Support for Difference-in-Difference/Synthetic Control/RD Methods
amichuda opened this issue · comments
Not sure if this falls in line with where you want this package to go, but given the recent slew of papers in the diff-in-diff literature, I was wondering if adding support for diff-in-diff directly (as well as the various ways of estimating treatment effects), could be a cool enhancement. But not if you think of diff-in-diff as a standalone method on the same level as a panel regression or whether diff-in-diff is "a method implemented using panel regression."
The same goes for synthetic control and RD. Would be happy to send a pull request for basic support.
Assuming, you're interested, one thing that might be needed would be helper functions to get the data into the right form needed for diff-in-diff/event study/ RD?
I think these would be a great fit since they are (nearly) linear models. Happy to review/accept PRs.