My research spans microeconomic theory, political economy, and the economics of science.
Mechanism Design and Algorithmic Game Theory
I study the behavior of no-regret learning algorithms in strategic environments, particularly pricing games. This includes investigating whether algorithmic pricing leads to supracompetitive outcomes and distinguishing between collusion and misspecification in learning dynamics.
I have derived necessary and sufficient conditions for extensive-form coarse correlated equilibria to be equivalent to Nash equilibria, with ongoing work extending these results to normal-form games.
Political Economy
In collaboration with Marco Battaglini and Tom Palfrey , I study group-based voting models and collective decision-making, applying structural methods to traditional political economy voting models. My work on policy dynamics with Giri Parameswaran examines how policy momentum and learning shape political outcomes, including when sophisticated agents can manipulate policy to induce or prevent slippery slope dynamics.
Science of Science
Joint work with Nic Fishman studies how journals should design acceptance rules when testing is cheap and p-hacking is a concern. When researchers can run many specification searches at low cost, traditional screening based on p-values breaks down. We show that journals can harness the same sequential search that enables p-hacking by requiring robustness checks: if the underlying effect is real, cheap testing makes it easy to produce many corroborating results, while false positives cannot consistently replicate across specifications.
For published work and preprints, see Papers .