Difference between fixed effects estimation and i.year i.region (time and region fixed effects)

[ana maria giraldo]

Hello all,

I am working with panel data, I have observations from the years 1930 until 1970 and I have three different regions.
I have read a lot about fixed effects and but I am still trying to understand what is the difference between:

xtreg, fe
and using time and region dummies: xi reg y i.time i.region (plus all other explanatory variables)

and which one would you recommend me for evaluating:
years of schooling= ln(migrantshare) + ln(urbanshare)


Thanks for the help

[Carlo Lazzaro]

Ana Maria:
as per your description, you're seemingly dealing with a small N, large T panel dataset.
hence, I would recommend to take a look at -xtgls- and -xtregar-, whereas -xtreg- is out of debate with such a T dimension.

[ana maria giraldo]

Hi Lazzaro,

Thank you for your help, I see the difference now