Staggered Difference-in-Differences Estimation for Antitrust Analysis: A Review of Literature and Recommendations for Practitioners

Abstract The aim of this paper is twofold: first, we discuss literature developments surrounding difference-in-differences (DiD) methods with staggered treatment mechanisms. Second, we provide a resource for sound DiD analysis in antitrust expert testimony in light of these developments. We review relevant papers and their most important conclusions. We then discuss the antitrust implications of three important topics: parallel trends, the not-yet-treated group, and data with customer entry and exit. We supplement this discussion with Monte Carlo analysis, in which we compare the performance of DiD estimators and quantify certain types of bias. Finally, we discuss the sensitivity and robustness checks that should underlay expert testimony going forward. DiD theory has come a long way in the academic literature since the 2010s, and we distill that knowledge into what we consider to be the standards for robust DiD results going forward.