st: FE OLS with few clusters

[Marc Peters <marcpeters32c@gmail.com>]

Dear all,


I have a dataset with units clustered into countries and want to
specify a fixed effect regression model. My problem is that the number
of countries in my dataset is very small (only 8) while the number of
observations/country is larger (n=40-350), do I have too few clusters
to run a FE OLS regression? If so, is there some other kind of model I
could run?

Thanks for your help,


Marc
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/