CMP - Simultaneous models & Sample Selection

[Yasmine Shemeis]

Dear all,

I am trying to estimate a wages equation with an endogenous ordered discrete variable (health) by CMP as follows,

(EQUATION 1): cmp (WAGES = HEALTH X1 X2) (HEALTH= X1 Z1), indicators($cmp_cont $cmp_oprobit)

But I am likely to have a selection issue for individuals who are out of the labour force and my selection equation is identified as follows

(EQUATION 2): I = X1 Z2

I have checked the cmp help file and many resources but I am still unclear about how to incorporate the IMILLS ratio in my EQUATION 1

Any help will be appreciated.
Thanks

[David Roodman]

I think you would just do:

cmp (WAGES = HEALTH X1 X2) (HEALTH= X1 Z1) ( I = X1 Z2), indicators(I*$cmp_cont I*$cmp_oprobit $cmp_probit)