Fighting and Voting: Violent Conflict and Electoral Politics
Abstracts
“Riots and Votes”
Steven Wilkinson & Chris Haid, University of Chicago
Abstract: This paper uses a new dataset on riots and votes in the Indian state of Gujarat to examine how the Feb-April 2002 riots in that state affected the votes of the major parties in the December 2002 state elections. More generally, we look at the ways in which violence alters electoral preferences. The paper uses multi- level models, as well as more qualitative evidence to demonstrate that; a) the riots were engineered for political purposes; b) that they had a significant effect on turnout as well as in changing voter preferences in the December elections, which the incumbent BJP won by an unexpectedly large margin.
“Electoral Geography and Conflict in Kenya: Examining the Local-Level Incidence of Violence in Rift Valley Province after the 2007 Election” Kimuli Kasara, Columbia University
Abstract: Perpetrators frequently use violence to change local demography and social scientists have examined the use of violence to expel people from urban neighborhoods, ancestral homelands, or strategically important areas. This paper examines whether exclusionary violence follows an electoral logic by studying the local-level incidence of violence and displacement in Kenya during the period following the disputed 2007 general election. Although those involved in the violence had various motivations, I argue that if some perpetrators used violence in order to change future electoral outcomes, we should expect a greater incidence of violence in and displacement from localities: 1) that are electorally competitive; 2) in which residents are expected to have a lower propensity to return once displaced and 3) in which voters could influence the outcome of parliamentary races.
In order to examine the relationship between electoral geography, violence and displacement, I use an original GIS of local-level electoral outcomes in elections prior to 2007 as well as data on violent events and displacement in Kenya from December 2007 to March 2008. The data provide some support for the hypothesis that some politicians and their supporters engaged in redistricting through violence during this political crisis. I also present evidence on other correlates of violence and displacement, such as poverty, local demography, and physical geography.
“The Formation of the State: Evidence from Colombia” Jim Robinson (Harvard)
Abstract: Max Weber defines the state as the entity that has the monopoly of legitimate violence in a territory. Yet many states in Latin America, Africa or Asia lack this monopoly and thus the capacity to govern effectively. In this paper we develop a new perspective on the formation of the state, specifically the establishment of a monopoly of violence. We build a model to explain the incentive of central states to eliminate non-state armed actors (paramilitaries) in a democracy. The model is premised on the idea that paramilitaries may choose to and can influence elections. Since paramilitaries have preferences over policies, this reduces the incentives of the politicians they favor to eliminate them. The model also shows that in non-paramilitary areas policies are targeted at citizens while in paramilitary controlled areas they are targeted at paramilitaries.
We then test the predictions of our model using data from Colombia between 1991 and 2006. We first present regression and case study evidence supporting our postulate that paramilitary groups can have significant effects on elections for the legislature and executive. Next, we show that the evidence is also consistent with the implication of the model that paramilitaries tend to persist to the extent that they deliver votes to candidates for the executive whose preferences are close to theirs. We show that this effect is larger in areas where the Presidential candidate would otherwise not do well. These results illustrate that, consistent with our model, there appears to be a symbiotic relationship between some executives and the paramilitaries. Finally, we use roll-call votes to illustrate the legislative quid pro quo between the executive and paramilitaries. In particular, while the executive proposed a highly lenient law governing paramilitary de-mobilization (the Justice and Peace Law) `paramilitary legislators’ voted to change the Constitution to remove Presidential term limits.
