Job Market Paper: "Breaking Rebel Economies: State Indiscriminate Violence During Civil War" (Under Review) Abstract: Why does indiscriminate violence against civilians by state forces weaken some rebel groups but strengthen others? The existing literature argues that indiscriminate violence is counterproductive because it increases civilians’ willingness to support the rebels. However, indiscriminate violence can also exhaust civilians’ ability to provide support. To adjudicate these competing effects, I distinguish between two strategies of indiscriminate violence: punitive and economic. I argue that the effectiveness of indiscriminate violence depends on the strategy used and the degree of rebel territorial control. Consistent with the existing literature, I find that punitive violence is rarely effective, particularly in contested areas. In contrast, economic violence used in rebel-controlled areas exhausts civilians and weakens rebel organizations. Leveraging archival interviews, process tracing, and a difference-in-difference design, I find strong support for this prediction using original data from rebel-controlled areas of the Mekong Delta exposed to economic indiscriminate violence during the Vietnam War (1961-1975).
Peer Review Articles: "Political Entrepreneurs or Bandits? The “Criminal” Origins of Peripheral Rebellions" with Janet Lewis (forthcoming at Perspectives on Politics) Abstract: How and why do armed groups that become known as “rebels” initially use violence? New datasets show that such violence is often small in scale. Numerous empirical examples indicate that it is also often ambiguous--not easily identified as a precursor to anti-state rebellion. This paper seeks to explain these patterns. We argue that a variety of fledgling nonstate armed groups find small-scale, anonymous anti-state violence useful, despite the risks. Therefore, armed groups that later become distinguishable as “rebels” or “bandits” often initially use this similar repertoire of violence. The resulting ambiguity of this violence--for outsiders from states to scholars--presents an opportunity for aspiring rebels, since states struggle to discern the threat they pose. Ambiguity lessens when aspiring rebels opt to use offensive, larger-scale violence. We illustrate our claims with three historical case studies that enable close examination of early armed group violence, as well as 12 brief case vignettes. Our analyses show the promise of integrating research on rebel origins, criminality, and state formation. Link
"Civilian Victimization During Conflict," Oxford Research Encyclopedia of International Studies, with Alexander Downes (2023) Abstract: Since the end of the Cold War, scholars have increasingly sought to explain the causes of civilian victimization—the intentional use of violence against noncombatants—during armed conflict. The question of the effects and effectiveness of violence against civilians, in contrast, has received less scholarly attention. One strand of research examines the impact of wartime civilian victimization on postconflict political behavior and outcomes. A second strand investigates the effectiveness of violence during the war itself. The principal question this literature asks is: Does civilian victimization “work”? Put more precisely, can intentionally targeting noncombatants help belligerents achieve their wartime objectives, whatever those might be? Civilian victimization takes different forms and serves different purposes in different kinds of conflicts; scholarship on its effectiveness is thus divided into work on irregular (predominantly intrastate) wars versus conventional (predominantly, but not exclusively, interstate) wars. No matter what its particular form or in which type of conflict it is used, however, civilian victimization tends to follow two broad logics: coercive and eliminationist. Most scholarship on irregular wars examines the effectiveness of coercive victimization, whereas studies of conventional war look at the efficacy of both. For example, a key debate in the literature on civilian victimization in irregular wars concerns whether selective or indiscriminate violence is more effective at deterring civilians from shifting their allegiances to the adversary. A broad consensus holds that violence is effective only when selective, but new studies have found that indiscriminate violence can also work under certain circumstances. Similarly, there is broad agreement (with some notable exceptions) in the literature on conventional war that coercive civilian victimization—which is almost by definition indiscriminate—is ineffective. In contrast, scholars have yet to assess systematically the effectiveness of eliminationist victimization in conventional war. Link
"Restrained Insurgents: Why Competition Between Armed Group Doesn't Always Produce Outbidding," Texas National Security Review, with Mark Berlin (2023) Abstract: Contemporary civil wars frequently involve numerous armed groups. How do armed groups compete with rival organizations for popular support? Existing research posits that militant organizations operating in the same conflict will often compete for support by outbidding rivals with escalatory acts of violence. However, evidence from various conflicts suggests that armed groups often forgo violent escalation in competitive environments, presenting themselves as more moderate alternatives to the local population. Armed groups may strategically limit, rather than escalate, their levels of violence during competition to differentiate themselves from rivals. In doing so, they can carve out a niche of support that differs from that of their rivals and avoid the negative backlash that may result from violent escalation. In order to advertise their relative moderation, armed groups may restrict the lethality of violence against civilians and moderate their rhetoric. Examining these arguments, we utilize Arabic-language primary sources and event-level data to analyze competition between prominent jihadist groups in Algeria (1998-2004) and Yemen (2015-2021). Link
Dissertation: Breaking Insurgent Economies: State Indiscriminate Violence During Civil War (August 2025) Abstract: State indiscriminate violence against civilians remains a tragically common feature of civil wars. Beyond direct fatalities, such violence also contributes to widespread civilian suffering through mass displacement, famine, and economic devastation. This dissertation seeks to address two key questions on states’ use of indiscriminate violence. First, why do states target civilians with indiscriminate violence during civil wars? Second, why does indiscriminate violence against civilians weaken some insurgencies but strengthen others? Prominent works on civilian victimization emphasize a punitive logic of indiscriminate violence where states seek to coerce civilians into altering their behavior. Instead, this dissertation argues that state violence often follows an economic logic that aims to exhaust insurgent opponents by disrupting the economy of insurgent-controlled areas. States use indiscriminate violence to deprive civilians the ability to support the insurgency rather than dissuade them from doing so. Further, in contrast to the conventional wisdom which argues indiscriminate violence counterproductively strengthens insurgencies, I argue that the effectiveness of indiscriminate violence depends on the logic of violence used, punitive or economic, and the level of insurgent territorial control in the targeted area. Punitive violence is counterproductive, particularly when used in contested areas. However, I conclude that economic violence targeting insurgent-controlled areas effectively weakens insurgencies.
