Imputación de ingresos laborales: Una aplicación con encuestas de empleo en México

The aim of this paper is to make imputations of earnings to observations with missing earnings in the Encuesta Nacional de Ocupaciones y Empleo (ENOE). We present imputations by two methods and also a correction of estimations by reweighting observations with reported earnings. Then, we analyze the possible bias in estimations of mincer equations and labor poverty derived from ignoring observations with missing earnings. The results show that when missing earnings are not considered in estimations, there are differences in the parameters that define the relationship between the human capital level and earnings, as well as those that describe the determinants of labor poverty, compared to the results obtained from estimations that consider all observations. Differences are more significant when we analyze labor poverty.