Reexamining ballots and bullets

Burundi military and police forces stand in formation as they guard the venue for the Conseil National pour la Defense de la Democratie – Forces pour Defense de la Democratie (CNDD-FDD) party congress in the capital, Bujumbura, April 25, 2015. Burundi President Pierre Nkurunziza spoke for the first time on seeking a third term moments after the ruling party nominated him. Nkurunziza’s opponents have said another term for him would be unconstitutional. REUTERS/Thomas Mukoya
Electoral violence is an unfortunately common feature of headlines around the world. Burundi’s government has violently targeted its opposition before and after the country’s recent elections. In Iraq, powerful Shiite political parties have their own militias, while other militias are hoping to enter electoral politics. Pakistan’s military is aggressively targeting Karachi’s most important armed political party, yet another chapter in a long history of election-linked violence in the city.
Elections can tip unstable post-conflict environments into renewed warfare, yoke non-state actors into networks of armed patronage and undermine the quality of democracy. Rather than alternatives, elections and coercion are often intertwined. Policymakers are increasingly trying to understand how to free democratic politics from the shadow of the gun, but are limited by a narrow concept of this violence, its actors, and causes.
In recent research published in Comparative Politics (ungated), I identify new insights about the relationships between violence and electoral politics, the simplest but most important of which is that “electoral violence” is far more diverse than commonly appreciated. Multiple kinds of actors, from local landlords to ruling parties, use violence in pursuit of very different goals. These varieties of electoral violence have different causes and require different policy responses.
Most of what we conventionally associate with electoral violence can be described as “intra-systemic,” in which coercion, displacement, killing and intimidation are deployed to win elections. This violence is sometimes carried out by government security forces (as Susan Hyde and Emilie Hafner-Burton have discussed) against dissidents and opponents. Sometimes, however, non-state allies of the regime, such as militias and armed wings of ruling parties, operate as shock troops to shatter electoral challenges. They kill or displace electoral rivals and their potential supporters. Contemporary Ethiopia and 1950s Burma are examples of these pro-regime logics of violence. Coercion is particularly important for competitive authoritarian regimes that manipulate illiberal elections with “thugs” and security forces.
Opposition parties can also arm themselves and fight back against rulers and their backers. This can lead to full-scale militarized elections, fusing ballots and bullets. At various points, Bangladesh, Kenya and India’s West Bengal have experienced highly militarized elections campaigns: ruling and opposition parties use violence, while trying to simultaneously get their own voters to the polls, build patronage networks and develop policy positions. This type of “pro-systemic” electoral violence can also undermine democratic representation and accountability, by influencing voters’ choices through threats and actively rigging and manipulating campaigns and voting.
In other cases, however, electoral violence is “anti-systemic.” Rather than winning elections, violence is intended to prevent democratization or destroy democratic politics altogether. The Burmese military crackdown of 1988-1990 deployed violence to shatter democratization. Support for armed groups by elements of the apartheid regime in early 1990s South Africa was intended to undermine the democratic transition. Key party-linked paramilitaries in Weimar Germany ultimately sought to replace German democracy with communist or fascist dictatorship. The Pakistani Taliban launched sustained attacks on rallies and candidates during the 2013 general election, reflecting its deep ideological opposition to the electoral process and specific parties.
Given these disparate kinds of violence, it is also important to recognize there is no single cause of electoral violence either. To avoid conceptual stretching, scholars must disaggregate types of election-related violence both in theory and data. The Pakistani Taliban’s onslaught against electoral practice was likely driven by very different mechanisms, and took on markedly different forms, than politicians’ support for criminal gangs in post-New Order Indonesia.
Consequently, policymakers need to respond to the specific logic(s) of electoral violence at work. Sometimes governments are to blame for violence, while in other cases armed opposition parties or unaligned local actors are responsible. Assigning this responsibility requires careful analysis of the motivations and capabilities of the key political players. Groups using violence to win elections will respond to very different incentives than those trying to forestall or overthrow democracy, while local private armies and formal state security forces have distinct strengths and vulnerabilities. The typology I advance in my article provides a framework for making these distinctions.
Similar diversity can also be found among electoral armed groups themselves. Taking armed groups seriously, as I show in International Studies Quarterly (ungated), is essential for understanding militarized elections. Usually we think of these groups as simple proxies for rulers, deployed to achieve government aims behind a veil of deniability. In many cases this is absolutely correct, with armed groups lacking independence from the regime.
Nevertheless, some electoral armed groups are autonomous of state patronage and support. They are powerful free agents that bargain with or defy rulers. Governments cannot easily shut down these organizations and instead are forced to make concessions for their cooperation or support. As regimes maneuver to advance their ideological and electoral interests–and armed groups respond– several distinct armed orders can emerge, including the incorporation of armed groups into the state, their normalization as a part of the political landscape or even open conflict with the state. For instance, rural private armies of the Philippines and armed Kurdish parties of northern Iraq have bargained with central rulers, while in India’s Punjab in the early 1980s a militant group that was supported by Congress Party elites for electoral reasons turned against the state, triggering a decade-long insurgency.
These “armed politics” between states and armed actors are most important where violence and elections enduringly co-exist. In these contexts, whether Iraq or north India, standard distinctions between civil conflict and electoral violence, war and peace, and political parties and armed groups break down. Improving our understanding of these politics is crucial for making sense of a contemporary political landscape that all too often blends violence and voting.
Paul Staniland is an assistant professor of political science, co-director of the Program on International Security Policy, and co-founder of the Program on Political Violence at the University of Chicago. He is the author of Networks of Rebellion: Explaining Insurgent Cohesion and Collapse (Cornell University Press, 2014).