Restricted 2SLS/IV Regression

[corydeburd]

Dear Statalist,

I'm trying to run a model that operates like a normal IV except with inequality restrictions on some coefficients. For instance, y = b1*x1 + b2*x2 + c + epsilon, z1 and z2 are exogenous instruments for the x's, and I impose b1 > 0.

Is there a canned Stata (or other) package to run this? If not, where would I begin on coding it?

Thanks for all your help,
Cory

[corydeburd]

Thanks friends, would appreciate any advice on this topic even if it is "I don't think what you're looking for exists."