st: TSCS with ordinal dep. variable and serial correlation

[Michael Hess <salsa.fresca@gmail.com>]

I am trying to analyze a dataset where the dependent variable is 
ordinal, five categories.  The dataset is TSCS and has observations for 
57 countries over 29 years.  There is a lot of serial correlation in the 
model centered around the dependent variable which trends upward.

I have been using the reoprob command by Guillaume Frechette to run the 
model, but have not found much literature on best ways to get rid of the 
autocorrelation in a maximum likelihood model.  Adding time dummies 
reduces, but does not eliminate, the problem.

Has anyone else dealt with this?  Any suggestions for me?

Michael L. Hess
University of New Orleans
*
*   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/