st: Two stage model question

[Joanne Flavel <joanne.flavel@flinders.edu.au>]

I am trying to estimate a two stage panel data model to control for 
selection. I'm using random effects probit in the first stage to 
generate the correction but my concern is that my second stage outcome 
is not linear. I have tried looking through the archives in the hopes of 
not asking something which has already been answered and noted one entry 
from 2003: (RE: st: Inverse Mills Ratio after MLOGIT, Fri, 18 Apr 2003) 
the two step approach with control for selection only produces 
consistent estimates when the second stage is a linear regression model.

Is there a way to produce consistent estimates with a non linear second 
stage? I have not been able to find another model that is feasible for 
me to do this given I am using panel data and with all the econometric 
issues I am trying to address. Any advice and suggestions are welcome

Thanks,
Joanne

--

Joanne Flavel
National Institute of Labour Studies
Flinders University

*
*   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/