Improving Educational Pathways to Social Mobility: Evidence from Norway’s Reform 94

Full Norwegian population registers; cohorts born around 1 January 1978, with placebo-year cutoffs

Norway, cohorts eligible or ineligible for Reform 94; earnings measured at ages 25-34 and hourly wages at 28-33.

Difference-in-discontinuity: births just before against just after the 1 January 1978 eligibility cutoff, differenced against the same calendar cutoff in placebo years to strip out school-starting-age effects. A strong design on population registers. The limit is that the treatment is a bundle — more general content inside the vocational track, PLUS a college-eligibility bridge, PLUS more apprenticeships — so it identifies the effect of a REFORMED vocational track, not of vocational education as such. Earnings stop at 34, so it is silent on the lifecycle-decay window that Hanushek et al. place at 43-50.

The reform sharply raised vocational enrolment. For men, vocational diplomas rose roughly one-for-one at the expense of academic diplomas with NO decline in college attendance or completion — the reform pulled in men who were not going to college anyway. Disadvantaged men's average annual earnings at 25-34 rose 5% (+$2,329, 2015 dollars), with reduced teenage criminal charges; the pooled effect is about 2%. For women the earnings effect is null, though disadvantaged women showed about a 6-point (20%) drop in having no high-school credential. Net: the earnings gap between disadvantaged and advantaged children narrowed about 20% overall and 30% among men, while the overall GENDER earnings gap WIDENED by about 8%. Only about 20% of disadvantaged men's gains are explained by the degree-composition shift, implying the rest came from improved vocational-track quality.

Low. Eligibility is assigned by date of birth relative to a policy cutoff, differenced against placebo years. Nobody chose their birth date to get the reform.

Agrees with Silliman & Virtanen on the key point — adult earnings gains from vocational education, no decay through the early-to-mid thirties, and no cost to college-going — using a different country, a different design and full population registries.