Dear all,
Couple of recent posts point out difficulties obtaining the marginal effects after IVprobit command. This seemed a rather simple issue of not including the "predict(pr)" option. From here i remember a problem I had in the past which now have resurface.
In specific, what is the correct syntax (or otherwise method) to estimate marginal effects after using ivprobit?
My main concern goes to the fact that older versions of stata (using version control) produce different results, and that some of them (at least point estimates) could be estimated without using ivprobit.
Please see the example below
Code:
version 14
use http://www.stata-press.com/data/r13/laborsup, clear
ivprobit fem_work fem_educ kids (other_inc = male_educ)
margins, dydx(other_inc fem_educ kids male_educ) atmeans  predict(pr) post
est sto m14a
ivprobit fem_work fem_educ kids (other_inc = male_educ)
margins, dydx(other_inc fem_educ kids male_educ) predict(pr) post
est sto m14b

version 15
use http://www.stata-press.com/data/r13/laborsup, clear
ivprobit fem_work fem_educ kids (other_inc = male_educ)
margins, dydx(other_inc fem_educ kids male_educ) atmeans  predict(pr) post
est sto m15a
ivprobit fem_work fem_educ kids (other_inc = male_educ)
margins, dydx(other_inc fem_educ kids male_educ) predict(pr) post
est sto m15b
probit fem_work fem_educ kids other_inc  male_educ
margins, dydx(other_inc fem_educ kids male_educ) predict(pr) post
est sto m15c

. est tab m14a m14b m15a m15b m15c, se 

-------------------------------------------------------------------------------
    Variable |    m14a         m14b         m15a         m15b         m15c     
-------------+-----------------------------------------------------------------
   other_inc | -.02143636   -.01401501   -.01357387    -.0097802    -.0097802  
             |  .00241808    .00098359     .0025939     .0014994    .00098284  
    fem_educ |  .08337907    .05451292    .08650361    .06232728    .06232728  
             |  .01056916    .00660068    .01103787    .00709898    .00603163  
        kids | -.07191827   -.04701989   -.08525341   -.06142649   -.06142649  
             |  .01888225    .01233965    .01964523    .01394461    .01283831  
   male_educ |  (omitted)    (omitted)    -.0269815   -.01944062   -.01944062  
             |                            .00268593    .00221035    .00748863  
-------------------------------------------------------------------------------
                                                                   legend: b/se
Thank you
Fernando