< Back to previous page


An Economical Perspective on Early school Leaving in Vocational Education and Training.

This doctoral dissertation employs quasi-experimental empirical methods on administrative population data to investigate the causal relationship between high school dropout and its selected predictors, labour market consequences, and policies. Within this broad research agenda, contributions are made in four specific research topics. I first use coarsened exact matching and treatment effect bounds and find that children who were raised in a relatively modern family structure, a same-sex family, have a lower probability to drop out of high school than children who were raised in a different-sex family. I then consider the consequences of school dropout, by comparing the labour market outcomes of high school dropouts with labour market outcomes of high school graduates in both Belgium and the Netherlands. In Belgium, I use parental educational background as an instrumental variable and find no returns to a high school diploma. Conversely, in the Netherlands, I exploit standardized exit exams in a fuzzy regression discontinuity design and find that, unlike in Belgium, the returns to a high school diploma are positive and can be explained by a high school diploma serving as an important signal of productivity on the labour market. Finally, I evaluate modular education in Belgium, a dropout prevention policy in which conventional courses are divided into smaller components or modules. Using a difference-in-differences framework, I find that modular education is an effective policy to tackle high school dropout, with largest effects on the foreign origin students.

Date:2 Mar 2015  →  1 Mar 2020
Keywords:School Dropout, Quasi-Experimental Methods, Administrative Panel Data
Disciplines:Labour and demographic economics
Project type:PhD project