Year-dummy, xtpoisson, FE

[Ward Bruurs]

Dear Stata-fan

Is it necessary to include year dummies (i.year) in a regression if I have already included Fixed Effects? The FE already has country and year dummies?

Regards

[Andrew Musau]

Official xt commands in Stata with the -fe- option only control for unit effects (defined by your panel identifier in the xtset). Therefore, if your xtset command is

xtset firm year

then year effects are not included.

[Ward Bruurs]

Official xt commands in Stata with the -fe- option only control for unit effects (defined by your panel identifier in the xtset). Therefore, if your xtset command is

xtset firm year

then year effects are not included.

So you would suggets a xtpoisson ... i.year, fe robust regression?

[Andrew Musau]

If you want to account for year effects, yes.