look at logistic regression

Best wishes David Booth

There is a bivariate logit and probit if you have two 1,0 dependent variables.

2SLS requires continuous variables.

Or if you have participation (no, yes) and then wages if yes then a heckman sample selection would do as you don't observe wages for non-workers

You can try IVProbit. In STATA, the command is ivprob.

Or you can do it manually (First stage --> Predicted value --> Second stage).