I recently received my PhD in Econometrics and Statistics from the University of Chicago Booth School of Business, where I was mentored by Azeem Shaikh and Christian Hansen between 2020 and 2024. In Autumn 2024, I joined Amazon as a Postdoctoral Scientist, where I work on panel data methods and machine learning for policy evaluation with Eric Tchetgen Tchetgen.
My research focuses on causal inference and econometrics, with a particular interest in the design and analysis of randomized experiments.
I am currently on the 2024-2025 academic job market.
Working Papers
Inference for Two-stage Experiments under Covariate-Adaptive Randomization (Job Market Paper)
Revision Requested at the Journal of Econometrics.
Abstract
This paper studies inference in two-stage randomized experiments under covariate-adaptive randomization. In the initial stage of this experimental design, clusters (e.g., households, schools, or graph partitions) are stratified and randomly assigned to control or treatment groups based on cluster-level covariates. Subsequently, an independent second-stage design is carried out, wherein units within each treated cluster are further stratified and randomly assigned to either control or treatment groups, based on individual-level covariates. Under the homogeneous partial interference assumption, I establish conditions under which the proposed difference-in-"average of averages" estimators are consistent and asymptotically normal for the corresponding average primary and spillover effects and develop consistent estimators of their asymptotic variances. Combining these results establishes the asymptotic validity of tests based on these estimators. My findings suggest that ignoring covariate information in the design stage can result in efficiency loss, and commonly used inference methods that ignore or improperly use covariate information can lead to either conservative or invalid inference. Then, I apply these results to studying optimal use of covariate information under covariate-adaptive randomization in large samples, and demonstrate that a specific generalized matched-pair design achieves minimum asymptotic variance for each proposed estimator. Finally, I discuss covariate adjustment, which incorporates additional baseline covariates not used for treatment assignment. The practical relevance of the theoretical results is illustrated through a simulation study and an empirical application.
Semiparametric Estimation of Treatment Effects in Observational Studies with Heterogeneous Partial Interference (with Zhaonan Qu, Ruoxuan Xiong and Guido Imbens)
Revision Requested at the Journal of Business & Economic Statistics.
On the Efficiency of Finely Stratified Experiments (with Yuehao Bai, Azeem Shaikh and Max Tabord-Meehan)
We study the efficient estimation of a large class of treatment effect parameters that arise in the analysis of experiments.
Publications and Forthcoming Papers
Inference in Cluster Randomized Trials with Matched Pairs (with Yuehao Bai, Azeem Shaikh and Max Tabord-Meehan)
Journal of Econometrics, 245(1), 105873. (2024)
Inference for Matched Tuples and Fully Blocked Factorial Designs (with Yuehao Bai and Max Tabord-Meehan)
Quantitative Economics, 2024, 15(2), 279–330. (2024)
Revisiting the Analysis of Matched Pair and Stratified Experimental Designs in the Presence of Attrition (with Yuehao Bai, Meng Hsuan Hsieh, and Max Tabord-Meehan)
Journal of Applied Econometrics, 2024, 39(2), 256–268. (2024)
Proximal Causal Inference for Synthetic Control with Surrogates (with Eric J. Tchetgen Tchetgen and Carlos Varjão)
The 27th International Conference on Artificial Intelligence and Statistics (AISTATS). 2024
Learning Intuitive Policies Using Action Features (with Mingwei Ma, Samuel Sokota, Max Kleiman-Weiner, Jakob Foerster)
International Conference on Machine Learning (ICML), 2023
Work in Progress
Randomization Inference for Two-Sided Market Experiments (with Azeem Shaikh and Panos Toulis)
Abstract
Randomized experiments are increasingly employed in two-sided markets, such as buyer-seller platforms, to evaluate treatment effects from marketplace interventions. These experiments, including the recently introduced Multiple Randomization Designs (MRDs), must reflect the underlying market structure, making them particularly challenging to analyze. In this paper, we propose a randomization inference framework to analyze outcomes from such two-sided experiments. Our approach is finite-sample valid under sharp null hypotheses for any test statistic and maintains asymptotic validity under weak null hypotheses through studentization. Moreover, we provide heuristic guidance for choosing among multiple valid conditional randomization tests to enhance statistical power. Finally, we demonstrate the performance of our methodology through a series of simulation studies.