Research Projects

Islamic Ties and Rebel Territorial Control in African Armed Conflicts

Grant-funding provided by the Association for the Study of the Middle East and North Africa (ASMEA)

​

Presented at Association for the Study of the Middle East and Africa (ASMEA) Annual Conference

Panel on Islam and Violence, November 7, 2022; Washington, D.C.

​

Presented at the International Studies Association (North East) Annual Conference

Panel on Civil Wars and Rebellion, November 4, 2022; Baltimore, Maryland

Rebel groups who purport to operate on behalf of Muslim populations account for a sizable fraction of armed conflicts in Africa. Though these groups share ties to Islam, they display significant variation in strategies, tactics, and capabilities. In this paper, I investigate the role that these ties play in a group's willingness to contest and control territory. My argument is two-fold: First, that rebel groups with ties to Islam demonstrate a systematic pattern of contesting areas in which they expect support from local Muslim populations. Secondly, groups who depend more on these ties---placing higher premiums on the support of co-ethnics---more strongly utilize this targeting strategy, and may even be deterred from contesting otherwise valuable areas without sufficient levels of support. To evaluate my theory, I employ quantitative and qualitative information from conflicts in Nigeria, Côte d'Ivoire, and the Central African Republic. These cases represent a wide range of unique dynamics as well as clear variation in the extent to which rebel groups place premiums on the support of Muslim populations. My arguments are evaluated within a mixed-method framework. Using a novel method for latent estimation of territorial control across space and time, quantitative sub-national analyses test the significance of local concentrations of Muslims on a group's willingness to contest and control different localities. The results of these models are then used as evidence to support a qualitative comparison of the cases to determine whether groups with stronger ties to Islam more intently select their targets. This project provides important insights into the role of Islam in African armed conflicts. More generally, the investigation sheds light on the relationship between ethnic geography and wartime dynamics.

Presented at the Midwestern Political Science Association (MPSA)
Panel on Modelling Conflict Processes, April 8, 2022; Chicago, Illinois

Territorial control is an elusive concept which is difficult to observe and measure. Its inherent endogeneity to behaviors and events within civil wars intensifies the empirical challenge, but also highlights the significance of finding solutions. A general model of territorial control should explain within-conflict variation of authority across actors, space, and time. However, it must also be systematic, applicable and comparable across conflicts. Emerging approaches take advantage of event data to make inferences about levels of control. These approaches, however, assume that actor behavior is bound to a particular category of warfare. This assumption is unnecessarily constraining and does not align with an empirical record which demonstrates evolutionary behavior. In this paper, I combine elements of emerging approaches to produce a model which estimates territorial control across all types of conflict. Relying on information regarding the timing and location of battles, guerilla skirmishes, the outcomes of these clashes, indiscriminate attacks, and other types of political violence, my model takes advantage of multiple data sources to produce a continuous measure of control by each actor. The estimates are continuous across space and time, take into consideration the control of other actors operating in the same areas, and uncertainties which arise from partial information are explicitly built into the model.

Presented at the Southern Political Science Association (SPSA)
January 15, 2022; San Antonio, Texas

Despite the fact that post-war states rarely transition into consolidated democracies, most studies of post-civil war regime transitions have focused exclusively on democratization. In this study, I shift the focus to factors which permit a more widespread occurrence: autocratization after war. Consistent with previous literature, I argue that the power structures leftover from war are carried into the post-war political environment, and that a stable balance of power promotes democratization. While strong adversaries place pressures on leaders to democratize, asymmetrical power structures permit leaders to consolidate and autocratize after war. Rather than use war outcomes alone as a proxy for relative power, I produce a theoretical framework for variable power structures within war outcomes to reconcile the fact that a particular power distributions can lead to multiple possible outcomes. Using a conceptualization of relative power which relies on the battlefield effectiveness of the warring parties, I introduce a measure which employs the location of fighting, relative to major cities, as a proxy. While initial regression models show support for my hypotheses, I demonstrate that after accounting for the effect that power distributions have on observing a given war outcome in the first place, more balanced power distributions actually create incentives for leaders to cheat and clamp down on power while when one side has a clear advantage, they are able to transfer that advantage into peaceful political participation. My results hold generally across war outcomes, though the findings for negotiated settlements are particularly robust.

Co-authored with Post Basnet.
Presented at the American Political Science Association (APSA)
Panel on Party Politics, October 1, 2021; Seattle, Washington

How does the wartime behavior of rebel groups impact their survival as a post-war political party? We argue that rebel groups leave legacies with civilian populations, impacting individual attitudes and collective memories which translate into long-term voting preferences. Particularly, rebels who rely on civilians and build institutions to maximize rebel-civilian cooperation leave fond legacies for which they are rewarded at the ballot box, while rebels face electoral consequences for coercive wartime behaviors. Using an original dataset of electoral outcomes for 57 post-rebel parties across 198 elections over a 30 year period in 52 countries, we test hypotheses derived from our theory through a Cox regression framework. We find that rebels who develop an extensive array of governing institutions during war---and particularly legal systems---are much more likely to survive, and for longer periods, than groups who fail to develop these institutions. We also show that the nature of these institutions matter---rebels who take a conciliatory approach to war-related justice are destined to survive, while those which take a coercive approach are much more likely to fail. Contrary to theoretical expectations, we show that rebels who rely on civilians to support their war efforts are unable to translate that support into voter loyalty after war, but institutionalizing civilian support through a tax system mitigates the post-conflict risks associated with wartime civilian dependence. Finally, while rebels which victimized civilians ensure short-term representation, their coercive legacies lead to failure over the long-run. Our findings are robust across a variety of model specifications and statistical corrections, and we show that our explanatory variables significantly increase the out-of-sample predictive performance of random survival forests.

