Dear all, this is my first message on the statalist, I thus hope to do it
right.
I have a problem,
I am working with an unbalanced Panel of about 25 countries and maximum of 7
years (for some countries). As I stated before, the Panel (or better the
TSCS) is unbalanced since the time dimension is not the same for all the
countries. In addition, I have some time-invariant variables. Given that I
work with countries, a random effect is not appropriate and a fixed effect
estimator would not allow me work with time invariant variables. I do not
want to work with a Hausman Taylor estimator.
I want to do a Pooled Least Squares, but, given that the panel is
unbalanced, I fear to give a too big importance to some countries with
respect to others. I thought to use regress with a pweight option but I do
not see how I could do it (the 25 countries are not randomly chosen). Can
anyone help me on thisor can suggest some readings?
Thank you very much for your help,
regards
Vincenzo
_________________________________________________________________
MSN Amor: busca tu � naranja http://latam.msn.com/amor/
