One of my favorite RSS feeds is Business Insider’s Chart of the Day. This chart came up a few weeks ago and made me think. It’s quite staggering to imagine that college tuition has outpaced inflation as rapidly as it has (~10x vs. ~4x over 30 years). The graph made me think: Has the value of a college education increased 2.5x? If it has, then there isn’t necessarily a bubble. There are three ways to think about the value of a college education to evaluate this question.
- The most obvious is the average income comparison between an average high school graduate (only) and an average college graduate (only). Using US government statistics, we find that in 1978, a college-educated graduate made ~55% more than an individual with only a high school education. In 2008 (the last year I have data for), a college educated graduate made ~87% more – which amounts to a 60% increase in the gap. Now there are obviously nuances in that figure (which I’ll let the policy experts correct in the comments if they choose), but this 60% figure is a decent order-of-magnitude impact (and even holds true if I adjust for the change in disposable income using the individual, under-65 poverty line as the base level of expenditure), and barely covers the 2.5x increase in tuition costs over inflation.
- Related to the above are the other effects of college education on lifetime income. It’s been demonstrated that college educated individuals live longer than those who don’t, and its commonly understood that college is a prerequisite for other income-boosting opportunities like graduate school. One’s lifetime income is also much more likely to trend upwards in life with a college education than with only a high school diploma. But there’s an interesting wrinkle here: does college education make you live longer and get promoted, or is it just an indirect way of finding individuals who tend to be wealthier and more intelligent? I unfortunately don’t have the data (or the time 🙂 — this is a casual blog post after all!) to run all the calculations needed to understand the impact, but I would hazard a guess that it’s probably overly-aggressive to assume that these second-order lifetime income benefits can close the gap between tuition costs and inflation.
- The last “source of value” for a college education is the subjective value of meeting lifelong friends, having new experiences, and expanding your intellectual horizons. Just because its last and extremely intangible, doesn’t mean there’s not enormous value here. But, the question to ask here is not whether or not college has large intangible value (of course it does), but whether or not you believe that intangible value to have increased significantly (by over 2.5x as we’ve concluded the increase in lifetime income likely isn’t enough to explain the 2.5x increase in tuition cost relative to inflation) since 1978. I personally think that’s being overly aggressive.
I can’t pretend to be the expert on this, but at this point, I’d conclude that whereas the value of a college education has increased dramatically (maybe even by as much as 75-100%), it hasn’t gone up enough to justify a 2.5x increase relative to inflation. If you accept my conclusion then the obvious question is what is causing tuition to increase so much? Two possible explanations jumped out at me:
- Tuitions haven’t caught up with the value of a college education: While my conclusion above was that the value of a college education hasn’t increased 2.5x from 1978 until today, one possible explanation of tuition price is that the value of a college education is still higher than what students pay for it, and, if that were true, we would expect tuitions to continue to increase.
- Tuitions are higher than the value of a college education and are being propped up by a combination of two things:
- Tuitions are being boosted by a subsidy cycle: Free money from the government is one of the easiest ways to get price increases. While we often think of the US government’s subsidized loans and tax writeoffs for college tuition as a means to help more people attend college, an equivalent way of thinking about it is that it gives colleges free rein to increase prices without worrying about reducing the number of people who enter. In a “normal” market, this would be the end of it (slightly higher prices, but more people entering college), but because college education has become such a political affair (every family always thinks its “too expensive”, and every politician promises to make it cheaper), we always get more subsidies from more politicians which feeds back into the original problem.
- Families have bad expectations around the value of a college education: One explanation, which is making the rounds of the policy wonk blogosphere, is that this is all a big bubble. In the same way that people felt tech stocks in the late 90s and real estate in the 2000s were a good buy, its entirely possible that families have uninformed expectations about the value of a college education and thus believe the higher tuitions are worth it. If this is true, then we could be on a collision course with a generation of families (like in the 80s) suddenly realizing “the emperor (of college tuition) has no clothes!” (which might be precipitated by a long stagnation/decline in the wages of college educated individuals) followed by a potential crash in tuitions.
What should be done? Truthfully, it depends on which of these conclusions is correct. If its just that tuitions haven’t caught up with the value of a college education, then it makes sense that tuitions are increasing, and it may even make sense to increase tuition grants/subsidized loans (as it likely means not enough people are getting a college education because they couldn’t get access to the capital to pay for it). However, if tuitions are over-valued, then it would be advisable to end the college tuition subsidy cycle and implement policies which potentially “soften the blow” of the coming college education value and tuition crash.
But to make the right policy decisions and tradeoffs, its important to first understand which of the two explanations is correct. And that’s a whole ‘nother set of data and analyses…