re:st: xtivreg first stage

[Christopher F Baum <baum@bc.edu>]

<>
I am running a fixed effects xtivreg.
I am interested in the coefficients of the first-stage regression, and
have, therefore, specified "fe, first".
The result is a table in which the dependent variable is __000003 ,
and in which all instruments appear, both included and excluded.
Can anyone explain the meaning of this result?

The depvar in that regression will be a tempvar if you have applied a timeseries operator, e.g. with grunfeld.dta,
 xtivreg invest (D.mvalue=kstock) time, fe first
will give __000003 as the depvar. The FSRs in any IV method are the regressions of each RHS endog on ALL exog, included and excluded.

You might prefer the more elegant handling of FSRs in B-S-S ivreg2 and S-S xtivreg2, available from SSC. There are some options available there to save the FSRs as separate sets of estimates, etc. as well as additional diagnostics.

Kit



Kit Baum   |   Boston College Economics and 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/