Research

Job Market Paper

• Civilian Loyalties, Curfews, and Legibility in Counterinsurgency. Under Review.
  • Abstract

    States often implement population control measures to combat insurgencies, but how do they decide where to impose such measures, and how do these policies reshape counterinsurgency dynamics? This study examines curfews—ostensibly nonviolent yet indiscriminate restrictions on civilian movement—as a tool to enhance state legibility and information flows. While curfews improve government oversight, they also disrupt civilian life, creating a tradeoff between security and public backlash. Using Turkey’s counterinsurgency campaign against the Kurdish insurgency as a case study, I find that curfews trigger civilian displacement, which mitigates the state’s information problem by altering the demographic composition of conflict zones. This facilitates intensified military operations in depopulated urban areas. Additionally, I show that curfew deployment is shaped by civilian loyalties: the government imposes more curfews in insurgent-dominated districts and fewer in competitive and government-dominated districts. These findings contribute to broader debates on state control, legibility, and the strategic use of repression in counterinsurgency campaigns.

Working Papers

Drafts are available upon request.

• Pave the Way to Diversion: Increased Saliency of Foreign Policy at Times of Economic Hardships. (with Efe Tokdemir). Under Review.
  • Abstract

    Most research on the diversionary use of force focuses on timing of conflict initiation and target selection, often treating the public as passive observers of the process. This research note challenges this assumption by examining how leaders try to configure most-needed public attention before resorting to use of force for diversionary purposes. We contend that economic downturns that trigger leaders' diversionary attempts require them first to boost salience of foreign policy issues in the eyes of the public. We measure foreign policy salience in U.S. presidential speeches from 1945 to 2019 using a large language model (LLM), and examine its mediating role between economic conditions and aggression in international politics. Moreover, rather than using quarterly or yearly data, we utilize monthly-level data for economic indicators as well as political outcomes to acquire a finer-grained view of causal story. The results are in line with our theoretical expectation, while not supporting existing accounts of diversionary theory: we do not find a direct effect of inflation on use of force abroad; whereas, inflation is positively associated with increased levels of foreign policy salience in presidential speeches, which in turn increases the likelihood of use of force. Our findings suggest that leaders first elevate foreign policy issues rhetorically before resorting to force; and hence, it can be a signal for upcoming aggression abroad. These findings highlight the importance of public priming in the diversionary use of force, and offer a new perspective on the role of presidential rhetoric.
• Understanding Elite Defections in Competitive Authoritarian Regimes: Evidence from Turkey. (with Hilal Sert). Under Review.
  • Abstract

    What role can voters play in influencing elite behavior and checking democratic backsliding? We argue that credible signals of voter discontent can trigger defections from ruling-party elites. While most research emphasizes how polarization limits voters’ ability to constrain autocrats, we highlight a different channel: voters can shape elite incentives by signaling dissatisfaction with undemocratic practices in subnational elections. Using survival analysis on an original dataset of Turkey’s ruling party (AKP) legislative candidates from 2002 to 2023, we find that voter backlash, following the attempt to overturn the 2019 Istanbul mayoral election, significantly increased elite defections. We further show that elites’ organizational embeddedness and autonomy shape their responses: those deeply tied to party structures remain loyal, while high-profile, such as former ministers, are more likely to defect. Our findings highlight that voter signals, by reducing elite uncertainty and reshaping strategic calculations, can weaken ruling parties internally and create openings for opposition actors and opportunities for democratic resilience.
• Insurgent Sabotage of Effective Local Governments: Evidence from Turkey.
  • Abstract

    Conventional counterinsurgency theory holds that effective local governance should reduce insurgent violence by weakening civilian allegiance to armed groups. Yet insurgents are not passive observers of governance improvements. I argue that when municipal control shifts to parties with strong service-delivery capacity, insurgents strategically escalate violence to prevent declining popular support. In particular, they increase attacks on security forces to provoke coercive counterinsurgency responses that alienate civilians from the state, thereby undermining the political benefits of improved public goods provision. I test this argument in the context of Turkey’s long-running conflict with the PKK, focusing on the 1994 municipal elections in which the Refah Party (RP) gained control in several Kurdish-majority districts. Using a close-election regression discontinuity design, I show that districts narrowly won by the RP experienced significantly higher levels of PKK attacks on security forces. This pattern is consistent with a sabotage–provocation mechanism in which insurgents respond to capable local governance by inducing enemy-centric counterinsurgency. These findings advance research on political violence by conceptualizing local governments as autonomous actors in counterinsurgency and linking literatures on development aid and provocation.
• Framing Domestic Extremism: How Labels Influence Public Support for Counterterrorism Measures.
  • Abstract

