Reading the Research on Education and Earnings: Twins and Birth Months

Why this is hard

In many countries, people with more years of education tend to earn more. The problem is that a simple comparison cannot tell you whether the gap comes from education itself, or from the fact that people with more ability or more family support tend to stay in school longer. Economists have spent decades finding ways to separate the two. Among all the research on learning and income, this is one of the longest-running debates.

What the research found

Using birth months (US). Angrist and Krueger (1991) noticed that in the US, the combination of school starting age and compulsory schooling laws means people born early in the year can legally leave school with less education than those born late in the year. Roughly 25% of potential dropouts stayed in school because of these laws. Using quarter of birth as an instrument, they estimated a return to education close to the ordinary estimate, suggesting that conventional estimates are not badly biased.

Using a change in the law (UK). Oreopoulos (2006) used the rise in the UK’s minimum school-leaving age from 14 to 15. The change led almost half of the 14-year-olds of that era to stay in school one more year. Estimates from this natural experiment showed that the benefits of the extra schooling were very large, even though the law affected a large share of the population rather than a small, unusual group.

Comparing identical twins (US and Japan). Identical twins share their genes and their family background, so comparing them removes much of the difference in ability and upbringing. Ashenfelter and Rouse (1998) studied about 700 identical twins in the US and estimated an average return of about 9% per year of schooling. Simple cross-section estimates were biased upwards, but only slightly. In Japan, Nakamuro, Inui and Yamagata (2017) also used twins and, after correcting for ability differences and errors in self-reported schooling, estimated a return of 9.3%, concluding that conventional estimates in Japan are not seriously distorted either.

The big picture. Card (2001) reviewed studies like these and found that estimates using compulsory schooling laws or differences in access to schools are typically as large as, or larger than, conventional estimates.

How to read it

These studies measure average tendencies using designs that come as close to cause and effect as observational data allow. Even so, the effect for people who stayed in school longer because of their birth month or a law may not apply in the same way to a working adult deciding whether to go to graduate school. Twin studies, too, cannot fully explain why one twin studied longer than the other.

It also matters that “education” here mostly means years of formal schooling. For adult learning, studies from Germany, the UK and Japan often find that the link with pay shrinks or disappears once ability and other factors are taken into account (see our articles on training in Germany and self-study in Japan). Don’t carry the schooling results straight over to adult learning.

A small step for today

Take one reading habit away from this. Whenever you see a story about an income gap, ask: are the two groups being compared also different in ways other than learning? Twins and birth months are exactly the tools researchers use to check that.

Then use “Studied for 30 minutes” in the app as an amount you can keep up without expecting quick results. What the research shows are average differences over years and decades.

Sources

• [X-card-2001] Card (2001) Estimating the Return to Schooling: Progress on Some Persistent Econometric Problems. Econometrica.
• [X-angrist-krueger-1991] Angrist & Krueger (1991) Does Compulsory School Attendance Affect Schooling and Earnings? The Quarterly Journal of Economics.
• [X-oreopoulos-2006] Oreopoulos (2006) Estimating Average and Local Average Treatment Effects of Education when Compulsory Schooling Laws Really Matter. American Economic Review.
• [X-ashenfelter-rouse-1998] Ashenfelter & Rouse (1998) Income, Schooling, and Ability: Evidence from a New Sample of Identical Twins. The Quarterly Journal of Economics.
• [X-nakamuro-2017] Nakamuro, Inui & Yamagata (2017) Returns to Education Using a Sample of Twins: Evidence from Japan. Asian Economic Journal.