Since the work of Tobin(1958) various statistical techniques were developed to estimate censored regression model. Not only parametric estimation technique, but also semi-parametric or nonparametric estimation method were developed. Recently semi-parametric estimation method about simultaneous equations with censoring was developed by Lee(1995).
But in case of nonlinear simultaneous equations with censoring there is no method except for 2QMLE by Hurd and Lee(1992). In real economy we often observe simultaneity, censoring, and nonlinearity.
In this paper a kind of simplified two stage estimation technique on the estimation of nonlinear simultaneous equations with censoring is suggested. This method use kernel-type nonparametric estimation method. Before this method we only depended on ML method. The advantage of our method lie on the avoiding computational difficulty of ML method and in some case we can provide our estimate with the initial value of iterative ML method. Computational results show that the proposed method can be another alternative of ML method.
