Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: re: IV regression with multiple IVs


From   Kit Baum <baum@bc.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: re: IV regression with multiple IVs
Date   Sun, 2 Nov 2008 16:00:48 -0500

<>
Nathan said

I am running an IV regression in Stata 10 and am instrumenting 2
variables, each with 2 instruments.

More specifically, variable x_t is instrumented with r_t and (r_t)^2
and x_t-1 is instrumented with r_t-1 and (r_t-1)^2.

The code I am using to do this currently is:

ivregress 2sls y  (x   x_l =  r   r2   r_l  r2_l), robust cluster(ccode)

2 denotes squared, _l is the lag.

Can anyone please advise if this is the correct code for what I am
attempting to do? Any help is appreciated.



It is not meaningful to talk about "this is being instrumented by that". All included endogenous variables (x and x_l) are instrumented by all exogenous variables, which in this case are the four in the list. If you are confident that current r is exogenous, then this code makes sense. You should do a test for overidentifying restrictions after running the regression to ensure that the assumptions of exogeneity are reasonable. I believe that after -ivregress- you can do this with -estat-; -ivregress- does not automatically produce the test as -ivreg2- does.

Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index