Is College Worth the Cost?: A Data-Driven Analysis of the Student Loan Bubble
As of July 2013, total student loan debt reached $1.2 trillion, surpassing credit cards and auto loans as the largest non-mortgage form of household debt in the United States. Given the rise of student loan debt, there has been much concern as to whether the value of a college education is worth the costs in tuition, fees and interest paid on loans. This has led many individuals and media outlets to raise the possibility of the existence of a student loan bubble.
Ever since student loan debt surpassed the amount of credit card debt in the nation to the tune of $830 billion in June 2010, professional and consumer publications have speculated about the possibility of a student loan bubble. In August 2010, a Wall Street Journal article quoted one source as saying, “The growth in education debt outstanding is like a lobster…The increase in total student debt occurs slowly but steadily, so by the time you notice that the water is boiling, you’re already cooked.” In March 2012, the Washington Post reported that more than 80% of bankruptcy lawyers have seen a substantial increase in the number of students seeking relief for student loans. The topic has recently come to light in professional publications. In July 2013, Professor Mitchell Franklin of Syracuse University wrote in The CPA Journal that, “the student loan industry will be the next area of failure that will cost taxpayers significantly.” Lastly, in a scathing editorial in Rolling Stone magazine, Matt Taibbi criticizes the federal government for making it easy to borrow money for higher education “saddling a generation with crushing debts and inflating a bubble that could bring down the economy.”
Before going into the intricacies of the student loan bubble, we must first develop a definition. Ivana Kubicova and Lubos Komarek define a bubble as an “explosive and asymmetric deviation of the market price of an asset from its fundamental value, with the possibility of a sudden and significant reverse correction.” While this definition is good in its description, it is too specific in its explanation of the bubble’s correction. A bubble does not necessarily need to “pop” as Kubicova and Komarek imply when they talk about a “sudden and significant reverse correction.” Bubbles may also deflate slowly and softly so as to garner little attention. Additionally, Kubicova and Komarek include bubbles that do not collapse in their definition by opening up the “possibility” of a correction. Since a significant number of asset-price bubbles are truly discovered only after the correction sets in, we will only include bubbles that collapse in our definition. An alternate definition can be found in the work of Hans Lind, who writes, “There is a bubble if the (real) price of an asset first increases dramatically over a period of several months or years and then almost immediately falls dramatically.” He goes on to give specific measurements to what quantifies a “dramatic” change in prices. However, including “dramatic” in the definition widens a bubble to include such things as a country’s national debt, which can increase gradually but exponentially over a long period of time. For the purposes of this paper, we can combine Kubicova and Komarek’s and Lind’s definitions to create our own frame through which to see the student loan bubble. Thus, an asset-price bubble will be defined as an asymmetric deviation of the market price of an asset from its fundamental value over a period of several months or years, that inevitably leads to a significant reverse correction as seen by a fall in prices.
In Irrational Exuberance, Robert Shiller describes several characteristics of asset-price bubbles. Among them, he mentions that consumer culture and the news media oftentimes propagate bubbles by contending that the asset is a good buy. To apply Shiller’s analysis to student loans, the current generation may have the mindset that the high cost of college tuition will eventually balance out in the long-run due to expected future earnings. College degrees have become the new high school diploma due to the switch to a services economy, which requires a greater knowledge base in order to obtain a job.
Shiller also mentions that bubbles oftentimes have a feedback loop, describing the initial increase in prices leading to more price increases due to speculation that prices will increase. While this loop is not necessarily found in the student loan market, I will posit that it is caused by a general demand for college. Students see their peers paying high prices to attend college and believe that the investment is worth it as a result. “If everyone’s doing it, then it couldn’t be a bad idea!” Because of this, demand increases, pushing up prices even higher and causing even more students to attend college (because they see an even larger number of students going to college as well).
The “Fragility of Anchors” is a topic that Shiller bases an entire chapter of his book on. He describes a situation in which people have difficulty thinking ahead to contingent future decisions. Decisions are made based on how individuals feel rather than a logical assessment of the situation. This rings eerily similar to students making a decision on their major without having a clear idea of their career path. Even the decision to go to college is oftentimes made without weighing the costs and benefits associated with loans and tuition.
