La elección del sector laboral y los retornos a la educación en Guatemala[Labour Sector Choice and the Returns to Education in Guatemala]

This paper estimates the returns to education in Guatemala, while attempting to account for self-employment and the presence of workers without monetary earnings in the economy, factors whose omission can potentially lead to sample selection bias. The analysis uses data from the Survey of Living Conditions (ENCOVI 2000) to obtain estimates of the rate of returns to different levels of schooling using a human capital earnings function. A multinomial logit model for sector choice is implemented to correct for selection bias as in Bourguignon, Fournier and Gurgand (2001). Non-linearities in the returns to education are strongly supported by the empirical results. It is found that OLS estimates which do not account for sector choice significantly overestimate the returns to primary and early secondary schooling, while underestimating the returns to late secondary education. The results also show lower rates of return in the self-employment sector and that workers who only complete primary or early secondary schooling have a higher probability of entering this sector.