When I introduced automatic difference-in-Sargan testing into xtabond2
(version 2.3.0, January 2007), I created a small bug. In the Mata
version, in system GMM, if there were no iv() options in the command
line and nocons was not specified, then the constant term was
incorrectly excluded from the instrument matrix (but retained in the
regressor matrix). Typically, users include time dummies in an iv(), so
the bug shouldn't be too harmful.
The bug has just been fixed, with version 2.4.8. You can get the update
by typing "ssc install xtabond2, replace".
--David
*
* 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/