“Ethnic Parties and Democratic Stability”
Kanchan Chandra (NYU)
Abstract: Classic arguments in empirical democratic theory identify the politicization of ethnic divisions by ethnic parties as one of the primary causes of democratic breakdown. These arguments have not, however, been systematically tested against cross-national data. This paper conducts an empirical test of this relationship, using new data on the politicization of ethnic divisions from an original dataset on the activation of ethnic identities by political parties (dubbed CDEI (Constructivist Dataset on Ethnicity and Institutions). The principal measure I use for the politicization of ethnic divisions is EVOTE, a variable generated from CDEI that measures the aggregate vote won by ethnic parties across countries in lower-house legislative elections closest to the year 1996. The measures for democratic breakdown are (1) a count of any civil wars in each country that break out within five years of the election to which EVOTE refers, generated using data from Fearon and Laitin 2003 and (2) a count of any transitions from democracy to dictatorship in these countries using data from Przeworski 2000. The measures for democratic instability are (1) a count of violent events short of civil war within five years of the election to which EVOTE refers, generated from the Banks Cross-National Data Archive and (2) a measure of downward movement on the eleven point scale of democracy constructed by the Polity IV dataset. I find little evidence of a direct relationship between the politicization of ethnic identities and either the breakdown or the instability of democratic governments.
“Persistent Fighting to Forestall Adverse Shifts in the Distribution of Power”
Robert Powell (UC Berkeley)
Abstract: Three striking features or stylized facts about both interstate and civil war are that (i) there are often periods of persistent fighting, (ii) fighting commonly ends in negotiated settlements as well as militarily decisive outcomes, and (iii) fighting sometimes recurs. We have few if any models that exhibit all of these features along an essentially unique equilibrium path. This paper presents a simple model of state consolidation in which fighting results from an effort to forestall adverse shifts in the distribution of power. The equilibrium path of the model displays these features, and the analysis links the pattern of fighting to the way that the distribution of power shifts during the consolidation process. Fighting occurs when the distribution of power is shifting rapidly and stops when the shift in the distribution of power slows or stops. Fighting resumes if the distribution of power again begins to shift rapidly. The analysis also shows that consolidation can occur without fighting if the process is sufficientlly slow.
“Time to Kill: The Impact of Election Timing and Sequencing on Post-Conflict Stability” Dawn Brancati (University of Washington at St. Louis) with Jack Snyder
Abstract: Elections constitute a fundamental element of post-conflict peacebuilding efforts in the post-Cold War era and are often held soon after conflicts end. Yet, the impact of early elections on post-conflict stability is unclear. While early elections may facilitate peace agreements, hasten democratization, and promote economic recovery, they may also entrench wartime combatants in power, undermine genuine democracy, and spark an immediate return to violence. Our goals in this project are, thus, two-fold. The first is to understand the factors that lead to early elections and the second is to analyze the effect of electoral timing and sequencing on post-conflict stability. We argue that in the short term, early elections can spark a return to violence because they increase the likelihood that combatants will reject the elections with weapons at the ready to engage in fighting, and that they can also provoke violence in the long-term by entrenching wartime combatants in power whose political behavior once in office fosters new conflicts. Our analysis, in contrast to previous studies of election timing, which are all qualitative in nature, is based on all civil wars that have ended between 1945 and 2006, and relies on duration models to examine the timing of post-conflict elections, as well as matching-methods to understand the causal relationship between electoral timing and sequencing and post-conflict stability.
“Concession amidst Contestation:
Explaining Why Violent Groups Participate in Elections”
Aila Matanock (Stanford)
Abstract: This project analyzes why violent groups participate in elections. In a new dataset on Latin America, I find that almost one in four groups make this transition. Existing literature on why violent groups participate in elections is in the very early stages of development, but it does suggest two general categories of important factors: state-level and group-level factors. I expand these, based on an initial probe into the data on Latin America, to suggest that institutional changes, often at a more sophisticated level than simply initiating elections, encourage groups to participate in elections. Several factors affect whether institutional changes are made, and which types of changes are made. These factors include group strength, and especially the types of resource structures that provide this strength, as well as inter-group competition. The next steps in the project are to refine measures of these hypotheses and then test the theory in a global dataset.
