A Practical Guide to Event Study Plots for Political Scientists

• In Event: Advances in Causal Inference

Abstract

Difference-in-differences (DiD) designs for estimating causal effects have grown in popularity throughout political science. Many DiD papers present their central results through an ``event study plot" -- a visualization that combines estimated average treatment effects for multiple post-treatment time periods alongside placebo tests of the main parallel trends identifying assumption. Despite their ubiquity, the methods used for the creation of these plots are inconsistent and in many cases can result in misleading inferences about both the treatment effects and the placebo checks. Building on and synthesizing recent contributions in the econometric literature on differences-in-differences designs (e.g Borusyak et al., 2021; Roth, 2022; Sun & Abraham, 2021), this paper illustrates some common pitfalls through a replication of four recently published papers in major political science journals. We identify three notable problems related to the incorrect specification of the placebo group, incorrect inclusion of ``always-treated" units, and the incorrect use of covariates. We conclude with recommendations for both applied researchers and for reviewers evaluating results from event study designs.

Authors

• Anton Strezhnev, University of Wisconsin, Madison

• Zikai Li, University of Chicago