    How does the labeling of violent domestic groups influence public support for counterterrorism policies, and how do partisan identities shape those effects? This study examines how Americans respond to domestic violent actors labeled as terrorists or extremists, arguing that labeling effects are conditioned by partisan alignment. Using a pre-registered survey experiment conducted in the United States, I first document partisan filtering in support for counterterrorism measures based on the perpetrating group’s ideology. I then test whether labeling the group as terrorist or extremist mitigates this bias. However, I find no consistent evidence that labeling increases support for counterterrorism responses. Instead, partisan identity and ideological proximity strongly shape attitudes toward counterterrorism measures. Further analysis shows that these divides operate through perceived threat, perceived credibility of the attack, perceived personal risk, and moral justification of violence. The findings suggest that elite and media framing may exert limited influence without deeper shifts in partisan perceptions, highlighting affective polarization’s central role in shaping responses to domestic extremism.
• Beyond Institutions: How Informal Alliances Shape Public Backing for Military Interventions.
  • Abstract

    How does collaboration with allies shape public support for overseas counterinsurgency operations? While much existing research on support for military interventions emphasizes formal institutions like the UN or NATO, this study examines ad hoc alliances, which are purpose-built and single mandate coalitions that operate outside formal multilateral institutions. Using a preregistered U.S. survey experiment, I find that collaboration with ad hoc alliances significantly increases support for troop deployments compared to unilateral action. However, it does not shift beliefs about threat, necessity, cost, or likely success. Instead, collaboration appears to function symbolically, resonating most with individuals predisposed toward internationalism and supportive of NATO. These findings challenge informational accounts of multilateralism and highlight the legitimating role of symbolic cues in shaping public opinion. This research note broadens our understanding of how governments can use even ad hoc/informal coalitions to secure domestic backing for military interventions.

Selected Works in Progress

• Foreign Meddling by Adverse Side-Taking. (with Eddy Yeung).
  • Abstract

    Foreign electoral intervention is an increasingly popular tool for authoritarian superpowers to influence politics in the international system. This paper conceptualizes a strategy of foreign meddling: adverse side-taking. We argue that foreign rivals can tip the balance of electoral support in favor of their preferred candidate by overtly endorsing another candidate. This strategy can be effective because rivals' endorsement can signal to domestic voters that the endorsed candidate, if elected, will likely make inappropriate foreign policy decisions for their country. To offer a first-cut empirical test of our argument, we administered a pilot survey experiment to American citizens and found that---consistent with the core prediction of our theory---candidates endorsed by US rivals suffered significant electoral backlash. Building on our pilot study, we design and preregister a complementary pair of candidate-choice and vignette-based experiments to unpack the impact, mechanism, and scope of adverse side-taking.
• Unity on the Pitch, Unity in the State? Football and Territorial Preferences in Spain.
  • Abstract

    How do major national rituals such as international football matches affect political attitudes on territorial organization? This paper investigates whether exposure to Spanish national team matches influences individual preferences over decentralization and national identification. Leveraging the timing of football matches and a rich panel dataset of political attitudes in Spain, I employ a quasi-experimental design that relies on the as-if random timing of matches relative to the survey fieldwork. I find that exposure to a national match significantly increases identification with Spanish identity and reduces support for greater regional autonomy. The findings contribute to our understanding of how seemingly irrelevant events shape attitudes toward national unity and political structure.
• Constituents in Peril: How Non-Violent Government Presence Impacts Insurgent Tactics.
  • Abstract

    Insurgents may risk their constituents' lives and protect themselves by conducting civilian-endangering attacks. Existing scholarship emphasizes the costs of civilian victimization on the responsible warring party. Extending this line of research, I ask under what conditions insurgent groups accept these costs, risking their own constituents' lives. In this research note, I argue that insurgents produce more civilian-endangering attacks when their constituents are more likely to support the insurgency because of a lack of political channels. I test this theory using micro-level event data on construction and violence in Iraq from 2004 to 2009. I show that the increase in non-violent government presence, such as mayoral offices or government centers, changes the composition of insurgent tactics, leading them to engage in civilian-endangering attacks less within regions densely populated by their constituents. The findings offer a new framework to understand insurgent violence against civilians, emphasizing the need for alternative non-violent channels to reduce civilian victimization.