How Much Should We Trust Staggered Difference-In-Differences Estimates?

How Much Should We Trust Staggered Difference-In-Differences Estimates?

Andrew C. Baker, David Larcker, Charles Wang

Series number :

Serial Number: 
736/2021

Date posted :

February 28 2021

Last revised :

January 27 2022
SSRN Share

Keywords

  • Difference in differences • 
  • staggered difference-in-differences designs • 
  • generalized difference-in-differences • 
  • dynamic treatment effects

We explain when and how staggered difference-in-differences regression estimators, commonly applied to assess the impact of policy changes, are biased. These biases are likely to be relevant for a large portion of research settings in finance, accounting, and law that rely on staggered treatment timing, and can result in Type-I and Type-II errors.

We summarize three alternative estimators developed in the econometrics and applied literature for addressing these biases, including their differences and tradeoffs. We apply these estimators to re-examine prior published results and show, in many cases, the alternative causal estimates or inferences differ substantially from prior papers.

Published in

Published in: 
Publication Title: 
Journal of Financial Economics (JFE), Forthcoming

Authors

Real name:
Andrew C. Baker