How to estimate a Multinomial Logit model with sample selection?

[Rosa Abraham]

Dear Stata users,

I want to estimate a multinomial logit model with sample selection. My selection equation is binary (to work, or not) and the regression equation is multinomial (informally employed, self employed, formally employed).
Is there a way to estimate this in Stata?

Thank you for your help,

Regards,
Rosa

[Carlo Lazzaro]

I do hope that Rosa may find the following old Stata thread: http://www.stata.com/statalist/archi.../msg00305.html helpful.

[Rosa Abraham]

Dear Carlo,
Thank you very much for the link. It certainly helps, in particular the reference to the Stata Journal.
Many thanks!

Rosa

[Alina Faruk]

Dear Carlo,
Thank you very much for the link. It certainly helps, in particular the reference to the Stata Journal.
Many thanks!

Rosa

Hi,

May I please know how you resolved the issue? I am facing the same problem.

[Rabab Al hasni]

Dear,

May I please know how you resolved the issue? I am facing the same problem.

Hi all

I am having the same problem, please could you please help me on how to solve this problem. In fact, I am attempting to use multinomial logit to examine which factors that would effect access to bank lending; using four unorder categorical variables. I think I cannot use heckman selection model because the assumption of my dataset normality not meeting probit assumption of normal distribution.

please could you please advice on how to use selection model with multinomial logit. I could not find a clear explaination on how to perform on STATA software.

Appreciate your kind help and efforts



Best regards,
Rabab