A0802
Title: A t-test for synthetic controls
Authors: Yinchu Zhu - Brandeis University (United States) [presenting]
Victor Chernozhukov - MIT (United States)
Kaspar Wuthrich - University of California San Diego (United States)
Abstract: A practical and robust method is proposed for making inferences on average treatment effects estimated by synthetic controls. A K-fold cross-fitting procedure is developed for bias correction. To avoid the difficult estimation of the long-run variance, the inference is based on a self-normalized t-statistic, which has an asymptotically pivotal t-distribution. The t-test is easy to implement, provably robust against misspecification, valid with non-stationary data, and demonstrates excellent small-sample performance. Compared to difference-in-differences, the proposed method often yields more than 50\% shorter confidence intervals and is robust to violations of parallel trends' assumptions. An R-package for implementing the methods is available.
