Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: FW: Marginal effects for the two-part model

From   "Michael Palmer" <>
To   <>
Subject   RE: st: FW: Marginal effects for the two-part model
Date   Wed, 2 Feb 2011 16:25:09 +1100

Thanks Maarten, 

I also thought this was the case (independent me) until I read a few
papers that argue otherwise.

See for example, 



Michael Palmer
PhD Candidate 
National Centre for Epidemiology and Population Health 
The Australian National University 
Ph. 6125 0538
M. 0437 867 940

-----Original Message-----
[] On Behalf Of Maarten buis
Sent: Tuesday, 1 February 2011 6:59 PM
Subject: RE: st: FW: Marginal effects for the two-part model

--- On Tue, 1/2/11, Michael Palmer wrote:
> The first part of the model is logit model for seeking hospital care 
> followed by an ols model for the cost of care for persons that sought 
> care.
> Simplified code is: 
> Logit care $xvar, robust
> Reg cost $xvar if care==1
> I am wanting to compute the conditional marginal effects but do not 
> know how to do this.

In this model the cost is independent of seeking care, so the marginal
effects of $xvar on the cost conditional on seeking care are just your
regression coefficients (the output you got from -regress-).

Typically, when estimating both parts you would want to allow for
correlation in the error terms across your two steps. I tend to think of
those correlations as variables that influence both the probability of
seeking care and the cost and are not in $xvar. If that is what you
want, then you can take a look at -help heckman-.

Hope this helps,

Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen


*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index