Prepared for graduate study on Political Event Data Management

This paper conceptualizes conflict event reporting bias within an economic framework where reporting is unbiased when supply and demand for information on conflict events are at an equilibrium. When the supply of information fails to meet demand, over-reporting (i.e., fake news) occurs. Under-reporting, on the other hand, is the result of a supply surplus—when there are too many events to report, but not enough capacity for events to be reported. Reporters act as economic agents who are responsible for maintaining a supply-demand equilibrium. Because more events tend to occur than can be reported, agents’ structure of incentives drives selective coverage of big stories with high payoffs and low costs of reporting, leading to systematic under-reporting of small, costly-to-cover events as the most typical form of bias. Using conflict casualties and geographic characteristics as indicators of reporting payoffs and costs, respectively, I quantitatively examine the impact of these characteristics on reporting bias in Afghanistan and find that low-intensity events in difficult-to-reach regions are systematically under-reported.

Prepared for graduate seminar on Ethnic Politics (with Jordan Nafa)

How does ethnic violence diffuse from one location to another in civil war? In this paper, we propose a general theory of strategic ethnic targeting in civil war. Our theory suggests that by understanding how each warring ethnic group selects its targets, we can make simple assumptions about the location of observed ethnic violence; that is, we expect ethnic violence is most likely to occur where both ethnic groups have overlapping incentives to target the same location. Our theory also suggests that ethnic groups are more likely to target locations which are geographically proximate to preceding conflict. We demonstrate the viability of our model using subnational data from the First Liberian Civil War. Taken together, we provide valuable insights into the complex process by which ethnic violence diffuses in civil war.

Ethnic Geography, Targeting Strategies and the Dynamics of Ethno-Territorial Civil Wars: An Agent-Based Model with Insights from Simulated Experiments

Prepared for graduate study on Simulation Modeling in the Social Sciences

This paper presents an agent-based computer model of ethno-territorial civil war where two ethnic groups (an incumbent and challenger) compete for territory for the purpose of expanding their war effort, maximize their chances of victory, and ensure their group's survival. Rebel groups emerge from ethnic rebellions and capture territory, providing access to new sources of co-ethnic support. Each group selects a strategy for expanding their war effort, with each strategy emphasizing a different characteristic of potential targets. I propose that groups which employ an ethnically-cognizant targeting strategy (prioritizing areas where co-ethnics are concentrated) will be most the most effective, and most likely to win. Additionally, this strategy should be even more effective in when a group's settlement pattern is spatially heterogenous. Results of simulation experiments confirm that although ethnically-cognizant targeting strategies are more effective in heterogenous environments, strategies which focus on controlling the population are more effective overall because it denies recruits to an adversary. When rebel groups are concentrated in homogenous regions, a focus on carving out consolidated territories becomes a more effective way to secure an existing support base. The experiments also demonstrate how ethnic geography and targeting strategies interact to produce systematic patterns of violence over space and time, which has important implications for stakeholders and introduces new avenues for future research on the complexities of civil war.

The Diffusion of Armed Rebellion: A Space-Time Analysis of Ukraine

Presented at the Midwestern Political Science Association (MPSA) on April 5, 2018

As the conflict in Ukraine kicked off, I noticed interesting patterns regarding its diffusion and began leveraging spatial information to understand why some areas were being targeted or experiencing rebellion while others were not. I wanted insights on the extent to which local political preferences and demographic features competed with external pressures in explaining these patterns. I assembled a small team (my friend Will and I) which geo-coded thousands of Ukrainian addresses which described polling locations and results from the 2012 parliamentary election. The algorithm I developed with this data to estimate support for pro-Russian candidates has a spatial resolution high enough to distinguish local preferences across even the smallest neighborhoods. After mining Twitter to collect geo-coded reports of conflict and correcting for reporting bias, my spatio-temporal regression analysis demonstrated that areas with high concentrations of pro-Russian support were more likely to be pulled into the conflict, and sooner, than areas with lower concentrations. I also found that the diffusion of conflict stalled in areas with polarized populations, with these communities facing the highest levels of violence.

Presented at ISSS-ISAC Joint International Security Conference
Panel on Human Security, October 13, 2017; Washington, D.C.

What factors drive Russian military intervention and other foreign policy behavior? Related rhetoric indicates Russia has a self-proclaimed responsibility to protect ethnic-Russians. Drawing from the international ethnic conflict literature, I posit a strength measure of transnational ethnic ties, with larger Russian minority groups possessing stronger transnational ethnic ties with Russia. I theorize that Russia’s self-proclaimed responsibility to protect is a function of the strength of its transnational ethnic ties with Russian minority groups abroad and the intensity of conflict that threatens them. Using subnational territory data from 1991 to 2015 I estimate the effect of this explanation on the probability of Russian military intervention in post-Soviet states. Initial findings support my hypotheses, but suggest that Russian foreign policy is driven by a multifaceted explanation that includes both geopolitical and legitimate national interests, as well as human security concerns on account of political constituency mechanisms that are influenced by transnational ties. The findings also suggest that Russia is less reactive to strategic rival dynamics than is currently assumed. Russian military intervention mostly disregards intervening in conflict for the sake of human security, most broadly defined, but is highly sensitive to conflict that threatens large Russian minority groups.