Pool versus individual coefficients
Posted: Thu Jul 31, 2014 11:11 am
I have estimated a model using a pool (a cross section of cities) with OLS. To check it, I had used the options for no fixed effects and all coefficients varying in cross section. I expected that would produce a set of individual cross section (city) models which would be the same as if I had estimated each separately. The coefficients are very different. Can you explain what is happening?