[Aayush Bakshi]

Jared Greathouse

Thanks for your reply,

With the inclusion of the first 1-way interaction term, I am hoping to allow some heterogeneity in the treatment effects by treatment cohort and calendar time. This is to show that the initial effect of Uber may grow the longer that it has been in a city. Does that make sense in what I am doing? Trying to include what Jeff Wooldridge mentioned in his reply