Job Market Paper
Optimal treatment assignment rules under capacity constraints (with Kohei Izumi), available at arXiv:2506.12225v2.
Abstract: We study treatment assignment under capacity constraints, where a planner aims to maximize social welfare by assigning treatments based on observable covariates. Such constraints are common when treatments are costly or limited in supply, but they complicate the analysis of optimal assignment rules because assignment probabilities must be coordinated across the entire covariate distribution. We develop a new approach that reformulates the planner’s problem as an optimal transport problem, which makes the constraints analytically tractable. Using a limits of experiments framework, we establish local asymptotic optimality results for two assignment rules, a Bayesian-type rule and a simple plug-in rule. We show that the former rule can dominate the latter rule, with simulations demonstrating sizable risk reductions. Finally, we illustrate the method with data from a Colombian school voucher program (Angrist, Bettinger, and Kremer, 2006).
Presentations: Midwest Econometrics Group at UIUC, Canadian Econometrics Study Group in Ottawa (poster)*, ES European Winter Meeting in Nicosia, Cyprus*. (*=scheduled)
Research
- Closed-form estimation of additively separable triangular equation models. (to appear soon)
- Applications of cross-fit variance estimator for model specification, testing overidentification, and structural parameter hypotheses (with Yukitoshi Matsushita and Taisuke Otsu), available at [.pdf].
- On large market asymptotics for spatial price competition models (with Taisuke Otsu), Economics Letters (2024), 234, 111468. doi: 10.1016/j.econlet.2023.111468. [.pdf] [Appendix]
- A characterization of the Esteban–Ray polarization measures (with Yoko Kawada and Yuta Nakamura), Economics Letters (2018), 169, 35–37. doi: 10.1016/j.econlet.2018.05.011.