Our orders are delivered strictly on time without delay
Do we always have to run a regression to estimate the treatment effect in a Differences in Differences Model?