Could PanelMatch be extended to continuous treatments?
My uninformed guess is that it should be straightforward, at least conceptually. For example, Fong, Hazlett and Imai (2018) could be used for the matching and then difference-in-difference is well defined for continuous treatments.
Am I missing something? What other complications would there be?
P.S. Thank you for producing such a useful and intuitive software. It's super helpful!
Could PanelMatch be extended to continuous treatments?
My uninformed guess is that it should be straightforward, at least conceptually. For example, Fong, Hazlett and Imai (2018) could be used for the matching and then difference-in-difference is well defined for continuous treatments.
Am I missing something? What other complications would there be?
P.S. Thank you for producing such a useful and intuitive software. It's super helpful!