Research
Working Papers
Heterogeneous Treatment Effects Analysis via Distribution Regression based Changes-in-Changes (Job Market Paper)
Abstract
Identifying and estimating the distributional effects of a policy intervention is of key interest in economics. In analyzing heterogeneous effects of a policy on labor market or health outcomes, for example, changes-in-changes proposed in Athey and Imbens (2006) is particularly appealing. It can accommodate endogenous treatment assignment and can identify the entire counterfactual distribution. Yet, challenges with incorporating control variables to address concerns akin to differential parallel trends in the difference-in-differences literature persist. I propose a semiparametric approach to changes-in-changes based on distribution regression that can flexibly accommodate potential observed confounders and can be applied to both continuous and/or discrete outcome variables. I derive large sample theory for the distribution regression based changes-in-changes estimator and for the functionals thereof. These include unconditional distributional and quantile treatment effects, average treatment effects, and decompositional treatment effects for the treated group. Bootstrap validity is also demonstrated for conducting inference in practice. Lastly, I apply the approach to study the heterogeneous effects of Earned Income Tax Credit on infant weights and find that the policy had higher concentrated benefits for lower birth weights and more muted effects across the birth weight distribution than previously reported.
Work-in-Progress
- Flexible Distribution Regression using Neural Networks (with V. Chernozhukov, I. Fernandez-Val, V. Quintas-Martinez)
- Panel Data Quantile Regression with Grouped Fixed Effects