“The Dark Side of Politics: Violence Against Civilians During Conventional Civil War” Laia Balcells (Yale)
Abstract: This paper analyzes the relationship between political loyalties and violence at the local level during civil war, with a particular focus on so-called conventional civil wars.These wars are characterized by stable frontlines and the use of both heavy and light weaponry by both incumbents and rebels. I distinguish between “direct” and “indirect” violence, which vary in the nature of their production: the former implies the use of light technology and face-to-face interaction with the victims. The latter implies the use of heavy technology and the lack of face-to-face interaction with the victims. Because of these differences, local political configurations have diverging predicted effects on each of these types: level of indirect violence should increase with electoral support for the enemy group –as the likelihood to imprecisely kill enemies grows. Direct violence, in contrast, should increase with the degree of political parity between political factions –as civilian strategic collaboration with the group increases and so it does the potential to precisely kill enemies. I test the predictions with self-collected data on violence perpetrated by different armed groups in 1,421 municipalities of Catalonia and Aragon during the Spanish Civil War (1936-1939).
“Powersharing under the threat of conflict” Stergios Skaperdas (UC Irvine)
Abstract: We examine a model of powersharing in a post-conflict environment and discuss some of its implications. The model is one in which the parties have difficulty creating enforceable commitments to powersharing. The level of a public good that can be called “governance,” “institutions,” or “property rights” is critical for choosing powersharing over continuing conflict. The history of past interactions, the contributions of the contending parties, but also outside actors are the three critical factors that determine the level of the public good and, thus, whether powersharing proves successful.
“Erecting Invisible Ramparts: A Formal Analysis of Political Violence and Electoral Support in a Networked Setting”
Chris Haid
Abstract: This paper is motivated by the context of grassroots politics in Kingston, Jamaica, where violence was used by parties to affect electoral outcomes by politically homogenizing neighborhoods in a single member district, first-past-the-post electoral system. The paper focuses on the initial use violence in elections and seeks explain where in geographic space violence occurs and how the choice of location effects the amount and flows of party support in different regions. A game-theoretic model played on a graph examines the role social interactions play in structuring political violence, social interactions, and political support in a networked population.
“The Making of Policy: Institutionalized or Not”
Mariano Tommasi with Carlos Scartascini
Abstract: This paper departs from well known models of policymaking in the context of formal political institutions, and examines the implications of enlarging the action set of political players, to include a number of political actions (such as road blockades or bribes) which are pervasive in many countries. The paper explores the implications of such an extension for the degree of institutionalization of policymaking (and relatedly, the strength and quality of policymaking institutions such as Congress), for the nature of resulting policies, and for the intensity of use of these alternative political technologies.
Different polities are characterized by different levels of institutionalization of their policymaking processes, and these differences tend to be reinforcing. When Congress and the political party system are the effective conduits of preference aggregation and political bargaining, various relevant actors will place their bets (that is, their investments) in those institutions, most citizens will believe that those are the spaces where relevant decisions are made, and this whole logic would reinforce and become self-fulfilling. On the contrary, if such institutional arenas are not taken too seriously, and everybody knows that the way of getting something out of the political system is to blockade a road or to bribe the president, those investments in the institutionalization of Congress or parties will not be undertaken, and the lack of institutionalization will be reinforced. Our model captures this in a stylized manner.
A related implication of our analysis, is that there is multiplicity of equilibria. Polities might be stuck with higher or lower levels of institutionalization. Polities in more institutionalized equilibria will behave as predicted in the “tidy” literature on political institutions and policies. Polities in less institutionalized equilibria might behave differently. This could have important implications for cross national empirical analysis on the effects of formal political rules on public policy.
Among the comparative statics results, we find that lower institutionalization (and greater use of alternative political technologies) are more likely the lower the cost of using these technologies, the higher the potential damage they can cause, the lower the wealth of the economy, and the more asymmetric the distribution of de jure political power. Relatedly, we find a strategic complementarity in the use of various alternative political technologies; for instance “bribes by the rich” and “protests by the poor” are likely to be countervailing forces that will tend to happen together in polities with weaker political institutions.
We present three levels of empirical evidence supporting our argument: international cross-country correlations, individual level survey responses for 18 Latin American countries, and a narrative of the case of Bolivia over the last 25 years, which is suggestive of a switch from a more to a less institutionalized equilibrium.