Currently, politicians are exploring proposals to alter student loan interest rates, which may inflate or deflate a possible bubble. The policies currently being debated are integral in determining the country’s long-term economic health. Policymakers must be aware of the potential for another asset-price bubble in the student loan market, if one exists, and should shape their stances on political issues around the possibility of a looming burst in the future. The “popping” of such a bubble would mean lower tuition prices, but demand for college would have to fall as well. A generation of students may deem college “not worth the cost” as the market price adjusts to the fundamental value of college. This would lead to lower enrollment in college and an undereducated population. In the long-term, this means a bout of structural unemployment and millions of students defaulting on their debt. This may cause bankruptcies from loan agencies and reduced confidence in the value of the American worker. Past bubbles, such as those mentioned previously, must be studied in order to gain a greater knowledge of the student loan bubble; however, examination of the peculiarities of the student loan market must also be taken into consideration in order to form policy proposals specific to the market. The purpose of this paper is to examine the possibility of a student loan bubble.
The two most recent bubbles in US history include the Dot-Com boom of the late 1990’s and the Housing Crisis of 2008-2009. The Dot-Com bubble was launched by the increasing popularity of the Internet and advent of start-up software companies that produced their products at a relatively low cost. Economists envisioned a “New Economy” in which inflation was virtually nonexistent. This is because the “Old Economy” was made of brick-and-mortar businesses in which the traditional rules of economics applied. The “New Economy,” which would consist mainly of virtual businesses and financial institutions, made economic data irrelevant to the success of these companies. IPO “squatters” bought names of typical addresses (www.business.com; www.loans.com) and trademarks (say, the World Wrestling Federation) for $50 to $100 dollars and waited for owners and investors to purchase them for exorbitant amounts. Venture capitalists saw claims for IPO addresses to be the wave of the future and invested heavily in start-ups with little experience and minimal business plans. The NASDAQ stock index exploded from 600 to 5000 points between 1996 and 2000 as dot-com companies raised billions of dollars overnight.
By early 2000, investors realized that the dot-com dream had devolved into a speculative bubble. The NASDAQ plunged from 5000 to 600 by 2002. Former start-ups such as Microstrategy went from $3500 to $4 per share. The “New Economy” concept became a fallacy. Investors were pouring money into the stock market and ignoring warning signs that the economy was going to head into a recession, losing millions of dollars as a result. Concurrently, accounting scandals erupted, adulterating consumers’ confidence in big businesses and the US stock market. While the Federal Reserve slashed interest rates in an attempt to stop the bleeding, the NASDAQ has never recovered to its 1996 level since.
The crisis in the US housing market has been seen as one of the largest speculative bubbles in economic history. Between 2000 and 2006 home prices rose dramatically (about 12% per year), fueling a home construction boom. The perceived value of houses can be measured in housing prices, while the fundamental value can be seen in the rental value of a home. The fundamental value of an asset equals the sum of its future payoffs. Rent can be used as the fundamental value of a house, because the payoff a house yields is in the form of the roof over the head of the occupant. Therefore, the present value of rentals in the future can be approximately measured by the rental value of the house. The “U.S. House Prices vs. Owner-Equivalent Rent” graph shows the deviation of housing prices from their fundamental value. The data was collected using the S&P/Case-Shiller measures for House Price Index and Owner-Equivalent Rent. 1999 was used as the base year to adjust home and rent prices to the Consumer Price Index. Dividing the house price index by the rent index, you would obtain a price-to-value ratio (see table). This ratio shows the overvaluation of housing prices compared to the fundamental value. If the price-to-value ratio is high, this means that housing is severely overvalued. The table indicates that the ratio peaked between 2006 and 2007, which was right before the bursting of the housing bubble. This eventually led to a correction in which housing prices became closer to rents, causing the ratio to drop and ultimately resulting in the “popping” of the bubble.
The crisis originated when credit for home loans became easier to come by, especially for low-income families. Between 2004 and 2005, the share of subprime (or risky) mortgages jumped from around 2% to 14% of all mortgage originations. Lenders eventually discovered that they could bundle up these subprime mortgages and sell them off to investors in a process known as securitization. Banks would gather thousands of these loans into a “pool,” divide this pool into shares and sell the shares as securities. Buyers of these securities would then gain the right to collect mortgage payments made by these homeowners whose mortgages have been pooled. These specific securities were called “mortgage-backed securities.” Investors underestimated the riskiness of these loans, believing that the value of these mortgage-backed securities was much higher than what it actually was. These beliefs were propelled by rating agencies such Standard & Poor’s, Moody’s, and Fitch, which gave many of these sub-prime mortgages a triple-A rating of very secure, causing investors to overlook how risky the loans actually were.
