Earnings over the Life Course: General versus Vocational Education

Not obtained at secondary depth; Swedish population registers

Swedish individuals in short (two-year) upper-secondary programmes, followed across nearly complete careers — the longest earnings horizon in this literature.

Register data with controls for GPA and FAMILY fixed effects, plus adjustment for further education and fertility. The sibling comparison is a real strength and would earn grade B under the archive's within-family rule; it is recorded C because track choice remains self-selected WITHIN family — siblings who chose different tracks differ, and the traits that drove the choice are heritable and are not absorbed by a family effect. Read at secondary depth; no magnitudes recovered.

Descriptive earnings patterns support both halves of the lifecycle trade-off: vocational skills raise relative short-term earnings and general skills raise relative long-term earnings, and the pattern survives controls for GPA, family fixed effects, further education and fertility. This is the best non-experimental support for the trade-off — and it is also exactly what a selection story predicts, which is why it cannot settle the question.

HIGH. Track choice is self-selected, and within-family selection is precisely the residual the family fixed effect cannot absorb. The archive's premise applies at full strength here.

Descriptively supports Hanushek et al.'s trade-off on Swedish registers. But it sits on the same side of the identification line as Hanushek — observational, self-selected track — and on the opposite side from Silliman & Virtanen, who study a comparable Nordic setting causally and find no decay.