Hello all,
I understand that we can use the ---hausman--- after ---xtlogit--- for testing fe vs. re specifications.
Here is my problem:
* panelid: zip code
* time: year (2005-2011)
* for most zip codes, the binary dependent variable is always zero.
* ----xtlogit, fe----- excludes all units where there is no variation in the dependent variable.
The FE model runs with 1,700 observations vs. 22,000 observations for RE model.
Is the use of ---hausman--- appropriate here? As a note, the test runs and rejects the default null hypothesis.
--Tanuku
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
