Is it possible to estimate a simultaneous equation model in which we have an endogenous and qualitative variable in one of the equations?
I have the following equation:
y=f(B, X)+Epsilon1
B=f(Y, X,Z)+ Epsilon1
All variables are continous variable, but B is a discret variable with 5 states.
What estimator can I use with cross sectional observation?
what estimator can I use with panel data observation?