The Effect of Tuition Subsidies on Student College Choices

• In Event: Promising Scholarship in Education: Dissertation Fellows and Their Research

Abstract

Public tuition subsidies enable students enrolled at public colleges to pay less for their education
than it costs. The institution-specic nature of tuition subsidies creates two possible controversies.
First, the presence of subsidies may distort choices between public and private colleges and even
between public colleges with dierent subsidy levels. Second, tuition subsidies may primarily go to
students whose educational choices are unaected by them. For these students, tuition subsidies
act as transfer payments from the taxpayers in their state. While all taxpayers fund these subsidies,
only those students who attend a public college benet from them. If public college students tend
to come from high-income families, tuition subsidies may be regressive.
Evaluating the validity of these concerns requires understanding whose choices are aected by
tuition subsidies and who benets from tuition subsidies. Such evaluations have been dicult due
to a lack of substantial, exogenous variation in tuition subsidies. In this paper, I avoid this challenge
by examining the dramatic decreases in tuition subsidies that occurred at California public colleges
from 2001-2010. Responding to a reduction in state funding, public colleges substantially decreased
tuition subsidies in the academic years of 2003-2004 and 2009-2010. Other public college variables
continued on their previous trends. The plausibly exogenous discrete jumps in subsidies at public
colleges contrast with the stable tuition trends at private and out-of-state colleges generating a
useful natural experiment to identify the causal eect of tuition subsidies on student choice.
I analyze the eects of tuition subsidies using administrative data on the application decisions of
over one million students and the enrollment decisions of 100,000 students. My detailed data enables
me to estimate a conditional logit model of college choice that can incorporate heterogeneity in
student response to tuition subsidies. This is critical because many important questions surrounding
tuition subsidies are about dierences in student behavior. I nd that the average eect of a $1000
decrease in tuition subsidies is a 3% drop in the probability of enrollment. This eect varies by
student characteristics, in particular student income.
Using my estimates, I conduct a counterfactual analysis that indicates that 6% of high-aptitude,
low-income students change their enrollment behavior when subsidies fall. Instead of enrolling at
University of California campuses, about a third of these students choose private colleges and the
majority of the rest choose two-year colleges. Low-aptitude students tend to move from California
State University campuses to two-year colleges though some with high-incomes attend private
colleges. These student shifts change the mix of students at public colleges and tend to make
tuition subsidies more regressive. Despite the large decreases in subsidies, the majority of students
(about 97%) do not change their enrollment choice indicating that the subsidies mostly serve as a
government transfer. The main eect of the tuition subsidy reductions is to redistribute billions of
dollars from public college students to other Californians.

Author

• Brianna Cardiff-Hicks, Stanford University