I am a PhD candidate in Economics at Boston College. My research interests broadly lie in econometrics, with a focus on developing robust and efficient semi-/non-parametric and machine learning methods to broaden identification and estimation strategies in causal inference and empirical research.
I am on the job market in the 2025-2026 AY.
Abstract: Regression analysis remains a cornerstone of empirical research. Most existing approaches are based on least squares which can be quite sensitive to outliers if the data distribution is far from Gaussianity. This paper proposes Distributionally Adaptive Regression Estimators (DARE), a framework designed to enhance both robustness and efficiency, particularly in settings where classical least squares methods exhibit poor performance. DARE efficiently aggregates distributional information through a weighted aggregation of quantiles, improving robustness to the unknown distribution of error terms. Under suitable regularity conditions, DARE achieves the same asymptotic variance as the infeasible likelihood-based methods while maintaining the same convergence rate as conventional estimators. A wide range of regression models are investigated. I establish asymptotic properties under both series and local linear frameworks. In addition, I develop machine learning implementations based on random forests that preserve asymptotic validity, and incorporate high-dimensional linear settings via lasso to obtain efficient estimators for the nonzero coefficients. Extensive simulations demonstrate substantial finite-sample improvements. Finally, I apply DARE in two empirical analyses of treatment effects. In both cases, DARE produces point estimates comparable to conventional methods while delivering notably tighter confidence intervals.
Distributional Relevance Instrumental Variable Estimation, (with Zhijie Xiao and Nicholas Sim)
Specification Analysis of Conditional Distribution Functions, (with Xiaojun Song and Zhijie Xiao), R&R at Econometrics and Statistics
Triangular Simultaneous Equations Without Instruments, (with Xiaojun Song and Zhijie Xiao)
Checking the Adequacy of Quantile Regression Models, (with Xiaojun Song and Zixin Yang)
Neyman-Orthogonal Goodness-of-Fit Tests for Distributions, (with Xiaojun Song), R&R at Economics Letters
Consistent Tests for Semiparametric Conditional Independence, (with Xiaojun Song), Statistics & Probability Letters, 2025, 216(1)
Detecting Heterogeneous Treatment Effects under Unconfoundedness
Testing for Essential Heterogeneity in the Marginal Treatment Effects Model, (with Yuhao Li, Minghai Mao, and Xiaojun Song)
Adaptive Wasserstein Generative Adversarial Networks
Math Camp (PhD Level), Summer 2023, 2024
Econometric Methods (TA, PhD Level), Spring 2024, 2025, 2026
Statistics (TA, PhD Level), Fall 2023, 2024
Principles of Economics (Discussion), Fall 2025
Math Camp, Summer 2018
Mathematical Statistics (TA), Spring 2019
Statistics (TA), Fall 2018
Last updated: Nov. 2025
