Discussion papers // Ibero America Institute for Economic Research 131
Within the attempts to understand Mexican economic inequality, returns to education have received a great deal of attention. The driving question has been: why are Mexican wages so unequal? This paper argues that not only the distribution of human capital matters, but also sociodemographic variables, which have their own dynamics and complex interactions with the former. A three-equation maximum likelihood specification in which employment, hours worked and log-wages, as well as their joint variance matrix is proposed, generalizing the Mincerian specification. The resulting is a complex story, where income profiles depend upon particular characteristics.