Statalist The Stata Listserver


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: xtivreg2 and Wooldridge's correction for attrition bias in a panel


From   "Patrick Van Horn" <pvanhorn@uci.edu>
To   statalist@hsphsun2.harvard.edu
Subject   st: xtivreg2 and Wooldridge's correction for attrition bias in a panel
Date   Fri, 29 Sep 2006 07:56:16 -0700 (PDT)

I have a panel of banks during the Great Depression with quarterly
observations. I am interested in seeing how they were managing their cash
reserves during this time. Therefore I the have reserves to assets ratio
as my dependent variable and my independent variables are bank
characteristics, macro variables, and monetary variables.
I was running the xtreg, fe command for a fixed effects model. However, my
data set suffers from attrition as some banks go out of business and leave
the data set for good.
Since the attrition will produce biased estimates in the fixed effects
approach, I was referred to Wooldridge's discussion of attrition in his
panel data book. He recommends an approach that uses first differences and
then an IV OLS regression. In his book, its on pages 585-590 if this is
not clear.
It seems that the xtivreg2 command does exactly what he recommends for
correcting for attrition. Am I correct in this assumption? Or do I
manually need to first difference and then run an IV estimation? Forgive
me if I am asking something that is extremely remedial or nonsensical. I
am an economic historian by training and not very well versed in
econometrics of this type.
Any input would be greatly appreciated.

Thank You

Patrick

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