Dear Statalist,
is there a program available in Stata to estimate fixed-effects with
sample selection? I found some older threats and citations in the
Statalist archive, but I am hoping that maybe something has become
available since then.
I would like to estimate a fixed-effects model with sample selection since
people non-randomly drop out of my sample. Wooldridge 1995 (Journal of
Econometrics) and in his graduate textbook proposes some tests for
non-random sample selection which showed this to be the case in my data.
He proposes to then proceed in three steps:
1. estimate the selection equation separately for each period, get the
inverse mills ratio
2. run the pooled OLS regression using the selected sample and including
these inverse mill ratios (one for each period) and time dummies
3. correct the standard errors for "general heteroskedasticity and
serial correlation, as well as the first stage estimation of the inverse
mills ratios".
(I use Stata SE 8.2 .)
Thank you for your help,
Kristin
*
* 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/
