Vocational Education and Employment over the Life Cycle

Not obtained at abstract depth; PIAAC 2012

PIAAC 2012, cross-national adults.

Logistic regressions with country fixed effects on cross-sectional PIAAC data. The weakest design in this cluster — grade D rather than C, because it is correlational work with no attempt at exogenous variation. It is included because it is the only INDEPENDENT test of the mechanism Hanushek et al. propose, and it fails it.

Confirms that vocationally qualified individuals have a higher employment probability early in the career and that this reverses later. But, in the authors' own words, in contrast to earlier findings they do not find that this effect varies systematically across countries with different vocational educational systems — directly contradicting the occupation-specificity mechanism Hanushek et al. propose. Follow-up work by the same authors places the crossover for men around age 55, roughly a decade later than Hanushek's 43-49.

HIGH. Cross-sectional comparison of self-selected tracks.

An independent team, on the same PIAAC data, reproduces the DIRECTION of the Hanushek reversal but breaks its MECHANISM. Two teams agreeing on a descriptive pattern and disagreeing on its explanation is what a selection artefact looks like from the inside.