Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

st: Appropriate version of Hansen J-test with xtabond2 and year effects


From   Ariel BenYishay <a.benyishay@unsw.edu.au>
To   statalist@hsphsun2.harvard.edu
Subject   st: Appropriate version of Hansen J-test with xtabond2 and year effects
Date   Tue, 24 Jan 2012 10:11:40 +1100

I'm using the xtabond2 package to run a difference GMM model.  My
model has year dummies, which enter "IV style":

xi: xtabond2 y l.(y) i.year, gmm(l.(y)) iv(i.year) noleveleq robust

In this case, from what I understand, the transformed year effects act
as their own instruments in the first stage. That makes them invalid
as classical instruments, right?  That is, when they are included in
the Hansen J-test (the default), they would tend to be rejected (i.e.,
they would push the chi-squared stat up, potentially leading to
over-rejection).  I note that xtabond2 automatically reports both the
standard J-test and version without the year dummies -- is the latter
(excluding the dummies) the more appropriate test?

Thanks!
--
Ariel BenYishay
Lecturer (Assistant Professor)
School of Economics
University of New South Wales
a.benyishay (at) unsw.edu.au

*
*   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 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index