Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.

# st: Heckman unconditional marginal effects decomposition for dummy variables

 From Adri Bestazza To statalist@hsphsun2.harvard.edu Subject st: Heckman unconditional marginal effects decomposition for dummy variables Date Fri, 7 May 2010 21:50:25 +0100

```Dear Statalisters,

Suppose I run:

heckman a b c d, select(b c d e f)

where "b" & "c" are continuous variable, whereas "d" is a dummy ("e" &
"f" are selection instruments not relevant for my question)

Suppose I want to apply the decomposition for unconditional marginal
effects for dummies:

E1(y)-E0(y)=P1(y>0)[E1(y|y>0)-E0(y|y>0)]+E0(y|y>0)[P1(y>0)-P0(y>0)]

following Stata Technical Bulletin, July 2000, for the Tobit model.

Let us say I want to find E1(y|y>0) & P0(y>0) for the dummy "d" at first

Would I do right if I ran:

heckman a b c d, select(b c d e f)
mfx compute if d==1, predict(ycond)
mfx compute if d==0, predict(psel)

?

If not, what stata command should I use to find E1(y|y>0) & P0(y>0)?

Best,