Eventually, the supply of houses exceeded demand. Defaults on mortgages began to increase due to a general decline in economic activity, pushing supply forward and causing prices to drop. As housing prices began to fall in 2007, a vicious cycle, or feedback loop, was created. Delinquencies increased as more people realized they could not pay back the loans they took out on their homes. As the default rate increased, the value of mortgage-backed securities fell, causing major losses for the banks that provided these loans. Banks tightened credit as a result, reducing household and business spending, which further caused housing prices to decline. Between October 2007 and April 2009, the financial sector’s mortgage-related losses went from $240 billion to $1.4 trillion as estimated by the International Monetary Fund. Confidence in the international banking system eroded as investors were fearful that what happened in the United States would happen at home. By the end of 2009, the Federal Reserve slashed the federal funds rate from 5% to nearly 0% in a last-ditch attempt to stimulate the economy. A global recession ensued nonetheless, bringing the unemployment rate over 10% by mid-2009.
Value of the Asset
The approach that was used in studying the housing bubble can also be used to study a bubble in student loans; however, there are certain complications with using a similar approach. This is because the value of a house differs significantly from the value of a college degree, which brings up an interesting point of comparison given our definition of an asset-price bubble. If a bubble means an asymmetric deviation of an asset from its fundamental value, this value can be determined by looking at what those earning a college degree get out of the asset: the present value of their future earnings as a result of the degree. The market price is easy enough to determine given the overall cost of college including tuition, room and board, and other college-related expenses (discounting for financial aid, scholarships etc.).
Also different from the housing crisis is the nature of the asset. Unlike a mortgage, there is no re-sellable asset associated with a student loan. The asset in this case is the certification one obtains in order to pursue a career in the form of a college degree. If the value of a house declines, the house is still there. The lender can repossess a house if the borrower fails to pay back the loan. A college degree cannot be repossessed, causing the lender to take on a majority of the risk. In housing, the use value one obtains from a house is shelter, which may differ from the market price of the house. In student loans, the use one obtains from a college degree is the prospect of a high-earning job, which also differs from the tuition one pays to receive a degree. Whether or not the use value and market price are connected in the student loans market as they are in the housing market is something to be further explored in the paper.
Additionally, going into debt to purchase a house is a one-step action. People do not need to go through a four-year process in order to obtain the asset and rarely fail to finish building or purchasing the house after a significant investment is made. Once the mortgage is awarded, the house is purchased and the asset obtained by the buyer. With student loans, individuals can put down a significant amount of money without ever obtaining a college degree if they fail to graduate. In 2011, the graduation rate for undergraduate students pursuing a bachelor’s degree at a four-year institution was 59% within six years. While this figure does not account for individuals who have transferred to different institutions, it is fair to say that a significant portion of the population invested money in their education without getting anything in return. This means that any debt taken on by those who fail to graduate is difficult to pay back, because these individuals will presumably be earning the income of someone who has not graduated from college, but will be required to pay for the tuition they took out a loan for. As far as an asset-price bubble is concerned, this may account for a large portion of the overall amount of student loan debt.
Detecting a Bubble: Analyzing Potential Sources
The Cost-Earnings Differential
If a bubble in student loans were to exist, then the price of college would significantly exceed its fundamental value. Tuition prices would have to be growing past the earnings one obtains from receiving a degree. My analysis is the inverse of this phenomenon. I compare the price of college to the earnings and track its progress over the past decade. The price, or investment, in the context of student loans would be the cost of college. The “fundamental value” of college can be measured through an earnings differential between high school graduates and college graduates (bachelors and higher). This is because the economic benefit of college is the higher income you receive over a high school diploma. The earnings differential data was obtained from the US Census Bureau’s inflation-adjusted annual median earnings for both high school and college graduates. Because it is difficult to predict earnings over the average working life of a person, the weighted median annual earnings must be smoothed to account for volatility. This aids in forecasting by ensuring that changes in the earnings differential were not impacted by recent economic conditions. While some may argue that the value of college has only recently been considerably high, smoothing tells us that this has been a long-term trend rather than a short-term phenomenon. Forecasting forty years into the future can be extremely difficult with only ten years of data, but the moving average takes into account past trends as well as present ones. To smooth the earnings, the moving average for the years beginning in 1991 was used to forecast the median earnings between 2001 and 2012. For example, the average of median annual earnings between 1991 and 2001 was calculated to find earnings for 2001. Then, the average of median annual earnings between 1992 and 2002 was calculated to find earnings for 2002 etc. The present value of the smoothed earnings was then projected out 40 years (the anticipated working time in a human life) with a discount rate of 4.4% (equivalent to the current interest rate on a 30-year treasury bond). From this, the value of going to college could be determined. The graph “Value of College” shows the projected earnings differential between 2001 and 2012. The diagram shows that the value of college increased from 2001 to 2007 and declined since the onset of the Great Recession in 2007. Despite this, the value of college has hovered around $545,000 in the past ten years, indicating a large difference between lifetime earnings for college graduates and high school graduates.