Re: st: endogeneity test

[Bobasu Alina <ally_bib@yahoo.com>]

Try to use ivregress gmm syntax and then run the endogeneity test for each 
variable and subsets of variables. 

not realyy sure if this is what you were asking.


----- Original Message ----
From: Dan Bergov <danbergov@yahoo.com>
To: "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Sent: Wed, May 11, 2011 12:45:46 PM
Subject: st: endogeneity test

Hello! I have the following question: I estimated with GMM a liniar equation and 
then performed AR test, robust to weak instruments. However when I run the 
condivreg option I have to mention the endogenous variable. So I run an 
endogeneity test which I noticed that is availabe for iv methods (ivregress 
syntax with estat endogenous postestimation option). My problem is: if I 
initially estimated my equation with GMM is it good if I perform also the 
ivregress gmm estimation which allows me to perform the endogeneity test? 
(gmm doesn't have this endogeneity test). The problem is that if i estimate with 
gmm option and on the other hand also estimate with ivregress gmm  I do not 
obtain the same results (I know that for instance 2sls is a particular case of 
GMM in the case of homoskedeastic errors) so I do not  know if it is correct to 
perform the endogeneity test after ivregress but admit that it is the same for 
GMM estimation.
I want to estimate with GMM and not ivregress gmm because I do not want to 
specify the endogenous varibale which is a requirement for ivregress.
I do not know if I made myself clear. Sorry. Hope for a reply.
Dan

*
*  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/


*
*   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/