I am looking for a Stata (or R/Matlab if there's no Stata) implementation of the model described by Greene (1994) (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1293115).

It is essentially a Heckit for ZINB instead of OLS.

Would anyone know if such an implementation exists? Also, a critical formula in this model (3.6, p18) is somehow missing from the Working Paper. If anyone knows of another version that has it, that would be greatly, appreciated.


Not an answer, but more of a comment.

I once spent time looking for this with no luck. This particular model is not even in the Winkelmann or Cameron and Trivedi's count data books.

Some possible solutions:

  1. Ask this on Statalist. Jeff Wooldridge may know the answer. Do note the SL policy on cross-posting.
  2. etpoisson with robust errors if your endogenous variable is binary.
  3. E-mail Greene himself.
  4. Flesh out your problem more in case there's another solution, like fmm.

If you figure it out, do post the answer(s).

| cite | improve this answer | |
  • $\begingroup$ Thank you. I will work on #1 and #3, etpoisson is not useful, as I read it, because it does not account for overdispersion or a mass-point at zero. What I'm looking for is something like etzinb. $\endgroup$ – JorgeG Jun 26 '14 at 17:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.