Suppose I add a time-fixed effect to a panel data regression that I want to estimate using OLS. My question is conceptually when should I also cluster by time (in addition to adding fixed effects)?
-
Login or Register
- Log in with

Comment