Eric Hanushek on the Education, Skills, and the Millennium Development Goals
What this covers
Eric Hanushek of Stanford's Hoover Institution joins Russ Roberts to argue that what matters for economic growth is not how many years students spend in school, but what they actually learn. Hanushek presents evidence that years of schooling is a deceptive measure of human capital—a ninth-grader in Peru, despite matching enrollment records with peers in Singapore, performs like a fourth-grader on comparable tests. His central claim rests on a stark empirical pattern: roughly 75% of variation in growth rates across countries can be explained by measured cognitive skills in math and science. This holds even more strongly in developing economies, where the gap between enrollment and actual learning is widest. Hanushek argues the Millennium Development Goals should therefore pivot from measuring inputs vs outputs—from counting years attended to verifying that students reach a functional skill threshold, or PISA Level 1.
The conversation moves through the causal machinery behind this relationship. Hanushek uses U.S. Census data on immigrants—comparing earnings of people educated in their home country against those educated in America—to isolate skill differences from institutional and cultural factors. He argues that institutions like external examinations create incentives for learning, establishing an indirect link between school structure and growth through skill acquisition. The discussion turns to why realizing these gains is hard: teacher quality dominates school outcomes, yet education sectors resist paying for performance or allowing choice. Evidence from Pakistan's private schools—which outperform government schools despite paying teachers a fraction of the salary—suggests competition can overcome institutional inertia. Hanushek acknowledges that complementarity of skills and institutions matters; skills yield higher returns in open economies with secure property rights. The economic gains at stake are enormous: projections across 76 countries show that lifting current students to baseline competence is worth roughly six times GDP in present value.
Hanushek argues that the quality of education—measured by actual cognitive skills on math/science tests—rather than mere years of schooling is what causally drives long-run economic growth, and that international development goals should therefore target skill outputs, not enrollment inputs.
- Years of schooling is a poor proxy for human capital because many students complete grades without learning basic skills
- Cross-country variation in growth rates is largely explained by measured cognitive skills, and immigrant earnings data suggest the relationship is causal
- Realizing skill gains requires better teachers, accountability, and choice, which face institutional resistance
Parents demand expert credentials for medical care but passively accept school decisions despite unstable educational practice.
- Although parents know little about medicine, they carefully scrutinize and select healthcare providers for their children, yet they tend to passively trust school boards, curricula, and administrators as experts even though what is considered best educational practice changes with fads and is rarely the same five years later.
“we are very careful about how we consume the people who provide medicine and health for our children. It's interesting to me that we struggle to do that with education.”