New Estimator Solves Bias in Models, Improving Accuracy of Economic Predictions.
The article introduces a simple method to estimate models with censored endogenous variables and sample selection bias. By using generalized residuals, the approach corrects for inconsistencies caused by the endogeneity of the censored variables. The study also presents two new tests to check for weak exogeneity. An example analyzing the trade-off between fringe benefits and wages shows that the estimation method is effective. Additionally, a simulation indicates that the new tests have strong detection abilities.
