06. Education — The Elite-University Brand and Teacher Quality

Overrated · The elite-university brand effect

College-admissions industry and rankings-driven media treat a selective university's name as itself causing higher earnings. Stacy Berg Dale and Alan Krueger tested this using the College and Beyond dataset, students entering college in 1976 and 1989. They matched students admitted to and rejected from comparable colleges, controlling for the traits that get students admitted in the first place. Once selection bias was controlled for, the more selective school produced no earnings premium on average. The average tuition of schools applied to — not the school attended — predicted earnings better than selectivity itself.

Underrated · Teacher quality

Critics of "value-added" scoring call a teacher's effect on test scores a noisy statistical artefact, not a durable difference. Raj Chetty, John Friedman and Jonah Rockoff tracked about 1 million US students from elementary school (1989–2009) into early adulthood via linked tax records (1996–2011). They used teacher turnover as a quasi-experimental design to rule out bias from student sorting. Students assigned to higher value-added teachers were more likely to attend college, attended higher-ranked colleges, and earned more as adults. Replacing a bottom-5%-value-added teacher with an average one raises a single classroom's lifetime income by about $250,000 in present value.

The gap

Dale and Krueger's headline finding — no average earnings premium once selection is controlled for — is the version that travels. The same 2002 paper also found a real earnings premium for disadvantaged students who attended the more selective school; that half rarely travels with it. Economist Jesse Rothstein and others have challenged whether Chetty et al.'s value-added estimates are free of bias from non-random student assignment; the design was built to answer that critique, and the debate continues. Both studies draw on US data only, so this chapter cannot claim the global coverage the rest of the book aims for.

Example: The brand on the diploma predicts less, on average, than the tuition bill of the schools a student merely applied to; the name on the classroom door, tracked for two decades, predicts six figures of lifetime income.

Verdict. The elite-university brand is oversold as an average effect and undersold as a targeted one; teacher quality is undersold across the board.

Overturned if: a replication outside that sample finds a general selectivity premium regardless of background, or a study outside the US finds Chetty et al.'s value-added effect does not hold.

Sources. 2002 · Stacy Berg Dale & Alan B. Krueger, "Estimating the Payoff to Attending a More Selective College," Quarterly Journal of Economics 117(4):1491-1527 — academic.oup.com · 2014 · Raj Chetty, John N. Friedman, Jonah E. Rockoff, "Measuring the Impacts of Teachers II," American Economic Review 104(9):2633-2679 — aeaweb.org