use a mixed method approach to capture the dynamics at play by combining quantitative and qualitative analysis. The project will provide a first global dataset on violent electoral conflicts from 1989 to 2015. We will also make an in-depth analysis of Kenya and Zambia. These countries have, for instance, similar electoral systems, but markedly different experiences of violence around elections. Jointly, the contributions will advance our understanding of the causes of electoral violence and its consequences for political institutions, peace and democracy.
Final report
The Institutional Roots of Electoral Violence
Violence constitutes a critical threat to the integrity of elections and defies the very notion of democratic conduct. The overall aim of the project has been to investigate how the risk of election-related violence is influenced by the evolution of institutions before, during, and after transitions from autocratic to democratic rule. In contrast to previous research, which primarily has focused on factors in the short-term to explain the prevalence of violent elections, the project has centered on developments over longer time periods.
The project has been implemented through three main components. First, to address a data gap, we have collected and made publicly available a global dataset – the Deadly Electoral Conflict (DECO) Dataset – that maps lethal events of electoral violence 1989–2017 across the world. In a second component, drawing on new cross-national data on democracy and political institutions with long time-series and statistical methods, we have captured how the configurations of political institutions affect the risk of large-scale political violence. In a third component, we have carried out qualitative analyses of and comparisons within and between Kenya and Zambia to trace how institutional evolution over time affects the presence or absence of violence in connection with elections.
The project advances the research agenda on electoral violence and the relation to political institutions, over time and across countries, theoretically and empirically.
First, DECO provides the research community with a unique data source and offers new insights into the prevalence of electoral violence globally and over different categories of violence. A key improvement of DECO over other related dataset, is that it provides event data over a close to thirty-year period. Another advantage is its stringent methodology used to parse out election-related violence from other forms of political violence. DECO builds on and is compatible with the georeferenced event data UCDP-GED, allowing users to examine direct links between organized violence, electoral contestation and the broader dynamics of political conflict. A key finding from DECO is that violence is distributed fairly equally over three categories of electoral violence: 1) state-based violence between a government and an armed non-state actor (35%), 2) non-state violence between armed non-state actors, such as rebel groups, political parties, or ethnic groups (29%), and 3) one-sided violence, in which an armed state or non-state actor targets unarmed civilians (33%). DECO also documents a fourth category of violence between an organized armed actor (predominantly state forces) and civilians who are armed, which constitutes only 3% of the DECO events. Such categorization of events enables assessment of how different political institutions or interventions affect different forms of election violence across and within countries and over time, as well as more fine-grained analysis of the consequences of such violence.
Second, the project uncovers how legacies from authoritarian rule cast a long shadow into the multiparty era and contribute to explain the presence or absence of election-related violence. In particular, the comparison between Zambia, where the first multi-party elections were peaceful, and Kenya, where the first multiparty elections were violent, showcases the importance of political legacies (Brosché, Fjelde and Höglund 2021). The study points to how – in the context of authoritarian regimes in multi-ethnic societies – leaders who rule by means of a narrow ethnic support base create interethnic relations characterized by competition and fragmentation. With the introduction of multiparty elections, such exclusionary approaches heighten the risk of electoral violence by foreclosing the forging of cross-ethnic elite coalitions and by contributing to more competitive intercommunal relations. In this context, leaders can gain politically by resorting to electoral tactics that rest on hostile group narratives. By contrast, strategies that rely on a more broad-based ethnic support base (inclusionary approaches), are more likely to produce interethnic relations that alleviate the risks of violence in founding elections by fostering inclusive inter-elite bargaining and by contributing to more amicable intercommunal relations, thereby lowering incentives for hostile out-group mobilization.
Third, the project provides new insights about the particular configurations of democratic institutions that dampen the risk of large-scale political violence, with particular emphasis on how elections combined with other political institutions contribute to more or less peaceful political outcomes. Our findings, drawing on cross-national time-series data over a long time-period, indicate that competitive elections is only one of several institutional means through which citizens can hold their political leaders accountable, and that elections in themselves might not be sufficient to reduce the risk of large-scale political violence. In a first study (Hegre, Bernhard, and Teorell, 2018), results show how a strong civil society, in particular, contributes to democratic processes in between elections that help reduce the risk of interstate conflict. A second study (Fjelde, Knutsen, and Nygård, 2021) shows that to reduce the risk of intrastate conflict, the electoral mechanism of vertical accountability needs to be combined with strong horizontal accountability mechanisms, in the form of institutional constraints on government power from a strong parliament or independent judiciary.
The project has given rise to several new research questions. One relates to the need for research to engage more in-depth with the actors involved in electoral violence. For example, it has become evident that the research field adopts a too simplified view of actors by using a binary conceptualization of government versus opposition actors. Such a dichotomy fails to capture violence happening between actors within the opposition and also does not recognize that within the government there are several different actors – the police, government affiliated militias or political supporters – that employ violence to influence the elections. This limitation in previous research uncovered in the course of the project serves as a foundation for a new project funded by the Swedish Research Council which addresses actor constellation and patters of election violence. Another research question that has come to the forefront through our work on this project is the longer-term consequences of election violence for institutions and political practices. Our case studies highlight how institutional choices and political discourse can produce electoral environments that are conducive to violence. Our results also show, however, that once election-violence unfolds along ethnic and/or partisan lines, polarization and animosity remains a challenge for future democratization trajectories. We are currently working on several studies that examines these legacies.
The project findings have been disseminated through several channels. Project members have presented the research at national and international conferences and in relevant academic environments, including at the Annual Meeting of the American Political Science Association, Development Research Conference, Amsterdam University, and Varieties of Democracy (Gothenburg University). We have engaged with relevant stakeholders both in connection to field work in Zambia and Kenya, and through policy-academic dialogue organized by, for example, Folke Bernadotte Academy. We have also authored op-eds and popular science articles for reaching a broad audience. All the articles and book chapters published are available open access.
