[Limdep Nlogit List] double hurdle models

[Brett Smith]

Thank you Bill

That all makes sense. I still have a related question (explaining why I was
looking into this in the first place).

1) The code {probit;hold$, Selection;Tobit;MLE : topic E31.2.1 which I read
as a double-hurdle model , probit with tobit} produces the same result as
{probit;hold$, Selection;MLE  : which I read as selection mode with
regress}. I am not sure if the inclusion of ;tobit is making a difference. 

Regards,

Brett



-----Original Message-----
From: limdep-bounces at limdep.itls.usyd.edu.au
[mailto:limdep-bounces at limdep.itls.usyd.edu.au] On Behalf Of William Greene
Sent: Thursday, 20 May 2010 12:11 AM
To: Limdep and Nlogit Mailing List
Subject: Re: [Limdep Nlogit List] double hurdle models

Brett
If you want to fit a double hurdle model with rho = 0, you can
just fit the two equations separately: a probit model and a truncated
regression or truncated poisson.  I don't see what SELECT has to do
with this, however.  SELECT, by construction, with RHO = 0 is just a
linear regression. That is why SELECT won't let you set RHO=0. IF
RHO=0, you have to use a different routine (REGRESS or TOBIT).
/Bill Greene


----- Original Message -----
From: "Brett Smith" <Brett.Smith at uwa.edu.au>
To: "Limdep and Nlogit Mailing List" <limdep at limdep.itls.usyd.edu.au>
Sent: Tuesday, May 18, 2010 8:45:09 PM GMT -05:00 US/Canada Eastern
Subject: [Limdep Nlogit List] double hurdle models

Dear all

I would like to restrict rho=0 in a double-hurdle estimation, but I cannot
see how to do this.

I have the general code
Probit...;hold$
Select; tobit; ...$

The estimate of rho=1 gives me the same model estimate as the Heckman
2-stage selection model (using ;mle does not converge, so I am
investigating).

Thank you  in advance 

Brett 

_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au
_______________________________________________
Limdep site list
Limdep at limdep.itls.usyd.edu.au
http://limdep.itls.usyd.edu.au