search

insongkim / PanelMatch

117 stars 35 forks source link

Treatment reversal #77

Closed guilhermekirch closed 2 years ago

guilhermekirch commented 3 years ago

Hi, I have a question about your very useful package. When I want to study a treatment reversal (like the authoritarian reversal in the paper) I need to redefine the treatment variable or is there a parameter in the PanelMatch function that control this behavior? Thank you in advance!

adamrauh commented 3 years ago

Hi there, thanks for using the package. It sounds like you might be interested in setting qoi = atc.

guilhermekirch commented 3 years ago

Hi Adam, thank for your reply. I think I was not clear in my question. I will use the dem event to ilustrate my point. The atc (with dem as treatment variable) is the effect that democratization would cause on a country that remains authoritarian (dem_(i, t) =0, dem(i, t-1) = 0, relative to another country with dem(j, t) = 1, dem(j, t-1) = 0). The treatment reversal is the att of a country that reverses from democracy to authoritarian (dem(i, t) = 0, dem(i, t-1) = 1, relative to another country with dem(j, t) = 1, dem(j, t-1) = 1). If I am right, it is necessary to redefine the treatment variable to do this analysis. It would be nice if your package has a function to automate this process.

adamrauh commented 2 years ago

A year late, but I believe this has now been solved in subsequent versions of the package. If I understand correctly, @guilhermekirch is describing what is now implemented as the "art". We previously had this incorrectly labeled as the atc. Now, the att, atc, art, and ate are all correctly implemented (hopefully).

guilhermekirch commented 2 years ago

Dear Adam: thank you very much!!