Calculate Minimum Detectable Effect (MDE) for clustered Difference-in-Differences

[Ashley Litwin]

Hi all,

I have a dataset that is at the state x year-quarter level. The main analysis is a standard (clustered on state) differences-in-differences (y ~ treat post treat*post). A reviewer wants to know if the study was powered ex ante, and specifically wants to know the MDE. I've only calculated the MDE before for a randomized experiment. Is there a way in Stata to calculate the MDE for a clustered DiD?

Thanks!