When Covariates Mask Pre-Trends

Honest difference-in-differences methods judge how far parallel trends may fail after treatment by how far it failed before treatment, as shown by pla…

Continue in Briflio to read, react, comment, and share.