Carter Davis and Taylor Nadauld, Finance Department
The original project that I set out to research with this ORCA grant was to investigate how the real estate bubble spread leading up to the financial crisis, using an incredible 90 gigabyte dataset of mortgage information. Two unfortunate events occurred during the course of this project: the external hard drive that stored the data was lost and a programming error on my part destroyed the data on the BYU supercomputer. Dr. Nadauld and I were devastated. Instead of investigating how bubbles spread, I set out to estimate the slope of the supply curve of higher education. Dr. Nadauld and I have been working for the past year to analyze how federal loan and grant policies affect the cost of higher education. Estimating the slope of the supply curve of higher education is both relevant to our current research and has important policy implications. I did this project with three BYU students: Joseph Cooprider, Michael Gmeiner, and Nick Hales.
The rising cost of higher education is of concern to prospective students, colleges and universities, and employers in the United States. The National Center for Education Statistics reports that from 1981 to 2012 the average total tuition, fees, and room and board rates for full-time undergraduate students at 4-year institutions has risen from $9,544 to $23,066 (measured in 2011-2012 dollars). In an article published on November 20, 2014 on the New York Times website, David Leonhardt reported that a survey of 1,000 individuals conducted by NBC News and the Wall Street Journal found that, “(of) a long list of domestic and foreign policy proposals, none received more support –82 percent— than reducing the cost of student loans.”
In basic economic theory, the slope of the supply curve indicates how demand shocks affect a particular market. A demand shock is when the quantity demanded of a particular good increases or decreases due to factors other than price. Intuitively, the slope of the supply curve indicates the degree to which prices rise (fall) and how much the quantity of goods sold in a market increases (decreases) when demand increases (decreases).
In the context of higher education, understanding the slope of the supply curve is important to understand what occurs to tuition and the amount of university degrees offered when demand shocks occur. For instance, during the financial crisis, the amount of people seeking university degrees vastly increased. This is a positive demand shock. The slope of the higher education supply curve indicates how much tuition changed and how the amount of university degrees changed due to this demand shock.
Estimating the slope of a demand curve or a supply curve is often quite challenging. This is because the only events that economists actually observe in the market are the equilibrium prices and quantities, and not actually the demand and supply pressures that are driving the prices and quantities. For example, the price of oil in the world has fallen by almost 50% recently because of both increasing supply and falling demand. It is quite difficult to determine how much the price change is actually due to demand or supply pressure though, because these actual pressures are not observed in the market. To overcome this difficulty, the standard way to estimate a supply curve is to find a factor that moves demand but not supply and statistically “trace out” the slope of the supply curve using this factor. This factor is referred to as the exogenous variable.
We used the percent of basketball wins from a previous season as the exogenous variable. Specifically, we thought that a successful basketball season may attract additional applicants to a university, but not affect the amount of degrees a university offers in a given year. To conduct this analysis, we meticulously matched basketball data, tuition data, and enrollment data that was publicly available. This took an incredible amount of time, and the data is available upon request (carterdavis@byu.edu).
The table below gives the estimated relationship between basketball wins and enrollment at universities in the United States. These results are statistically significant, indicating that when a university has an excellent basketball season, enrollment increases at the university the next year significantly.
We use these estimated results to trace out the supply curve, as discussed above. We found that the slope of the supply curve was not significantly different from zero. This indicates that we likely have a “weak” exogenous variable. In other words, our basketball wins treatment effect is not powerful enough to accurately identify the supply curve. It could also be the case that the supply curve is actually just flat, which is less plausible.
Note: Each regression includes institution fixed effects and clustering at the institution level. Applicants and enrollment are in log terms.