Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/193346 
Year of Publication: 
2018
Series/Report no.: 
IZA Discussion Papers No. 12052
Publisher: 
Institute of Labor Economics (IZA), Bonn
Abstract: 
Although recent evidence suggests that the aggregate gender gap in access to Higher Education in Peru has been closed, differences in enrollment between the poor and the rich are still notably larger among girls. This paper explores the factors behind these gender differences in access to Higher Education. Specifically, we assess whether larger socioeconomic disparities among females can be explained by long-run factors crystalized in Higher Education preparedness (i.e., cognitive and non-cognitive skills), rather than by short-term economic constraints. We employ a rich longitudinal data set that allows for the estimation of a structural model of skill formation from early childhood. Our results show that cognitive abilities are strong predictors of enrollment for both genders, whereas non-cognitive skills are only determinant among boys. We also provide strong evidence of gender-specific short-term barriers in access to post-secondary schooling: while differences in skills are the major determinants of the wealth gradient for males, the female gap remains large even after accounting for these factors. Further analysis reveals that access to Higher Education among girls is overly sensitive to marginal costs of enrollment, suggesting that at least part of this gradient might be explained by lower expected returns rather than credit constraints. Overall, these findings illustrate the importance of early human capital investments on educational attainment, but also point to the prevalence of short-term restrictions that disproportionately affect females in disadvantaged households.
Subjects: 
higher education
gender
skills
JEL: 
I23
J16
J24
Document Type: 
Working Paper

Files in This Item:
File
Size
532.59 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.