Is The Case For College Too Good To Be True?

Recent research, The Hamilton Project, mentioned in the New York Times, and on Greg Mankiw's blog, about the value of a college degree states:

On average, the benefits of a four-year college degree are equivalent to an investment that returns 15.2 percent per year. This is more than double the average return to stock market investments since 1950, and more than five times the returns to corporate bonds, gold, long-term government bonds, or home ownership. From any investment perspective, college is a great deal.

15 percent is the unleveraged return. Although college students often borrow funds to pay for tuition, the 15 percent return is computed on the total cost of the education and not on just the students' savings and funds invested, the equity portion, of the investment.

The 15 percent return is a consistent return over time. It is almost Madoff like, but these are the real average returns, not made up numbers. If someone came to professional or knowledgeable investors, and promised them double the stock market return over their working lives, they would think he was a scam artist and fraudster.

It is very difficult, if not outright impossible, to consistently surpass the investment returns of the stock market, yet alone achieve double the market return for many years, especially without leverage or without substantial risk.

Why The Double The Market Return?
Employers do not spend money they do not have to spend. Excessive compensation eats into corporate profits and lowers stock market share price. Why would employers pay a 15 percent return on a college investment when, most likely, employers would offer a lower return than 15 percent and many college graduate employees would accept less than 15 percent? Heck, any return above the risk free rate would make a college education worthwhile.

There are several possibilities to justify a double market return on investment:

• There are wide statistical variances, high risk, in college graduate earnings and college graduates demand compensation for that risk. To earn double the market return, the systemic risk of college grad earnings would have to be double that of the stock market, which is equivalent to double the risk of the US economy. Since college grads have a lower unemployment rate than non-college grads, it is unlikely they face double the systemic risk of the economy. But, that is overall and not on an individual case. There maybe classes of college grads for which their is excessive earnings' risk not identifiable until they enter the workforce and for which they do not have self-knowledge until they are in the workforce.


• The labor market for college graduates is highly inefficient with high transaction costs and employers are willing to pay a high salary for college graduates. However, market and capitalistic forces would work to erase or lower the inefficiency and transaction costs until they disappeared.


• Demand for college graduates exceeds supply to the extent that their premium is double the market return. If true, it raises the question of why employers have not found a substitute strategy for college grads to avoid the high pay premium. Productivity can be increased from increasing capital investment as well as hiring more productive college grad employees.


• There is survivorship bias in the reported results. The average return is computed after the fact. If some college graduates drop out of the labor market and they have lower wages than the colleges graduates that remain, the results will be skewed upwards. Professional investors know that mutual funds, hedge funds, etc. close the funds with low returns and keep funds with high returns to make average returns after the fact look better than they are. [Addendum} The cost of college for non-graduating students has to be included in the calculation. To only look at graduates understates the total risk and total cost to high school graduates who go to college. Not all college entrants will graduate and looking only at graduates understates the amount of investment and overstates earnings because it does not include the cost to high school graduates who enter college but do not graduate and their lower earnings.

Any, or all, of the above possibilities are the explanation for the high college grad wage premium.

Based on my own investment knowledge, I would guess most of the premium is survivorship bias. If all the college grads in any year were followed prospectively and their earnings tabulated going forward, I think the average return for their college investment results would be much closer to the average stock market return over the same period, if not below.

Addendum:
I am not saying a college degree is not a worthwhile investment. I am just questioning the extent of the above market return. I do believe a college investment will yield a positive return, just not double the stock market return.