Research
Recent Submissions
Panel Instrumental Variable Regression Models with Varying-intensity Repeated Treatments: Theory and the China Syndrome Application with Jaerim Choi and Dakyung Seong
Abstract: Instrumental variable models with repeated endogenous treatments are popular in empirical research using pooled cross-sectional or short panel datasets. This paper proposes a novel semi-parametric approach that explicitly considers treatment effect dynamics by allowing for 1) path-dependency in the contemporaneous treatment effect and 2) a direct carryover effect from last period's treatment. We show that if either of these new features is present and external instruments are serially correlated, the textbook two-stage least-squares estimator is invalid. We apply our proposed estimation and inference approach to revisit the work of Acemoglu et al. (2016). Using their industry-level data, we find that the contemporaneous impact of increased Chinese import competition on US manufacturing employment depends on the import exposure of the preceding period. In particular, the trade shock an industry received over 1991-1999 magnifies the negative impact of the trade shock over 1999-2011. The magnifying pattern is mild until the import exposure increase over 1991-1999 exceeds a certain threshold.
Dynamic Regression Discontinuity under Treatment Effect Heterogeneity with Yu-Chin Hsu, conditionally accepted
Abstract: Regression discontinuity (RD) is a popular tool for the analysis of economic policies or treatment interventions. This research extends the classic static RD model to a dynamic framework, where observations are eligible for repeated RD experiments and, therefore, treatments. Such dynamics often complicate the identification and estimation of longer-term average treatment effects. Previous empirical papers with such designs typically ignored the dynamics in the model or adopted restrictive identifying assumptions. This paper presents identification strategies under various sets of weaker identifying assumptions and proposes associated estimation and inference methods. The proposed methods are applied to revisit the effect of California local school bonds in the seminal study of Cellini et al. (2010).
Publications
Inference on Optimal Treatment Assignment with Tim Armstrong, Japanese Economic Review , 2023
Instrumental Variable Estimation With First-stage Heterogeneity with Alberto Abadie and Jiaying Gu, Journal of Econometrics, 2023
- Testing Monotonicity of Conditional Treatment Effects under Regression Discontinuity Designs with Yu-Chin Hsu, Journal of Applied Econometrics, 2021
A Censored Maximum Likelihood Approach to Quantifying Manipulation in China's Air Pollution Data with Dalia Ghanem and Junjie Zhang, Journal of Association of Environmental and Resource Economics, 2020
Two-sample Instrumental Variables Regression with Potentially Weak Instruments with Jaerim Choi, Stata Journal, 2020
Stata addon weaktsiv can be installed on Stata
Testing for Treatment Effect Heterogeneity in Regression Discontinuity Design with Yu-Chin Hsu, Journal of Econometrics, 2019
Weak-instrument Robust Inference for Two-sample Instrumental Variables Regression with Jaerim Choi and Jiaying Gu, Journal of Applied Econometrics, 2018
Testing for Rank Invariance or Similarity in Program Evaluation with Yingying Dong, Review of Economics & Statistics, 2018
Oracle and Adaptive False Discovery Rate Controlling Methods for One-sided Testing: Theory and Application in Treatment Effect Evaluation with Jiaying Gu, Econometric Journal, 2018
Estimation and Inference of Distributional Partial Effects: Theory and Application Journal of Business & Economic Statistics, 2019
Distributional Tests for Regression Discontinuity: Theory and Empirical Examples with Xiaohan Zhang, Review of Economics & Statistics, 2016
Estimation of Censored Panel-data Models with Slope Heterogeneity with Jason Abrevaya, Journal of Applied Econometrics, 2014
Working Papers/Projects
Long-run Direct vs. Total Effects in Panel Models with Endogenous Varying-intensity Treatments”, with Jaerim Choi and Dakyung Seong
Lack of Identification Diagnostics in GMM Contexts with Stephen G. Donald and Yu-Chin Hsu
Decomposing the Average Treatment Effect in Regression Discontinuity Models with Yu-Chin Hsu and Ying-Ying Lee
High-dimensional Regression Discontinuity Estimation Away from the Cutoff, with Yu-Chin Hsu