Violence constitutes a critical threat to the integrity of elections and defies the very notion of democratic conduct. The overall aim of the project has been to investigate how the risk of election-related violence is influenced by the evolution of institutions before, during, and after transitions from autocratic to democratic rule. In contrast to previous research, which primarily has focused on factors in the short-term to explain the prevalence of violent elections, the project has centered on developments over longer time periods.
The project has been implemented through three main components. First, to address a data gap, we have collected and made publicly available a global dataset – the Deadly Electoral Conflict (DECO) Dataset – that maps lethal events of electoral violence 1989–2017 across the world. In a second component, drawing on new cross-national data on democracy and political institutions with long time-series and statistical methods, we have captured how the configurations of political institutions affect the risk of large-scale political violence. In a third component, we have carried out qualitative analyses of and comparisons within and between Kenya and Zambia to trace how institutional evolution over time affects the presence or absence of violence in connection with elections.
The project advances the research agenda on electoral violence and the relation to political institutions, over time and across countries, theoretically and empirically.
First, DECO provides the research community with a unique data source and offers new insights into the prevalence of electoral violence globally and over different categories of violence. A key improvement of DECO over other related dataset, is that it provides event data over a close to thirty-year period. Another advantage is its stringent methodology used to parse out election-related violence from other forms of political violence. DECO builds on and is compatible with the georeferenced event data UCDP-GED, allowing users to examine direct links between organized violence, electoral contestation and the broader dynamics of political conflict. A key finding from DECO is that violence is distributed fairly equally over three categories of electoral violence: 1) state-based violence between a government and an armed non-state actor (35%), 2) non-state violence between armed non-state actors, such as rebel groups, political parties, or ethnic groups (29%), and 3) one-sided violence, in which an armed state or non-state actor targets unarmed civilians (33%). DECO also documents a fourth category of violence between an organized armed actor (predominantly state forces) and civilians who are armed, which constitutes only 3% of the DECO events. Such categorization of events enables assessment of how different political institutions or interventions affect different forms of election violence across and within countries and over time, as well as more fine-grained analysis of the consequences of such violence.
Second, the project uncovers how legacies from authoritarian rule cast a long shadow into the multiparty era and contribute to explain the presence or absence of election-related violence. In particular, the comparison between Zambia, where the first multi-party elections were peaceful, and Kenya, where the first multiparty elections were violent, showcases the importance of political legacies (Brosché, Fjelde and Höglund 2021). The study points to how – in the context of authoritarian regimes in multi-ethnic societies – leaders who rule by means of a narrow ethnic support base create interethnic relations characterized by competition and fragmentation. With the introduction of multiparty elections, such exclusionary approaches heighten the risk of electoral violence by foreclosing the forging of cross-ethnic elite coalitions and by contributing to more competitive intercommunal relations. In this context, leaders can gain politically by resorting to electoral tactics that rest on hostile group narratives. By contrast, strategies that rely on a more broad-based ethnic support base (inclusionary approaches), are more likely to produce interethnic relations that alleviate the risks of violence in founding elections by fostering inclusive inter-elite bargaining and by contributing to more amicable intercommunal relations, thereby lowering incentives for hostile out-group mobilization.
Third, the project provides new insights about the particular configurations of democratic institutions that dampen the risk of large-scale political violence, with particular emphasis on how elections combined with other political institutions contribute to more or less peaceful political outcomes. Our findings, drawing on cross-national time-series data over a long time-period, indicate that competitive elections is only one of several institutional means through which citizens can hold their political leaders accountable, and that elections in themselves might not be sufficient to reduce the risk of large-scale political violence. In a first study (Hegre, Bernhard, and Teorell, 2018), results show how a strong civil society, in particular, contributes to democratic processes in between elections that help reduce the risk of interstate conflict. A second study (Fjelde, Knutsen, and Nygård, 2021) shows that to reduce the risk of intrastate conflict, the electoral mechanism of vertical accountability needs to be combined with strong horizontal accountability mechanisms, in the form of institutional constraints on government power from a strong parliament or independent judiciary.
The project has given rise to several new research questions. One relates to the need for research to engage more in-depth with the actors involved in electoral violence. For example, it has become evident that the research field adopts a too simplified view of actors by using a binary conceptualization of government versus opposition actors. Such a dichotomy fails to capture violence happening between actors within the opposition and also does not recognize that within the government there are several different actors – the police, government affiliated militias or political supporters – that employ violence to influence the elections. This limitation in previous research uncovered in the course of the project serves as a foundation for a new project funded by the Swedish Research Council which addresses actor constellation and patters of election violence. Another research question that has come to the forefront through our work on this project is the longer-term consequences of election violence for institutions and political practices. Our case studies highlight how institutional choices and political discourse can produce electoral environments that are conducive to violence. Our results also show, however, that once election-violence unfolds along ethnic and/or partisan lines, polarization and animosity remains a challenge for future democratization trajectories. We are currently working on several studies that examines these legacies.
The project findings have been disseminated through several channels. Project members have presented the research at national and international conferences and in relevant academic environments, including at the Annual Meeting of the American Political Science Association, Development Research Conference, Amsterdam University, and Varieties of Democracy (Gothenburg University). We have engaged with relevant stakeholders both in connection to field work in Zambia and Kenya, and through policy-academic dialogue organized by, for example, Folke Bernadotte Academy. We have also authored op-eds and popular science articles for reaching a broad audience. All the articles and book chapters published are available open access.