xuyiqing
commented
3 years ago This may be due to the fact that the number of treated units is too small. You can consider using parametric bootstrap instead.
On Mon, Oct 12, 2020 at 7:15 AM sebbych notifications@github.com wrote:
Closed #44 https://github.com/xuyiqing/gsynth/issues/44.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/xuyiqing/gsynth/issues/44#event-3866850875, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB2PKGD52BFSISOCIXAAW7LSKMFOJANCNFSM4OR6BOYA .
-- Yiqing Xu
Assistant Professor Department of Political Science Stanford University http://yiqingxu.org/
Hi, I have applied the Gsynth function to my dataset.
But, I find it strange that the ATT curve is not always within the confidence interval. It is not something that I was expecting.
My dataset has a staggered treatment (among the treated units, not all treated units begin the treatment in the same year). But some units also leave treatment earlier than other treated.
I thought it can be an explanation of the problem, but the problem stays the same if I reclassify all units as treated all the periods after the first year of treatment.
Please find attached my data frame and my script.
gsynth_problem.zip
Thank you for any explanation or reference that can help me interpret or correct it! Thanks for an awesome package!
PS: if I run the analysis with the MC estimator, the problem disappears.