I wrote:
>>No guarantees on identification, etc., of course.<<
That's a little vague. You should be able to set up the problem to be
identified but may run into empirical unidentification.
There's an article by Rabe-Hesketh and Skrondal that shows how to set up
multilevel nominal choice models. I can't recall which journal it's in,
but see http://www.gllamm.org, which has references to nearly everything
they've done.
Jay
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
