Dear Stata listers,
I have a question regarding XTGLS and the use of dummy variables for panel units. Firstly, I would like to refer to http://www.statalist.org/forums/foru...-square-values where Clyde Schechter advices Panos to not use dummies for company fixed effects when using xtgls, since "-xtgls- will automatically absorb the fixed effects for the panel variable you specified when you -xtset- your data." Clyde also asks whether the panel variable in -xtset- was some other variable at a different level of aggregation, and this is what my question is about.
My dependent variable is on county level (c=county=aggregate of municipalities) and my independent variables are both on county level and municipal level. I want to run -xtgls- to be able to control for heteroscedasticity, autocorrelation and also apply analytical weights. Now, my data is xtset on id=municipality and time=year. What are the pitfalls of including county fixed effects (i.c)?
Best regards
Hanna L
I have a question regarding XTGLS and the use of dummy variables for panel units. Firstly, I would like to refer to http://www.statalist.org/forums/foru...-square-values where Clyde Schechter advices Panos to not use dummies for company fixed effects when using xtgls, since "-xtgls- will automatically absorb the fixed effects for the panel variable you specified when you -xtset- your data." Clyde also asks whether the panel variable in -xtset- was some other variable at a different level of aggregation, and this is what my question is about.
My dependent variable is on county level (c=county=aggregate of municipalities) and my independent variables are both on county level and municipal level. I want to run -xtgls- to be able to control for heteroscedasticity, autocorrelation and also apply analytical weights. Now, my data is xtset on id=municipality and time=year. What are the pitfalls of including county fixed effects (i.c)?
Best regards
Hanna L
