[Limdep Nlogit List] Murphy and Topel correction

William Greene wgreene at stern.nyu.edu
Tue Apr 15 06:12:09 EST 2008

Dear Francesca. 
First, I'm not sure the "multinomial logit" model is what you
want for two binary variables. The MNL is for a single variable
that indicates which among two or more outcomes occur. You
might have two variables that connect via a bivariate probit
model, however. (There is no workable bivariate logit model.)
However, in either case, I'm afraid there is no built in routine
in LIMDEP for this case that extends the one in R10-44 to a 
pair of endogenous variables.
Bill Greene

----- Original Message -----
From: "Francesca Fabbri" <Francesca.Fabbri at lrz.uni-muenchen.de>
To: limdep at limdep.itls.usyd.edu.au
Sent: Monday, April 14, 2008 3:52:56 PM (GMT-0500) America/New_York
Subject: [Limdep Nlogit List] Murphy and Topel correction

Dear Sir,

I would like to estimate a model where I have a continuous variable (wage)
as a function of two endogenous binary variables (which follow a
multinomial logit model). I think I need to estimate a 2 step model, like
the one described in the LIMDEP manual (R10-44). My question relates to
the Murphy and Topel corrections for the standard errors. Is there a
routine already developed?

Thanks and regards,


Limdep site list
Limdep at limdep.itls.usyd.edu.au

Professor William Greene
Department of Economics
Stern School of Business
New York University
44 West 4th St., Rm. 7-78
New York, NY   10012

More information about the Limdep mailing list