The economics of a college degree

By Felix Salmon
July 12, 2010
BusinessWeek has a big feature on the value of a college degree, and naturally has <a href=" " data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

BusinessWeek has a big feature on the value of a college degree, and naturally has presented it in the format of a ranking. College rankings are profoundly silly things, and this one is no exception; it purports to calculate the extra amount of money that graduates of certain universities earn, compared to the amount that they would earn if they hadn’t gone to college, thereby coming up with some kind of dollar figure for value-for-money. It then ranks “30-year net return on investment”, which ranges from $1.69 million at MIT to just $998 at Black Hills State University in Spearfish, South Dakota.

There’s a lot to dislike here, and the BW story twists itself into knots listing one disclaimer after another. For one thing, the survey doesn’t look at how much people actually spend on college tuition: it just looks at the headline rack rates, failing to take into account any kind of grants or student aid. It also uses a 30-year-long dataset, which includes a lot of years when women were chronically underpaid. That penalizes women’s colleges.

What’s more, the survey does a very bad job of quantifying the benefits of a liberal-arts degree. Let’s say you go to college and then earn $45,000 a year working in the theater, or you end up with a steady job in public administration or social services. You’re clearly better off in many different ways than a high-school graduate earning the same amount — you’re probably happier in your job, you’re doing what you want, and you have more job security. But the BW methodology would give you a negative return on your university tuition, on the grounds that you missed out on earning money while you were at college.

There are other problems with the survey, too, which aren’t even hinted at in the BW story. Pay for college graduates has been rising over the past 30 years, while pay for individuals with no more than a high-school education has been falling sharply. But the survey pays no attention to those trends, and assumes that the income prospects for someone with no college education are the same now as they have been, on average, over the past 30 years. Needless to say, that’s ridiculous.

The survey also assumes that students who drop out of any given university will make no more money than if they hadn’t enrolled at all. That’s trivially false in the case of top-ranked universities like MIT and Stanford, where you could probably make the case that dropouts end up making more money than graduates.

And the survey also concentrates solely on income, rather than wealth. Getting large paychecks is one — but only one — way to get wealthy. And I’m pretty sure that college graduates in general are wealthier than non-graduates earning the same amount of money. If nothing else, they have a tendency to marry each other, so even if they don’t end up earning a lot of money themselves, they can still benefit from their spouse’s higher income.

Even with all those caveats, one thing jumps out from the survey, which again BW doesn’t mention: even the very worst-performing colleges — the ones coming 851st and 852nd out of 852, for instance, where fewer than one in three students even graduate — have significantly positive “annualized net ROI”s. John Carney looks at these numbers and concludes that “the ROI on going to college is worse than the S&P” — not when the ROI on the S&P is negative, it isn’t. Clearly anybody who’s got a good chance of graduating from university would be much better going to university than investing their tuition money in an S&P 500 index fund.

The one thing which most annoys me about the survey, however, has nothing to do with the methodology, and rather the way that the results are presented: you have to do a lot of clicking and scrolling to try to read them, and they’re not searchable. So while the results carefully make the distinction between public and private colleges, they don’t make the distinction between for-profit and non-profit private colleges. Are there any for-profit colleges in the list? How do they stack up? That’s one datapoint I’d be fascinated to see, but I can’t find it anywhere.


Comments are closed.