Search

In this article, we present the eqregsel command, which estimates and provides bootstrap inference for sample-selection models via extremal quantile regression. eqregsel estimates a semiparametric sample-selection model without an instrument or a large support regressor and outputs the point estimates of the homogeneous linear coefficients, their bootstrap standard errors, and the p -value for a specification test. [ABSTRACT FROM AUTHOR]

In this paper we analyze career dynamics for US workers who have more schooling than their peers in the same occupation. We use data from the NLSY79 combined with the CPS to analyze transitions into and out of overeducated employment, together with the corresponding effects on wages. Overeducation is a fairly persistent phenomenon at the aggregate and individual levels, with 66% of workers remaining overeducated after 1 year. Overeducation is not just more common but also more persistent among blacks and low-AFQT individuals. Further, the hazard rate out of overeducation drops by about 60% during the first 5 years spent overeducated. However, the estimation of a mixed proportional hazard model suggests that this is attributable to selection on unobservables rather than true duration dependence. Lastly, overeducation is associated with lower current as well as future wages, consistent with scarring effects. [ABSTRACT FROM AUTHOR]