A Design Argument to not use TWFE with Differential Timing Diff-in-Diff

Short post today, but I’m going to make a short argument in favor of the newer diff-in-diff estimators, not based on heterogenous treatment effects, but on the basis of design.

When you take a design approach to causal inference, you are thinking very carefully ex ante about how you would try and replicate the randomized trial. And in the randomized tr…