Re: st: marginal effects from -ivprobit-

["Brian P. Poi" <bpoi@stata.com>]

On Sun, 4 Dec 2005, Adrienne L wrote:

> Greetings listservers,
>
> I am trying to extract the marginal effects from an -ivprobit-
> estimation.  When I use the -mfx- command after my estimation, the
> estimates provided are identical to those after -ivprobit- (not the
> marginal effects).  I tried using the old -divprob-.  Those marginal
> effects are different from -ivprobit- (and appear to actually be the
> marginal effects).  I would prefer to use the -ivprobit- -mfx-
> combination because of the flexibility in the error structure.  I am
> using Stata/SE 9.0.  Any help you can provide is much appreciated.
>
> Adrienne
By default, -predict- after -mfx- calculates the linear prediction, so 
-mfx- will simply return the estimated coefficients.

To get the marginal effects on the probability of a positive outcome, you 
need to use the predict(p) option with -mfx-:

   . sysuse auto, clear
   . ivprobit foreign mpg (turn = length)
   . mfx compute, predict(p) eq(foreign)


The eq(foreign) option tells -mfx- to compute marginal effects only for 
the equation for foreign.  Without that option, -mfx- will also show the 
marginal effects of the extra instruments, which are identically zero.

   -- Brian Poi
   -- bpoi@stata.com
*
*   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/