To examines these questions, the dissertation’s empirical strategy begins with three chapters on the Vietnam War. Using quantitative and qualitative archival data, I investigate the effects of several different indiscriminate violence strategies including forced resettlement, chemical crop destruction, bombing, and collective punishments. To address the broader generalizability outside of the Vietnam case, the dissertation then identifies the use of indiscriminate violence strategies by state forces during all major insurgencies in Southeast Asia between 1825-2025. Finally, the dissertation explores the broader generalizability of the theory of effectiveness with three other case studies of foreign intervention in Asia: Japanese in China (1937-1945), French in Indochina (1945-1954), and Soviets in Afghanistan (1979-1989). I conclude by discussing implications for the political violence literature, policy alternatives to economic violence, and approaches for mitigating the effects of economic violence on civilians. Link
Working Papers: "Controlling Hearts and Minds: Forced Resettlement and Rebel Strength in Counterinsurgency" with Alexander Downes and Kelly Greenhill (Under Review) Abstract: This paper investigates the conditions under which forced resettlement—the coerced relocation of noncombatant populations to defended, government-controlled, locations—weakens insurgencies. We argue that resettlement, because it targets everyone in a given locale, does not automatically establish complete state control over relocated populations; hence, what incumbents do after resettlement matters. We contend that a policy of resource control, which lowers the probability that resettled civilians can smuggle resources to the insurgency from government-controlled areas, enables counterinsurgents to catch defecting civilians and also coerce the majority into collaborating with the state, thereby decreasing rebel strength, defined as a reduction in at least two of the following: number of insurgents, number of insurgent attacks, and insurgent territorial presence. We test our theory against two alternatives—benefits provision and violent punishment—by collecting data on every major case of forced resettlement—a total of 32—from 1816 to 2006. In a medium-N analysis, we find that resource control outperforms the alternatives. We then process trace two prominent cases of forced resettlement—the Malayan Emergency (effective) and the Strategic Hamlet Program in South Vietnam (ineffective)—and show that the causal logic of our theory operates as expected.
"Causally Identifying the Effects of Indiscriminate Violence: Re-assessing the Findings of Design-Based Inference Studies of the Vietnam War" with Alexander Downes Abstract: The use of natural experiments to isolate the causal effects of civilian victimization during armed conflicts is increasingly common. Many of these experiments rely on detailed sub-national historical data. Although possessing significant potential to advance the empirical study of violence, we argue that such causal inference designs face serious risks when they provide only general historical context and highly focused qualitative evidence for validating methodological assumptions. Neglecting attention to a design’s specific historical context obscures problems with the theoretical and casual interpretation of studies’ findings for authors and readers alike. We illustrate the pitfalls of neglecting such context by qualitatively reevaluating two prominent studies (Kocher, Pepinsky, and Kalyvas 2011; and Dell and Querubin 2018) of civilian victimization during the Vietnam War. Both studies attempt to causally identify the effect of indiscriminate violence on insurgent activity; both conclude that it is counterproductive. We demonstrate that despite the use of sophisticated statistical techniques, neither study satisfactorily defends its theoretical and causal claims, leaving the effect of indiscriminate violence—and the effectiveness of less violent, hearts-and-minds strategies—during the Vietnam War an open question.
In Progress: “Calculated Frightfulness: The Effectiveness of Collective Punishments during Civil War" “Tools of Competitive State-building: Civilian Victimization Strategies and Territorial Control” "Auditors or Looters?: The Varieties of Insurgent Taxation”