Research

Working Papers

"Competing Campuses: An Equilibrium Model of the U.S. Higher Education Market" (Job Market Paper) Link

Abstract: Universities compete for enrollment through price and non-price characteristics, including the number and type of undergraduate majors offered. Undergraduate majors are a key source of differentiation in U.S. higher education, and survey evidence suggests that students consider a university’s offered majors when they make their college choice. This research seeks an empirical answer to three questions. First, do the number and type of majors offered at a university affect applications and enrollment? Second, what are the preferences and costs that drive universities’ choices? Finally, what are the equilibrium effects of subsidies in the form of increased financial aid or subsidies awarded directly to universities to support Engineering, Math, and Science programs? To answer these questions, I use state-level data from ACT and College Board and information on university characteristics to estimate an equilibrium model of the U.S. higher education market.

The model features four stages: first, universities choose majors, admission criteria and price; second, high-school seniors observe universities’ choices and choose a set of colleges to apply to; third, universities offer admission to a subset of applicants; and finally, students make their enrollment decisions. Modeling the application, admission, and enrollment stages of the college choice process produces parameters that have a meaningful interpretation in an expectation-maximization framework. The model of university behavior features a trade-off between selectivity and revenue net of costs, with heterogeneous preferences for selectivity.

I use the simulated method of moments to estimate the model. Instrumental variables are utilized in estimation for both the demand- and supply-sides. On the demand side, the model implies correlation between the endogenous characteristics—price and majors—and unobserved quality. On the supply-side, the model implies correlation between universities’ heterogeneous costs and the endogenous characteristics. Instruments include the exogenous demand- and supply-shifters, graduate programs offered by universities, and features of the regional market structure. The demand estimates show that students are willing to pay over $100 per year for each major offered, with heterogeneity by type. Supply-side estimates show that universities place substantial weight on selectivity, and that the cost of supplying different majors varies by type. 
Financial aid awarded to students to attend public universities in a state results in increased price and selectivity stratification in the public sector, higher prices and increased selectivity in segments of the private sector, and adjustments to majors offered to manage costs. A $10 per student subsidy to public universities for each Engineering, Math, and Science major offered would increase the median number of majors in that category at public universities from 8 to 12, while also generating increased stratification by price and admissions selectivity in the public sector, and higher prices and admissions selectivity in segments of the private sector.

Published and Forthcoming

"Missed Exams and Lost Opportunities: Who Could Gain from Expanded College Admission Testing?," with Sarah Turner (UVA), AERA Open, 2019, Volume 5, Issue 2, pp 1-20. https://doi.org/10.1177%2F2332858419855030

Abstract: When students with the capacity to succeed in a four-year college do not take a college admission test, this represents a potential loss of opportunity for students and colleges alike.  However, the costs of testing—both pecuniary and non-pecuniary—may exceed the benefits for students who lack the interest in or qualifications for college attendance.  In states like Virginia, access to admission tests varies markedly with district and family circumstances.  We estimate that universal testing in Virginia could increase the number of high school graduates with test scores competitive for admission at broad-access universities in the state by as much as 40%—and at the most selective institutions, nearly 20%—with larger increases for low-income students.  Alternative policies that encourage testing among students with high demonstrated academic performance could realize nearly these increases without generating testing costs for students who are unlikely to attend a four-year college.

"Network Structure and Consolidation in the U.S. Airline Industry, 1990-2015," with Federico Ciliberto (UVA) and Jonathan W. Williams (UNC-Chapel Hill), The Review of Industrial Organization, February 2019, Volume 54, Issue 1, pp 3–36.  https://doi.org/10.1007/s11151-018-9635-y

Abstract: We study the effect of consolidation on airline network connectivity using three measures of centrality from graph theory: Degree; Closeness; and Betweenness. Changes in these measures from 1990 to 2015 imply: (i) the average airport services a greater proportion of possible routes; (ii) the average origin airport is fewer stops away from any given destination; and (iii) the average hub is less often along the shortest route between two other airports. Yet, we find the trend toward greater connectivity in the national network structure is largely unaffected by consolidation—in the form of mergers and codeshare agreements—during this period.

"Transitioning from High School to College: Differences across Virginia," with Jessie Romero (Richmond Fed) and Sarah Turner (UVA). Federal Reserve Bank of Richmond Economic Brief, 2017, 17-12. https://www.richmondfed.org/publications/research/economic_brief/2017/eb_17-12

Abstract: In Virginia, there are substantial differences across school districts in college enrollment and, conditional on college enrollment, attendance at high resource colleges and universities. School districts in low-income and relatively rural areas tend to demonstrate the weakest outcomes, but income and geography do not fully account for the observed differences. Whether limited enrollment at a broad range of colleges arises from gaps in academic preparation, difficulty in navigating the application process, or individual preference matters greatly for public policy.

Work in Progress

"Pricing Public Higher Education: Responses to Appropriations Changes in a Market Context", with Sarah Turner (UVA)


Abstract: As public universities have experienced an overall retrenchment in support from state governments in recent decades, tuition levels have also increased markedly. This analysis examines the question of how the posted tuition and the net price faced by students from different economic circumstances adjusts to changes in public subsidies or state appropriations. Our interest is in modeling the link between changes in state appropriations and prices for different types of students across the distribution of public colleges and universities. Within states, public colleges and universities differ substantially in how they adjust tuition levels and associated net prices to changes in appropriations. Public research universities, which are generally the most selective and most resource-intensive within a state, typically demonstrate greater changes in response to appropriations declines than their broad access counterparts. In turn, these public research universities tend to distribute tuition increases progressively across income bins with net price for low-income students increasing less than dollar-for-dollar with posted tuition increases. In contrast, net price changes are closer to dollar-for-dollar across all income groups at broad access institutions. Distinguishing further among state markets and institutions, universities in states with a relatively high representation of high-income families (who are likely to be price inelastic) and those with fewer private sector competitors are the most likely to move to high-tuition, high-need pricing strategies.

"Competition, Wages, and the Emergence of Computer Science Degree Programs in the U.S.," with Devaki Ghose (UVA) and Ekaterina Khmelnitskaya (UVA)

Abstract: How does the presence of endogenous entry costs affect entry decisions? When deciding whether to introduce a new program universities have to take into account that a large part of their costs, both variable and fixed, depends on the faculty wage. This wage, at the same time, is determined in equilibrium, where it is affected not only by demand from other universities but also by the demand from the industry. The equilibrium faculty wage, in turn, will be endogenous to the entry decisions of all the participants and the industry demand. We introduce a two stage game where in the first stage universities decide whether to enter a market while taking into account the costs that will be determined in the second stage through equilibrium faculty wages. We thus develop a framework that allows us to study not only how the standard entry game changes when the costs are endogenized, but also how the competition between the downstream and upstream industry for the same input affects entry. We use historic data on CS program adoption at US universities, university characteristics, faculty wages, and industry wages to understand the timing of each university's decision to introduce CS programs at the undergraduate and graduate levels. Implications for public policy are explored- in particular, we describe how the amount and timing of subsidies to universities affected the growth of CS as a discipline and the distribution of CS programs by university control and selectivity.