<>
I am trying to run a IV estimation using ivreg2. I want include only a
subset of the exogenous regressors from the main equation in the
reduced form equation. It seems that ivreg2 does not allow this
option. Is there any way that'll work? Any suggestions are
appreciated.
This is a FAQ. As explained in Baum-Schaffer-Stillman 2003 (freely
available from first URL) and IMEUS, single-equation (limited-
information) estimation with instrumental variables requires that all
of the exogenous variables in the regressor matrix also appear in the
instrument matrix. The 'first stage' regressions involve projecting
each endogenous variable on _all_ exogenous variables, including those
in the equation. You don't get to pick and choose which included
exogenous variables to use. This constraint is not imposed by ivreg2,
ivregress or ivreg: it is imposed by econometric theory.
Kit Baum | Boston College Economics & DIW Berlin | http://ideas.repec.org/e/pba1.html
An Introduction to Stata Programming
| http://www.stata-press.com/books/isp.html
An Introduction to Modern Econometrics Using Stata | http://www.stata-press.com/books/imeus.html
*
* 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/