Evaluating Education Investments: The Rates of Return Approach

When governments and individuals spend on education, are they making a sound investment? The rates of return approach provides a framework to answer this question by treating education as an economic investment, much like purchasing stocks or real estate. This method compares the costs of obtaining education-tuition, fees, and time spent studying-against the lifetime benefits, primarily in the form of higher earnings. Understanding how economists measure these returns helps policymakers allocate resources efficiently and helps individuals make informed decisions about their educational pursuits.

Private versus social returns: two perspectives on education investment 🔗

The rates of return approach distinguishes between two fundamental types of returns: private returns and social returns. These represent different perspectives on who benefits from educational investment and who bears the costs.

Private returns: the individual’s perspective 🔗

Private returns measure the benefits of education from the individual’s point of view. When calculating private returns, economists compare what an individual invests in education-tuition fees, books, and foregone wages while studying-against the higher earnings they expect to receive throughout their working life. According to World Bank research, the global average private rate of return to a year of schooling is approximately 9-10%, meaning each additional year of education generates about a 10% increase in annual earnings.

Private returns serve several practical purposes. They help individuals decide whether pursuing further education makes financial sense compared to alternative investments. They also inform policy decisions about student loan program design, helping set appropriate interest rates based on expected returns to borrowers.

Social returns: society’s perspective 🔗

Social returns take a broader view by including costs and benefits that extend beyond the individual. On the cost side, social returns account for government expenditures on education-money spent on buildings, teacher salaries, and administrative costs-in addition to private expenses and foregone earnings.

The benefits side of social returns is more complex. Beyond higher tax revenues from better-paid workers, education generates numerous positive externalities that benefit society as a whole. Research shows that educated individuals are less likely to commit crimes, more likely to vote and participate in democratic processes, and more likely to engage in behaviours that improve community health outcomes. Education also creates knowledge spillovers: workers become more productive when they work alongside highly educated colleagues.

Private rates of return typically appear higher than narrow social rates because governments heavily subsidise education. However, if economists could fully account for all social benefits-including reduced crime, better public health, and increased civic participation-social returns might well exceed private returns.

The internal rate of return: measuring profitability 🔗

To quantify returns to education, economists primarily use two methods. The most widely employed is the Mincer earnings function, named after economist Jacob Mincer, which estimates returns by examining how wages increase with additional years of schooling while controlling for work experience. The second approach is the internal rate of return (IRR), a discounting method that determines the interest rate at which the present value of educational costs equals the present value of future earnings benefits.

How the internal rate of return works 🔗

The IRR treats education like any financial asset with an initial investment and subsequent income stream. According to the Federal Reserve Bank of St. Louis, if one compares tuition costs with the higher incomes that college-educated workers earn-as if comparing the price and payoff of a financial asset-the estimated returns appear quite substantial.

Recent estimates indicate that annualised returns on a college education in the United States range from 12-14% after adjusting for inflation. For comparison, the average real return on U.S. stocks over the past century is around 7%. These calculations assume the individual successfully completes their degree within a reasonable timeframe.

Returns vary by education level and context 🔗

Returns to education are not uniform across all levels of schooling or all countries. Research compiled over decades shows that returns tend to be higher in low-income countries, where educated workers are scarcer. Primary education historically showed the highest returns in developing nations, though tertiary education now shows the highest returns even in low-income countries-a phenomenon economists describe as a “race between education and technology.”

Women consistently experience higher returns to schooling than men, with global averages showing female returns around 11.5% compared to 9.6% for men. This pattern holds across most countries and education levels, suggesting that investing in girls’ education remains particularly valuable.

Challenges in estimating returns to education 🔗

While the rates of return approach provides valuable insights, several methodological challenges complicate accurate estimation. Researchers must account for factors that can bias results either upward or downward.

