Bookmark and Share

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


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

Re: st: Marginal effects with an Heckman selection model


From   [email protected]
To   [email protected]
Subject   Re: st: Marginal effects with an Heckman selection model
Date   Mon, 16 Apr 2012 20:41:58 +0200 (CEST)

Thanks a lot for your reply.

I try to apply your suggestions but it only works when you estimate the heckman selection model by ML.
In my case, I estimated my model by the 2 step procedure because the ML estimation does not work.


margins, dydx(*) predict(psel)  works with the two step procedure

but the 2 other predictions does not work

// marginal effects on expected
// wage (not working is 0 wage)
margins, dydx(*) predict(yexpected)

// marginal effects on expected
// wage (excluding not working women)
margins, dydx(*) predict(ycond)




----- Mail original -----
De: "Maarten Buis" <[email protected]>
À: [email protected]
Envoyé: Lundi 16 Avril 2012 09:45:40
Objet: Re: st: Marginal effects with an Heckman selection model

On Sun, Apr 15, 2012 at 8:30 PM, <[email protected]> wrote:
> I'm estimating the charitable giving of households thanks to
> administrative data's. To do so I'm estimating an exponential Heckman
> selection model (Wooldridge, 2nd edition, chapter 17) with the two
> step procedure.

Notice that this is a multidisciplinary list. This has consequences,
one of those is that an article or book that is "world famous" in your
(sub-(sub-))discipline is likely to be completely unknown outside that
micro cosmos. So even if it is enough within your discipline to a use
a name/year or name/edition reference, this is completely
uninformative on this list. This is why the Statalist FAQ asks you to
give full references. You can find a link to the Statalist FAQ at the
bottom of this post (and any other post on this list). It is well
worth studying as it contains many tricks on how to ask answerable
questions on this list.

> I want to estimate the marginal impacts of my variables on the
> probability of giving to charities and on the amount given.

You do this by specifying the -predict()- option. You can see what you
can specify in that option by typing -help heckman postestimation- and
going to the entry on -predict-.

*----------- begin example --------------
webuse womenwk
heckman wage educ age, ///
select(married children educ age)

// marginal effects on
// the probability of working
margins, dydx(*) predict(psel)

// marginal effects on expected
// wage (not working is 0 wage)
margins, dydx(*) predict(yexpected)

// marginal effects on expected
// wage (excluding not working women)
margins, dydx(*) predict(ycond)
*----------- end example --------------

I leave it up to you how you are going to deal with the problem that
you get marginal effects on the log(donations) rather than donations.

Hope this helps,
Maarten

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


http://www.maartenbuis.nl
-------------------------- *
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


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