The Python code and cleaned, aggregated data used to produce this article can be downloaded at GitHub
The total balance of student loan debt in the US eclipsed that of credit card debt in 2010, making it the second largest pool of consumer debt behind mortgages. And the student loan debt load continues to grow at a rate that surpasses all other types of debt, totaling nearly $1.5TN at the end of 2018 according to the New York Fed: many (including Secretary of Education Betsy DeVos) have characterized the student loan debt situation in the US as a “crisis”, and some have begun describing it as a “bubble”.
Student loan debt is fundamentally different than all other types of consumer debt. If someone buys a car with a loan and then loses their job, they can sell the car and use the proceeds to pay at least some of the loan down. That can’t happen with a student loan: the asset that the loan paid for is (ostensibly) knowledge, and that can’t be repossessed by a University or sold secondhand. Student loans are also almost impossible to discharge in a bankruptcy, and interest starts accruing on some types of student loans even while the borrower is in school.
But what caused the student loan debt load to get to the level of “crisis” in the United States? In cursory depictions of the student loan situation, many of the same explanations for the sharp increase in the cost of postsecondary education are often offered, from the pedestrian (“University campuses look like the palace of Versailles these days!”, “College kids are majoring in underwater basketweaving!”) to the conspiratorial (the endowments of the most prestigious universities are essentially hedge funds, and by charging tuition that isn’t generally affordable, universities can provide generous financial aid and maintain charity status for their endowment gains). But rapidly rising costs don’t necessarily create bubbles or crises. That the cost of tuition rises doesn’t mean that it must be unaffordable relative to earlier periods, per se.
The crisis of affordability
The affordability of something, and the magnitude to which lack of affordability creates a crisis, is best expressed by simply measuring how many people can afford it — and by this standard, higher education in the United States is distinctly unaffordable. What initially catalyzed the research behind this article was a graphic in a Bloomberg news piece about student loan delinquency that presented something curious: student loan default rates were actually decreasing in 2010 and 2011 before precipitously spiking up to unprecedented levels. That change struck me as tragic: student loan debt service affordability was trending towards being manageable, but something happened and now, just a few years later, the word crisis is used to describe the situation. This reversal also raised a pointed question: why 2012?
The obvious answer is that 2012 is four years — one standard-length university education — later than 2008, the year in which the Global Financial Crisis commenced. The easy explanation for a rapid rise in student loan default rates in 2012 is that more people went to university in 2008 than otherwise would have because the financial crisis caused jobs to be eliminated, and those people were less able to acquire gainful employment upon graduating and subsequently defaulted on their loans.
That answer is mostly convincing, except that student loan default rates (specifically, 90-day default rates — that is, accounts that have been delinquent for at least 90 days) have more or less stayed flat since dramatically and sharply increasing in 2012. The dissipation of the global financial crisis didn’t significantly diminish the growth of student loan debt, and in fact, student loan debt was the only class of US consumer debt to not shrink during the global financial crisis.
The chart above is striking: in 2010, at the height of the global financial crisis, the outstanding volume of every single class of consumer debt shrank except for student loan debt, which grew at more than 10%.
What’s perhaps even more astonishing is that the compounded annual growth rate (CAGR) of student loan debt in the US between 2005 and 2018 is almost 10%, more than any other class of consumer debt. What this means is that, on average, student loan debt grew by 10% every year from 2005 to 2018.
But, again, the notion to explore within this data is not growing debt loads, but growing delinquencies: if starting salaries for university graduates were increasing more rapidly than student loan burdens, a university education would be seem increasingly affordable over time.
The jobs glut
I was accepted into a two-year graduate program in economics in 2007 that was set to start in 2008. When the news broke in 2007 that Lehman Brothers had just disintegrated, I remember thinking, “I’m really happy that I won’t have to deal with this for the next two years.”
The cohort of university graduates between 2008 and 2010 are amongst the unluckiest in modern history, with lower rates of employment relative to pre-recession levels than graduates during any other recession since the 1980s:
But there’s a problem with looking exclusively at unemployment rates in considering the student loan crisis: although the unemployment rate increased starting in 2008 and then receded back to pre-recession levels, underemployment levels, especially for new graduates, remain obstinately elevated, floating between 40 and 50% since the start of the financial crisis.
The New York Fed defines the underemployment rate as, “the share of graduates working in jobs that typically do not require a college degree.” Underemployment is basically what the underemployed racked up tens or hundreds of thousands of dollars in student loan debt to avoid. And underemployment for recent college grads is both significantly higher and more susceptible to shocks than for all other measured groups.
Unemployment isn’t a pressing concern in the United States right now; in fact, it is at historic lows. But underemployment is a persistent problem across recent graduate cohorts, and it is exacerbated by one other product of the Global Financial Crisis: a sustained increase in college applications.
More graduates, the same number of good jobs
Total enrollment at institutions of postsecondary education grew considerably during the global financial crisis, but it never returned to pre-crisis levels: the CAGR for postsecondary enrollment in the US is nearly 1% between 2005 and 2018.
Nearly 2.5 million more people enrolled at a postsecondary institution in 2018 than did in 2005; the global financial crisis caused an enduring increase in enrollment.
This step-change increase in enrollment corresponds to the step-change increase that can be seen in delinquency rates of student loan debt: the number of people going to college soared during the global financial crisis, pushing a large cohort of recent graduates into an anemic economy that couldn’t fully employ them. And enrollment has only fallen by about 15% from the mid-crisis peak, meaning an elevated number of college graduates is landing in an economy dogged by an underemployment rate that stubbornly refuses to descend below pre-crisis levels.
Why did enrollment rates not revert back to 2005 levels, even though the absolute number of 19–25 year olds increased only slightly? Perhaps the surge in college enrollees during the financial crisis created a stronger social stigma around not attending college; maybe expectation inflation by employers convinced a greater number of people that a college degree is a necessary precondition to a successful career. A recent study by the Upjohn Institute found a meaningful gap between the lifetime earnings of high school graduates and college graduates; perhaps the carnage of the global financial crisis scared high school students into applying to college so they could insulate themselves against future such situations.
Whatever the case, increasing student loan debt levels might merely be a symptom of an underemployment crisis. If underemployment rates haven’t moved while the number of college students has ballooned, it seems sensible to conclude that too many graduates are unable to find the gainful employment needed to cover their student loan debt service. It’s possible that a solution to this could be diverting would-be college students to specialty schools, like the coding camps that proliferate frenziedly in San Francisco. Another solution could be a government program that subsidizes or fully funds degrees in certain critical fields.
It’s important to recognize that the student loan debt problem exists on both sides of graduation day: yes, college has gotten more expensive, but, following the global financial crisis, the average postsecondary education also delivers much less earnings value. Having ended nearly a decade ago, it’s easy to dismiss the global financial crisis as a source of any contemporary ills, but the gloom of its pall hangs over the economy today.
Eric Benjamin Seufert is a media strategist and quantitative marketer who has spent his career working for transformative consumer technology and media companies, including Skype and Rovio. Eric runs Mobile Dev Memo, a trade publication dedicated to advertising and freemium strategy on mobile, and QuantMar, a knowledge-sharing site for performance marketers. Additionally, Eric authored the book Freemium Economics, which was published by Elsevier in 2014. Eric holds an MA in Economics from University College London.