Ability bias: overestimating the education effect 🔗

Perhaps the most discussed estimation challenge is ability bias. The concern is straightforward: people with traits the labour market values-intelligence, work ethic, perseverance-tend to both obtain more education and earn higher incomes. If employers can detect these valued traits, people with more education would earn above-average incomes even if their education itself had no effect on productivity.

Studies using twins help isolate this bias. Research comparing identical twins with different education levels found that controlling for shared genetics and family background still leaves significant returns to schooling, though approximately 12% of the estimated schooling impact may be attributable to ability differences. Another approach involves controlling for measured ability directly; studies controlling for IQ scores suggest that standard estimates may overstate returns by roughly 40%.

Selection and sorting effects 🔗

A related issue involves selection bias-the possibility that individuals who benefit most from education are precisely those most likely to pursue it. As researchers have noted, when the causal effect of education varies across individuals, there is no single “true effect.” The returns experienced by someone who attends university may differ substantially from those a non-attender would have experienced had they enrolled.

Economists address this through instrumental variable techniques, using factors like changes in compulsory schooling laws or school construction programmes as natural experiments. These approaches isolate the causal effect of education from self-selection, though they estimate effects for specific populations that may not generalise.

Adjustments for taxes and earnings growth 🔗

Accurate return estimates require adjustments for taxation and realistic earnings projections. Since workers pay taxes on their earnings, post-tax returns differ from pre-tax calculations. Social returns use pre-tax earnings differentials because taxes represent transfers rather than resource costs, while private returns should reflect after-tax income.

Earnings growth assumptions also matter significantly. The Mincer equation estimates an average marginal rate of return, but returns actually vary with years of schooling and can decrease over the working lifecycle. Using historical cross-sectional data to predict future earnings becomes problematic when economic conditions change substantially, as technological developments and economic complexity alter the value of different skills over time.

Measurement and data limitations 🔗

Practical data constraints introduce additional complications. Earnings data typically come only from employed workers, implicitly assuming zero unemployment. Since less-educated workers generally experience higher unemployment rates, this omission tends to understate education’s true economic value. The St. Louis Fed notes that calculations also assume workers survive to retirement age, potentially overstating returns for groups with lower life expectancy.

Furthermore, standard calculations exclude non-monetary benefits. Highly educated individuals report better health outcomes, are less likely to smoke, experience less depression and stress, and have longer life expectancy. If these health benefits were monetised, the rate of return on education investment could double compared to estimates based on wages alone.

Why returns to education remain relevant 🔗

Despite estimation challenges, the rates of return approach continues to guide important policy decisions. Governments use return estimates to determine where educational investments will have the greatest impact. The evidence consistently shows that education represents a highly profitable investment for both individuals and societies-often outperforming alternative investments in physical capital or financial markets.

The approach has evolved considerably since George Psacharopoulos published his landmark comparative study fifty years ago. Researchers now employ sophisticated econometric techniques, use instrumental variables to address selection bias, and increasingly account for non-monetary returns. Yet the fundamental insight remains: education is not merely an expenditure but an investment in human capital that yields measurable economic returns.

What do you think? Given the challenges in measuring returns accurately, how should individuals weigh economic considerations alongside other factors when making educational decisions? And how might changing labour markets and technological advancement affect these calculations in coming decades?

References

1. https://blogs.worldbank.org/en/education/50-years-after-landmark-study-returns-education-remain-strong
2. https://wol.iza.org/uploads/articles/278/pdfs/estimating-return-to-schooling-using-mincer-equation.pdf
3. https://documents1.worldbank.org/curated/en/442521523465644318/pdf/WPS8402.pdf
4. https://economist.asu.edu/universities-knowledge/high-return
5. https://www.stlouisfed.org/publications/regional-economist/2023/mar/return-investing-college-education
6. https://www.econlib.org/archives/2012/01/correcting_for.html
7. https://www.sciencedirect.com/science/article/pii/S0272775798000338
8. https://pmc.ncbi.nlm.nih.gov/articles/PMC3102252/