Working Papers:

"The Protest Dilemma: The Effect of Threat of War in Sustaining Non-Democratic Regimes" (with Mariam Malashkhia) (Job Market Paper) [Accepted at APSA 2025 and Comparative Politics and Formal Theory Conference 2025]

In non-democratic regimes, where elections do not provide a viable path to regime change, protests serve as the primary means of voicing political discontent. While there is a growing body of literature on protests, the role of external factors, such as threat of war, remains underexplored. This gap is particularly significant given that many non-democratic regimes operate under the persistent threat of invasion.  

In this paper, we develop a formal model showing that the threat of invasion can deter protests, ultimately benefiting the incumbent. Thus, our findings help to explain why regimes manipulate media to exaggerate invasion risks. The core mechanism driving this result is that citizens only gain from regime change if the country is not invaded, making protest a risky gamble. 

Several factors amplify this demobilization effect, including the likelihood of a successful invasion, the enemy’s strong incentives to intervene, and the low perceived costs of war. 


“Politics of Non-Full Democracies: Fraud, Post-Election Boycotts, and Protests” (with Mariam Malashkhia) [R&R at APSR]

Non-democracies are often characterized by electoral manipulation and contested results. Despite a growing literature on these regimes, opposition strategies remain underexplored. To address this gap, we develop a formal model analyzing the interactions among the incumbent’s, the opposition’s, and the citizens’ decisions. We show that a boycott --- forfeiting parliamentary seats --- can serve as a powerful signal of the opposition’s ability, thereby mobilizing large-scale protests.  This new information explains why citizens may support the regime one day but protest the next. Thus, in contexts where the opposition faces public outreach constraints, a post-election boycott becomes one of the few available and effective strategies. Given the boycott’s significance, we explore the factors that influence this decision.

Moreover, our model offers a new perspective  on the strategic role of fraud --- shaping post-election dynamics. Additionally, we argue that  peaceful transitions arise when credible threats of protests are coupled with parliamentary elections.


"Political Incentives for Health Insurance Provision" [Accepted at APSA 2025]

Adverse selection in private health insurance markets generates well-known welfare losses and is a standard rationale for government intervention. Yet such intervention is shaped by electoral incentives: voters care deeply about health care, and political parties compete on promises of coverage and subsidies. This paper develops a formal model of strategic interaction between political parties, private insurers, and voters. Parties compete over the level of mandatory social insurance—financed through uniform taxation—while individuals can purchase additional coverage in the private market. The analysis yields two main results. First, when the private market is already inefficient due to adverse selection, mandatory public insurance can either mitigate or exacerbate this inefficiency by altering the composition of the private risk pool. Thus, once political constraints are accounted for, intervention is not necessarily welfare-enhancing—even when market failure is present. Second, I characterize the welfare gap between the politically chosen and socially optimal levels of coverage and show how it varies with key economic and political parameters.


"The Effect of Internal Displacement on Political Attitudes" (with Mariam Malashkhia) [Junior Scholars in Quantitative Conflict Workshop]

How does war affect political attitudes? While much research on war centers on violence, the effects of displacement remain underexplored, despite over 120 million people being forcibly displaced.  This omission is significant, as displacement represents a unique form of victimization with long-lasting impacts. We address this gap by examining Georgians displaced by the 2008 Russian invasion. Using a natural experiment, we find that while displaced and non-displaced individuals demonstrate similar levels of support for Georgia joining NATO and the EU, displaced individuals are more likely to hold hawkish attitudes when questions are directed at those responsible for their displacement — in this case, Russia. However, displaced individuals, particularly those who have experienced violence, exhibit a heightened sense of solidarity with others who have endured similar hardships, namely with Ukrainians. In a world increasingly shaped by forced migration, understanding how displacement influences political preferences, fosters solidarity, and affects trust in institutions to mediate conflicts is critically important.


"How did the Russo-Ukrainian War Affect Voting Behavior and Party Affiliation in the US?" (with Mariam Malashkhia) [Accepted at MPSA 2025 and APSA 2025]

Per the "Almond-Lippmann consensus", the public is uninformed, has incoherent opinion on foreign policy and is irrelevant to decision making. However, after addressing the endogeneity issue, we find that support for Ukraine has a strong negative effect on identifying as a Republican. Moreover, the support for Ukraine, a foreign policy issue, has a larger effect than the variables consistently found in the literature to have a large and significant effect on party identification. As for identifying as a Democrat, the results from the probit model show that when other, potentially exogenous, variables are included as controls, the support for Ukraine has a strong, significant, and positive effect. However, when we add other, possibly endogenous variables, the effect disappears. The strong negative effect of Ukraine support on identifying as a Republican contradicts most of the existing work that foreign policy is a low salience issue.


Selected Work in Progress:

"Insurance Provision with Electoral Incentives and Market Power" (with Santiago Seoane)