Semiparametric two-stage estimation of sample selection models subject to Tobit-type selection rules (Q1318975)

The distributional restrictions of the commonly used econometric models of choice and sample selection are relaxed here in terms of semiparametric methods involving two stages. The model equations are \[ y_ 1 = x_ 1 \alpha_ 0 + u \qquad \text{and}\qquad y_ 2 = x_ 2 \beta_ 0 + v \] where \(\alpha_ 0\) and \(\beta_ 0\) are the true parameter column vectors and the values of \(y_ 1\) and \(y_ 2\) are observable only when they are positive. It is a censored regression model if \(\max\{0,y_ 1\}\) and the corresponding \(x\) are observable. But it is a truncated regression model if the positive values of \(y\), and their corresponding \(x\) are observable. An estimator is proposed for \(\beta_ 0\) and Monte Carlo simulations are performed to compare the performance of this estimator.