On Jan 3, 2008 12:01 PM, <[email protected]> wrote:
> Does anyone knows if there is a program to estimate Cragg's two-part model which
> I understand is very similar to Heckman's? I found a dhurdle.ado on the Web
> (created by J. Fennema) but the program does not seem to have ever been
> completed (and used), and there is no .hlp file. So, I hope something better
> must exist...or it does not?
>
> Thanks a lot.
>
> B. Milanovic
I am not aware of anything available other than what B. Milanovic
mentioned. However, he may need nothing more than the official Stata
commands to estimate the Cragg's model. The Cragg's model is a two
stage model where in the first stage one estimates a dichotomous model
that determines whether an individual participates in the market. In
the second stage, a truncated model is estimated based on the
assumption that when the individual participates in the market it will
consume or demand positive values (of goods or services). The original
Cragg's model was estimated with a probit and a truncated normal
[Stata does both]. However, the idea works for any dichotomous choice
model and any truncated model [Stata is also good at most, if not all
of these] if the two models can be estimated independently. B.
Milanovic will need to determine whether the two decisions he's
dealing can be separated or not. Bockstael, Strand et al. (1990) [Land
Economics, issue 66, pp. 40-49] find that when the researcher deems it
necessary to separate the two decisions, the Cragg's model seems to
work best.
--
P. Wilner Jeanty, Post-doctoral researcher
Dept. of Agricultural, Environmental, and Development Economics
The Ohio State University
2120 Fyffe Road
Columbus, Ohio 43210
(614) 292-6382 (Office)
*
* 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/