IV regression,the independent variable is Non-negative integer, endogenous variable is ordinary, which command can I use?

[Lingfeng Wang]

Thanks a lot, Now, I am facing a problem, the problem is just like the title said.
I want to conduct a reaserch the impact of education(1- primmary school; 2 - middle school; 3 - high school.. and so on) toward the people's number of reading books per year(0, ,1 , ..., ), but when I want to use the IV regression as the robust, The problem is I can't find the suitable command to fit these kind of data, I am confused about that, Coud you give me some advices?

[Liu Qiang]

It seems it depends on your choice of the model. If you take your dependent variable (number of books) a continuous variable, then --xtivreg-- would be an option. If you want to conduct an ologit model, you may have to use gsem.

[Joao Santos Silva]

Dear Lingfeng Wang,

If your dependent variable is a count (number of books) you should consider ivpoisson. This is valid even if the data does not have a Poisson distribution.

Best wishes,

Joao