Panel data regression using fixed effects

[Joe Mckennedy]

Hi everyone,

I have a question regarding Panel data, this is my stata command: xtreg LogPrice CPI Log_nonWorkers Log_incomeWorkers i.small i.year,fe cluster(ID)
I was wondering if it would be correct to run xtreg LogPrice CPI i.small i.year,fe cluster(ID), afterwards do the same command again but replace CPI with Log_IncomeWorkers
ID which is the firm in each city and the data runs from 1989-2004.


Kind regards,

Joe Mckennedy

[Carlo Lazzaro]

Joe:
welcome to this forum.
If you -xtset- your dataset with -firm- as your -panelid-, you cannot include -i.firm- as a predictor.
As far as your codes are concerned, you should go with the one that gives a fair and true representation of the data generating process you're investigating (skim through the literature in your research field in this respect).