Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: Endogeneity test (orthog), 2SLS autocorrelation (ivgmmN)


From   "Steven Stillman (LMPG)" <Steven.Stillman@lmpg.dol.govt.nz>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   st: RE: Endogeneity test (orthog), 2SLS autocorrelation (ivgmmN)
Date   Fri, 19 Sep 2003 00:00:38 +1200

Hi Maria.  As an author if ivreg2, I'll attempt to answer your questions.

The short answer to question 1 is no.  The hypothetical instrumental
variables model that would exist if your suspected endogenous variables are,
in fact, endogenous, must be identified.  The exogeneity test is a modify
hausman test, i.e. to test the null hypothesis (H0) that Z1 is endogenous it
compares the estimates from a model that is consistent and efficient if H0
is true and consistent and inefficient if H0 is false (IV) versus a model
that is inconsistent if H0 is true and consistent and efficient if H0 is
false (OLS).  This test can only take place if both models can be estimated.
If you run the correct command syntax of  ivreg2 Y X2 (X1 = Z1), orthog(X2)
to test the exogeneity of X2, you will get a message stating the current
model is exactly identified so a test is not valid.

The short answer to question 2 is soon.  We are currently in our final
testing phase of an update to ivreg2 which extends the package to produce
standard errors that are consistent in the presence of autocorrelation.

Steve 

> -----Original Message-----
> From:	D Jimenez Rubio [SMTP:djr109@york.ac.uk]
> Sent:	Thursday, September 18, 2003 11:39 PM
> To:	statalist@hsphsun2.harvard.edu
> Subject:	st: Endogeneity test (orthog), 2SLS autocorrelation (ivgmmN)
> 
> Dear Statalist users,
> 
> I am running a 2SLS where I suspect 2 of my variables could be
> endogenous. However, I only have adequate instruments for 1 of the
> suspected endogenous variables. Is it possible to use the instrumental
> variable regression (ivreg2) to test whether the variable for which I
> have instruments is endogenous or not and the option "orthog" afterwards
> to check whether the second suspected variable (the one for which I
> don't have instruments) is endogenous? In that way, I can test for
> endogeneity for the second variable without using instruments.
> I would like to ask you something else. I would like to know when is it
> going to be released the command ivgmmN which corrects for serial
> correlation in 2SLS. I suspect that my 2SLS regression suffer from
> autocorrelation and I don't know how to correct for it. 
> 
> Thanks a lot very much
> 
> Maria D.
> 
> *
> *   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/

The information contained in this document is intended only for the
 addressee and is not necessarily the views nor the official 
communication of the Department of Labour.  All final/official papers 
which are sent from the Department will be sent by non-electronic
means, on appropriate letterhead, signed by authorised personnel.
*
*   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/



© Copyright 1996–2021 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index