Quantile and interquantile regression models for returns to education by economic sector and vulnerable population in Colombia

Abstract

We investigated the returns to education by economic sector in Colombia, focusing on the relationship between educational levels (degree of highest educational level) and wages in different labor areas (economic sectors), as well as vulnerable populations such as women and migrants. Quantile and interquantile regressions were employed, correcting for selection bias through the inverse Mills ratio and using monthly data from Colombia’s Great Integrated Household Survey (GEIH) for 2019, to explore how the effect of education varies at different points of the income distribution and between these points. Using quantile regression provided a more comprehensive view of this relationship than traditional statistical regression approaches. Traditional Mincerian socioeconomic variables such as gender, experience, hours worked, marital status, relationship with the head of the household, and social security affiliation, were controlled for. Results show that while there is a positive effect between educational level and income in all economic sectors studied, this relationship varies in magnitude and form along the wage distribution.