I am trying to estimate a system of equations. There are two dependent
variables, y1 and yr2 and they are functions of the same set of covariates.
Y1 is continuous (and is estimated using least squares) while y2 is
censored at 0. After estimating both these equations, I would like to
perform a non-linear test on the estimates. I thought that -suest- was a
suitable candidate to do this.
Here are the questions I had regarding this procedure.
1. One of these covariates is endogenous to both y1 and y2. Hence, it needs
to be instrumented for, and I have a set of instruments. Hoewever, if i
want to use ivreg, it does not have a score option that is required by the
-suest- command.
2. How does one correct for endogenous variables in intreg?
Till now, I was planning to regress the endogenous variable on the set of
explanatory variables and instruments, and take the predicted values to the
equations that I estimate in the -suest- command. However, I believe that
the standard errors in doing that may not be correct.
