NEHURDLE: Stata module for estimation of models with corner solutions

nehurdle estimates models for dependent variables with corner solutions at 0. It collects the following maximum-likelihood estimators: Tobit (Tobin (1958)), Truncated Hurdle (Cragg (1971)), and Type II Tobit. It allows for both linear and exponential specification of the value equation, as well as for modeling exponential (multiplicative) heteroskedasticity, like in Harvey (1976), in both the selection and value processes where appropriate. In version 14, Stata introduced churdle, a command that allows estimations of models with bounded dependent variables. churdle is, in fact a Truncated Hurdle estimator that allows linear and exponential specifications of the value equation, as well as modeling heteroskedasticity both in the selection and the value equation. nehurdle differs from churdle in that nehurdle works on versions 11 and later, not just 14 and later, in that nehurlde only works on variables that are bounded from below at zero, a subset of variables on which churdle works, and in that nehurdle also has the Tobit and Type II Tobit estimators for linear and exponential specifications of the value equation that allows modeling of heteroskedasticity in the value and selection processes, while churdle only does this with the Truncated Hurdle estimator.