Research
You can also find my articles on Google Scholar.
Published articles and book chapters are listed on the Publications page. Below, working papers are grouped by topic.
War exposure and innovation
Public Support for Long-term Military Investments during Wartime: Evidence from Ukraine
(with Vera Mironova and Sam Whitt.)
When one’s country goes to war, there is a common “rally ‘round the flag” effect in public opinion, leading to both greater support for the government’s role in the economy and increased support for the military. In this study, we offer a novel perspective on the “rally ‘round the flag” effect by examining its long-term implications for the direction of the country’s economic development. We focus on the changes in preferences toward future investments in new military versus civilian technologies as war progresses. In a two-wave survey in Ukraine, we observe declining support for government resource allocation toward military innovations. This time-related decline is explained by increasing prioritization of other sectors of the economy, which we interpret as a resource diversification strategy. Individuals are less willing to invest heavily in long-term military innovations when victory seems doubtful and instead hedge by funding civilian sectors. Results are robust to controls for individual and settlement-level victimization and conflict exposure. Our findings have important implications for post-war security, reconstruction, and development.
War Exposure and Public Demand for State-led Innovation: Evidence from Ukraine
(with Vera Mironova.)
April 2026.
War may increase the political feasibility of state-led innovation by making technological capacity more visible and urgent. Using wartime surveys from Ukraine with fixed-sum budget tasks, we compare respondents more directly exposed to war with observationally similar respondents from the same pre-war settlements. Direct exposure raises the share of a reconstruction budget allocated to R&D by 3–5 percentage points. Economic exposure mainly crowds out healthcare, while family harm crowds out infrastructure investment. Exposure reduces support for importing technology, with little change in broad technology attitudes or support for war effort, suggesting reprioritization under scarcity rather than generic rally effects.
Ethics and survey research
Public Perspectives on War Trauma Research: Evidence from Ukraine
(with Vera Mironova and Sam Whitt.)
Many scholars have discussed the ethics of conducting survey research on sensitive issues of war trauma, but less is known about public perceptions of acceptable and unacceptable research practices. This study examines public opinion on the ethical conduct of research in a May 2025 survey of 2,000 respondents in Ukraine. Using a randomized survey experiment that varies the researcher’s identity, whether harm and trauma inquiry focuses on sexual violence, and pre-treatment conditioning on perceptions of rapport with researchers (trust, comfort, willingness to be forthcoming in responses), we find that who is asking questions (identity) is less important than what is being asked (sexual violence) and how it is asked (perceived rapport with researchers). Public concern about war trauma research is most clearly mitigated by perceived rapport with researchers. Our results reinforce how lessons from qualitative research could be more effectively applied to survey and experimental best practices.
Beliefs, media, and political communication
When Opinions Pass for Facts: Self-Esteem and Motivated Classification in Media Environments
(with Alexey Zakharov, Minh Dao, and Sydney Brown.)
Under review.
Modern political communication is saturated with claims that blur the line between evidence and interpretation. Citizens encounter opinion-based content not only on social media, but also in mainstream newspapers, newsletters, podcasts, television panels, and campaign messaging. When such claims are accepted as facts, people do not merely agree with a speaker; they update beliefs and policy judgments as if they had learned something objective about the world. Motivated-beliefs models suggest one reason this may happen: people sometimes distort what they accept as true because those beliefs protect self-image or identity (Bénabou and Tirole, 2002, 2016; Bracha and Brown, 2012). We study the reverse implication of that logic in a pre-registered experiment. If self-protection helps sustain self-deception, then temporarily bolstering self-image should reduce the willingness to treat congenial opinions as facts. To test this, we perform a survey experiment manipulating self-esteem. Respondents first complete a positive, negative, or neutral social-media-style task and then classify 24 short statements as facts or opinions. Statements vary by type, source, topic, and ideological valence. In the control group, respondents are substantially more accurate on opinions than on facts, and classification errors are systematically related to statement content rather than random noise. Positive treatment does not improve accuracy overall. In our preferred directional specification, however, it reduces by about 3.2 percentage points the tendency to classify congruent opinions as facts, with no corresponding increase in time spent on the task. Neutral-item placebo tests are close to zero, while fact-side effects are less uniform. These findings identify a demand-side mechanism through which self-image can shape what citizens count as evidence. The broader political-economy implication is that actors may influence not only what people prefer, but which claims people are willing to treat as factual in the first place.
Privacy, regulation, and digital markets
Searching for Data Privacy: Effects of the General Data Protection Regulation on Firm Ad Revenue and Data Privacy Concerns
(with Kristine Thomas.)
Modern websites optimize performance and target ads using personal data, enhancing consumer convenience but risking consumer privacy and data agency if done without consent. The EU’s General Data Protection Regulation (GDPR) mandates this consent and data protection. This study examines whether GDPR enactment reduced public concerns over data privacy using Google search data. First, we confirm that online search history reflects privacy concerns. To do so, we replicate prior findings that GDPR negatively impacted ad revenue and demonstrate that it positively affected the probability of ad-revenue-related searches. We also provide illustrative evidence that the two move in opposite directions. We find that user concerns over data privacy decline over time, amplified by increased online activity during COVID-19. Interestingly, we find that the initial drop in ad revenue post-GDPR was followed by partial recovery. These findings suggest privacy regulations can alleviate public concerns without long-term harm to the ad industry.
R&D, innovation, and intellectual property
Patent Protection and Innovative Entrepreneurship: Examining Effect of Alice Ruling on Startup Funding
(with Maxim Ananyev.) SSRN
We study the effects of patent protection on innovative entrepreneurship. We use exogenous variation in patent protection from the 2014 U.S. Supreme Court ruling in Alice Corporation Pty. Ltd. v. CLS Bank International, which established a two-part test for software patentability and sharply changed litigation incentives for software patents while leaving other fields largely unchanged. Using a database of U.S. startups, we show that weaker protection in software-related industries made it harder for firms to obtain startup funding or reach an IPO, and affected entry decisions for new startups.
Authoritarian politics, propaganda, and crisis
Information Manipulation and Repression: Theory and Evidence from the COVID-19 Response in Russia
Were COVID-19 and the associated restrictions used by authoritarian governments to tighten their grip on power? Using data from 83 Russian regions, we show that, yes, they did—the extent of information manipulation and political repression, two main tools of authoritarian regimes, were influenced by the strength of local civil society and institutions. Repression complemented propaganda: more politically motivated arrests were associated with an increase in information manipulation. Repression waves that followed the poisoning of Alexey Navalny, the opposition leader, in August 2020 and his arrest in January 2021 were more pronounced in those regions that manipulated the COVID-19 statistics more. The increase in the usage of authoritarian tools came at a price: misinformation reduced compliance with pandemic restrictions. In addition to two-way fixed effects, we use sensitivity analysis to account for possible omitted variables. Our findings confirm, both theoretically and empirically, the complementarity between propaganda and repression as instruments of authoritarian control.