Dear listers,
I estimated a system of simultaneous equations using 3SLS via -reg3-
which, by default, assumes the equations are correlated. I then wish
to investigate the consequences if the correlations are restricted to
zero. I did this by adding -corr(independent)- option to -reg3-. To
my surprise, the resulting error covariance matrix is not diagonal as
it should be (advertised on p.308, Reference Manual [N-R]).
Am I missing something here? I know for my original purpose I can
apply 2SLS using -ivreg- to each of the equations, but I am wondering
why the -corr(independent)- does not work as advertised.
The following code demonstrates the issue.
====== demo begins =========
use http://www.stata-press.com/data/r8/supDem
global demand "(qDemand: quantity price pcompete income)"
global supply "(qSupply: quantity price praw)"
reg3 $demand $supply, endog(price)
mat list e(Sigma)
reg3 $demand $supply, endog(price) corr(independent)
mat list e(Sigma) /* should be diagonal */
==== demo ends =======
HJW
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
*
* 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/
