# Re: st: Question regarding test for endogeneity in the contextof a count variable with a binary re

 From "Alfonso Miranda Caso Luengo" To Subject Re: st: Question regarding test for endogeneity in the contextof a count variable with a binary re Date Thu, 13 May 2004 14:16:14 +0100

```Dovev,

You can use the espoisson code that performs a FIML estimation of a endogenous switching Poisson model (have a look to Miranda (2004) Stata Journal 4(1):40-49). You could estimate the model with exogenous switching using the exs option. Then estimate the model with endog. switching. Finally, perform a boundary-likelihood ratio test for the sig. of rho. If rho is not significant, then you can treat your dummy as exogenous.

Best,

Alfonso.

===========================================
Alfonso Miranda
PhD Student

Economics Department
University of Warwick
Coventry CV4 7AL
E-mail: Alfonso.Miranda-Caso-Luengo@warwick.ac.uk
WebPage: http://www2.warwick.ac.uk/fac/soc/economics/research/phd/ecrgw/
===========================================
>>> M.E.Schaffer@hw.ac.uk 05/13/04 13:56 PM >>>
Dovev,

Quoting "Lavie, Dovev" <lavie@wharton.upenn.edu>:

> Does anyone know how to test for endogeneity in the case where the
> dependent
> variable (y) is a count variable and the endogenous regressor is a
> binary
> variable (z)? (see below). As far as I know, the Durbin-Wu-Hausman
> test
> works only with OLS. Any suggestions on how to determine whether
> variable z
> can be assumed exogenous are most welcome.

I think it's not hard to do in principle, but there's a big "if"
involved.  *If* you can estimate the model in two forms, one treating the
regressor z as endogenous, and one treating it as exogenous, then you can
do a Hausman test using the -hausman- command.  When estimating the
former, you'll need a procedure that will deliver a valid var-cov matrix
for the estimated coefficients as well as consistent estimates of the
coefficients themselves.

A quick

findit binomial endogenous

didn't turn up anything, so there may not be any such package around for
estimating a negative binomial with an endogenous regressor.  But maybe
you've taken care of this requirement.

Hope this helps.

Cheers,
Mark

> z = a0 + a1*x1 + a2*x2 + epsilon1 (probit model)
>
>
>
> y = b0 + b1*z + b2*x3 + epsilon2 (negative binomial model)
>
>
>
> Thanks,
>
> Dovev
>
>
>
>

Prof. Mark Schaffer
Director, CERT
Department of Economics
School of Management & Languages
Heriot-Watt University, Edinburgh EH14 4AS
tel +44-131-451-3494 / fax +44-131-451-3008
email: m.e.schaffer@hw.ac.uk
web: http://www.sml.hw.ac.uk/ecomes
________________________________________________________________

DISCLAIMER:

This e-mail and any files transmitted with it are confidential
and intended solely for the use of the individual or entity to
whom it is addressed.  If you are not the intended recipient
you are prohibited from using any of the information contained
in this e-mail.  In such a case, please destroy all copies in
Watt University does not accept liability or responsibility
for changes made to this e-mail after it was sent, or for
viruses transmitted through this e-mail.  Opinions, comments,
conclusions and other information in this e-mail that do not
relate to the official business of Heriot Watt University are
not endorsed by it.
________________________________________________________________
*
*   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/

*
*   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/
```