The impact of vocational training interventions on youth labor market outcomes: a meta-analysis

This paper is part of the output of the project 'Linking Education and Labour Markets: Under what conditions can Technical Vocational Education and Training (TVET) improve the income of the youth?' (short title: LELAM TVET4Income). The project, financed jointly by the Swiss National Science Foundation (SNSF) and the Swiss Agency for Development and Cooperation (SDC), aims to find out under what conditions and to what extent TVET can help to improve the labour market situation of the youth, especially in east developed, low and middle-income countries. The project consists of six teams coming from five different countries and four continents: Chile, Costa Rica, Benin, Nepal and Switzerland.

We summarize more than 1,600 effect size estimates from 89 studies evaluating the impact of vocational training on youth labor market outcomes globally. Two-thirds of effect sizes derive from experimental evaluations. Using Hedge's g as outcome measure, we employ robust variance estimation (RVE) to account for statistically dependent effect sizes. Univariate meta-regressions show a more than two-times larger impact of training on labor market outcomes (g=0.12***) than in related meta-analyses. Multivariate meta-regressions suggest that: (1) pure in-classroom or workplace-based training is less effective than combining both; (2) training affects outcomes through human capital accumulation in high income countries, while it acts a signaling and screening device in low- and middle income countries; (3) involving non-public actors in the design or implementation of interventions yields larger effect sizes; (4) the impact of training on conditional, e.g. formal employment or wages, is lower than on unconditional outcomes; (5) intention-to-treat impact estimates are relatively lower; and (6) effect sizes are smaller in male and female sub-samples than in mixed samples.

Edited excerpts from publication and publisher's website. Show less