Individual FE

[Neg Kha]

Hi,

I have a panel of people working at the same company (but working at this company's different offices) with their characteristics like race, education, age, etc. My outcome variable is their wages and the variable of interest is the gender dummy.
To my specification, I have added year fixed effects, age fixed effects and office fixed effects. And I used the reghdfe command.

So my question is:
Do I also need to add individual fixed effects?

Best,
Neg

[Carlo Lazzaro]

Neg:
how you did -xtset- your data?
That said, my gut-feeling is that you included too many fixed effects.
Evetually, as per FAQ, please share what you typed and what Stata gave you back. Thanks.

[Neg Kha]

Neg:
how you did -xtset- your data?
That said, my gut-feeling is that you included too many fixed effects.
Evetually, as per FAQ, please share what you typed and what Stata gave you back. Thanks.

Hi Carlo!

Sure, my code is:

reghdfe wage female education married i.age i.year , absorb(i.office)

I do not define a xtset since reghdfe command does not need it but is would be :

xtset id year

And variable id would be the employees identifiers.

[Carlo Lazzaro]

Neg:
I do not see any advantage with preferring the community-contributed module -reghdfe- vs -xtreg,fe- here, unless you're intrested in obtainng the fixed effect of -year-, too:

 
 reghdfe wage female education married i.age, absorb(FE1=office FE2=year)