I am a PhD candidate at the University of Rochester specializing in comparative politics and methodology.

My research examines media capture amid democratic erosion, focusing on how incumbents manipulate state resources to undermine the free press and how pro-government media shape their news reporting strategies. My work combines modern causal inference techniques with computational methods, drawing on a large corpus of news data that I collected during my PhD. I also share tools that I developed for my research, including CC-Downloader for collecting online news and Human Coding Dashboard for facilitating human annotation tasks.

I was a Democracy Center fellow for the 2025-2026 academic year. Before starting my PhD, I was a research associate at the University of Zurich, and I received an MA in Analytical Political Economy from Duke University and a BA in Economics and Sociology from Boğaziçi University.

Research

Working Papers

“Subsidizing the Messenger: How Soft Censorship Shapes Economic News”

Abstract

Scholars have extensively investigated the interest of incumbents in controlling media and the role of financial independence in safeguarding press freedom. Nonetheless, the literature lacks evidence on how governments employ ostensibly legal methods to fund connected media in exchange for favorable coverage. This paper identifies market-based exchanges as two-way favors between the government and politically connected media owners, exploiting the acquisition of the largest media company in Turkey by a pro-government conglomerate in 2018. First, I find sizable advertising favors from the government as the 2018 sale significantly increased state advertising in the acquired newspapers, despite a simultaneous decline in their circulations and private advertising. Second, these newspapers return the favor with biased economic news supportive of the government. My results establish that state advertising in Turkey is politically driven and incompatible with economic incentives, and the connected newspapers cater to the government, potentially at the expense of circulation.

Keywords: Media capture, government advertising, politically connected owners, triple differences regression, topic modeling, sentiment analysis, BERT, GPT

“Do Voters Prefer Women and Young Candidates? Re-evaluating Evidence from Conjoint Experiments” with Scott F. Abramson [R-Package]

Abstract

Conjoint experiments are widely used to study whether voters discriminate against candidates on the basis of characteristics such as gender and age. Recent meta-analyses interpret this literature as evidence against demand-side explanations for the under-representation of women and younger candidates in elected office. We reassess these claims by examining the finite-sample sensitivity of candidate-choice conjoint estimates. Building on recent advances in influence-function methods, we estimate the smallest share of respondents and experimentally generated electoral contests whose removal changes the sign of study-level and meta-analytic AMCEs. Across experiments in prominent meta-analyses of gender and age, we find that many estimates can be reversed by dropping small fractions of respondents or contests. These sensitivities also propagate to the meta-analytic level, where pooled estimates change sign after similarly small perturbations. The results show that conjoint-based claims about voter preferences can depend heavily on both respondent sampling and the random sampling of candidate contests implied by experimental designs.

Keywords: Gender bias, candidate-choice experiments, discrimination

Work in Progress

“Repression, Exit, and Voice in the Newsroom” with Horacio Larreguy

“Divergent News Environments: A Longitudinal Study Leveraging the Common Crawl Corpus”

“Internet Censorship and the Chilling Eﬀect on Sensitive News Coverage”

Teaching

I have served as a Teaching Assistant in the Department of Political Science at the University of Rochester. In 2024, I received the Graduate Student Teaching Assistant Award.

PSCI 238: Business and Politics

Term: Spring 2026 | Level: Undergraduate | Instructor: Professor David Primo
Course Topics: Non-market strategy for business, lobbying, corporate social responsibility.

PSCI 505: Quantitative Methods III

Term: Fall 2024 | Level: Graduate | Instructor: Professor Curtis S. Signorino
Course Topics: Maximum likelihood, survival analysis, Bayesian inference.

PSCI 282/482: Making Public Policy

Term: Fall 2024 | Level: Undergraduate / Graduate | Instructor: Professor Sergio Montero
Course Topics: Pareto concepts, externalities, coordination/commitment games.

PSCI 205: Data Analysis II

Term: Spring 2024 | Level: Undergraduate | Instructor: Professor Curtis S. Signorino
Course Topics: Bivariate/multivariate regression, maximum likelihood, experiments & causal inference.

PSCI 200: Data Analysis I

Term: Fall 2023 | Level: Undergraduate | Instructor: Professor Curtis S. Signorino
Course Topics: Descriptive statistics, distributions, confidence interval & hypothesis testing.