Fixed Effects Estimation

[Ayush Salgia]

Hi Everyone,

I am currently running a FE regression with panel data on OECD countries, looking at the effect of labour and capital growth on output growth. As part of this I am running a fixed effects regression, and want to be able to extract the fixed effects for each country in my panel data to compare them, but an unsure as to what command is needed for stata to generate these on a per country basis

Any help would be much appreciated!

[Clyde Schechter]

If it is a linear fixed effects regression you can run:

predict u, u

following the regression, and Stata will put the fixed effects in a new variable named u. (Evidently if you already have a variable named u, pick a different name for this.)

If your fixed effects regression is not a linear regression, then it cannot be done as those "fixed effects estimators" are actually estimators conditional on the fixed effects, and they cannot estimate the fixed effects themselves.

[Ayush Salgia]

Great got it thank you.