Dynamic Treatment Effect Estimation with Interactive Fixed Effects and Short Panels
We study inference on dynamic average treatment effect parameters for staggered interventions when parallel trends are only valid conditional on unobserved interactive fixed effects. Our identification strategy allows for any first stage system of moments that controls the column space of the unobservable trends including principal components, common correlated effects, quasi-differencing, and more. This result applies to data sets with either many or few pretreatment time periods.