An Averaging Alternative to Pre-Trend Testing

Nicholas L. Brown, Qiushi Bu

We study difference-in-differences (DID) estimation of average treatment effects on the treated when the researcher is uncertain about which pre-treatment periods satisfy the paral…