Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: IV with Sample Selection


From   aam00@aber.ac.uk
To   statalist@hsphsun2.harvard.edu
Subject   st: IV with Sample Selection
Date   Fri, 14 Nov 2003 03:20:00 -0000 (GMT)

Dear Stata Users,

I am trying to estimate returns to education with sample selection. In
particular, for my female sample I am applying parametric procedure
(Heckman two-step)to correct for the selectivity bias. To correct for the
endogeneity of schooling I use instrumental variable: spouse's education.
My problem arises when I estimate the labour force participation (LFP) via
probit (first step of Heckitt). Clearly, in the probit equation one should
also include the schooling variable (in my case it's measured as years of
schooling). However, I am not certain exactly what should appear in the
probit: observed schooling of the (female) respondent or instrumented
schooling (using spouse's level of schooling as an instrument). LFP
equation is a binary outcome: LFP=1 if the female receives a wage, hence
participates, and 0 otherwise.

Once I obtain Mill's ratio I continue to estimate the returns to education
for the females by including the instrumented level of schooling and
Lambda on the RHS of the equation. Now I have corrected for the possible
endogeneity and selection biases. however, the probit equation is causing
a concern: specifically, what should appear on the RHS-observed or
instrumented schooling?

I would very much welcome your comments on the above. May be you can
direct me to a paper that addresses this issue simultaneously.

Truly,

Altay Mussurov
PhD (c)
University of Wales Aberystwyth












*
*   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