
From  Marcello Pagano <pagano@hsph.harvard.edu> 
To  statalist@hsphsun2.harvard.edu 
Subject  st: ivreg2+first stage 
Date  Fri, 11 May 2007 06:02:22 0400 
From: Kit Baum <baum@bc.edu>
Date: May 10, 2007 8:24:25 PM EDT
To: statalist@hsphsun2.harvard.edu
Subject: ivreg2+first stage
sacrificial camel
Andrea said
I want to find the answer to a simple question.
Consider a panel data such as
y(it) = a + a1*x1(it) + a2*x2(it) + a(i) + u(it)
where x2 is predetermined or endogenous. I found a potential set of endogenous variable Z(it) to use as instrument. I know that applying ivreg to such equation provides consitent estimates of a, a1, a2 and u(it) [suppose all variable has been differenced to remove unobserved heterogeneity).
However, i'm also interested in estimating
x2(it) = b + b1*x1(it) + b2*Z(it) + a(i) + e(it),
the first stage of the iv regression. Is such an estimate consitently estimated by ivreg2, first?
Of course. If x1 and Z are appropriately uncorrelated with the e process, the FSRs of IV are consistent estimates. If diagnostics on the IV estimates indicate that you have not satisfied the orthogonality conditions (e.g. the SarganHansen test), then you should question whether the FSRs are appropriately estimated by OLS.
Kit
Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.statapress.com/books/imeus.html
* * 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 1996–2016 StataCorp LP  Terms of use  Privacy  Contact us  What's new  Site index 