Who benefits from student aid? The economic incidence of tax-based federal student aid

https://doi.org/10.1016/j.econedurev.2011.12.008Get rights and content

Abstract

Federal benefit programs, including federal student aid, are designed to aid targeted populations. Behavioral responses to these programs may alter the incidence of their benefits, a possibility that receives less attention in the literature compared to tax incidence. I demonstrate the importance of benefit incidence analysis by showing that the intended cost reductions of tax-based federal student aid are substantially offset by institutional price increases for a sample of 4-year colleges and universities. Contrary to the goal of policymakers, I find that tax-based aid crowds out institutional aid roughly dollar-for-dollar. Unfortunately, it is not clear how institutions utilize these captured resources, so that the ultimate incidence of the programs is uncertain.

Highlights

► I quantify the college price response to federal student aid. ► This is the first paper to consider the price response at the student level. ► Schools decrease grant aid roughly dollar-for-dollar in response to tax-based aid. ► This result holds for a sample of 4-year colleges and universities. ► It is unclear how institutions utilize the captured resources.

Introduction

Government benefit programs, including federal student aid, are designed to aid targeted populations. Behavioral responses to these programs may undermine their intended effects, and as a result, alter their designed welfare implications. Tax incidence, which explores the behavioral responses to taxation by examining where the burdens of taxation ultimately fall, receives much attention in the literature (Fullerton and Metcalf, 2002, Gruber, 1997, Kubik, 2004). The study of benefit incidence of government programs is less common. For example, until recently the assumption that the Earned Income Tax Credit benefits recipients had never been tested. Leigh (2010) and Rothstein (2010) address this omission, and find substantial erosion of benefits for nominal recipients via reduced wages in the labor market. These results suggest that the efficacy of benefit programs depends crucially on the extent of offsetting price changes. This paper adds to the benefits incidence literature by quantifying the institutional price response to tax-based federal student aid.

Funding for federal student aid, over $660 billion between 1998 and 2006,2 is based on the assumption that student aid decreases the cost of postsecondary attendance, allowing for more individuals to enroll in college and for students to graduate with less debt. The existing literature finds that student aid increases enrollment (Dynarski, 2000, Dynarski, 2003, Ellwood and Kane, 2000, Heller, 1997, Leslie and Brinkman, 1987, Nurnberg et al., 2012, Schwartz, 1985, Turner, 2011a), but how effectively these programs do so depends on the degree to which there are offsetting price changes. Yet, the literature examining the institutional price response to student aid is limited and generally focuses on tuition effects at the school level. The use of tuition increases to appropriate the benefits of federal student aid is referred to as the Bennett Hypothesis, named after former Education Secretary William Bennett.3 Long, 2006, Long, 2008 and Ikenberry (1997) discuss the existing work on the Bennett Hypothesis and note that there is weak empirical evidence supporting its validity.4

One possible explanation for these inconclusive findings is that instead of increasing tuition, schools may appropriate the benefits of federal student aid by strategically reducing institutional grant aid. I refer to this possibility as the price-discrimination Bennett Hypothesis. Unlike tuition increases that affect all students, the reduction of institutional aid allows schools to realize financial gains from increases in federal student aid while ensuring that no student is made worse off. The strategic use of institutional aid also avoids the highly visible and unpopular process of increasing tuition.5 Both policymakers and financial aid administrators are aware of the possibility that institutional aid will be replaced by tax-based aid. Former Education Secretary Richard Riley sent a letter to presidents of colleges and universities declaring that the goal of tax-based aid is to, “…provide additional help for families to pay for college and not simply substitute for existing sources of financial assistance” (Riley, 1998). In response, some financial aid administrators pledged that students would receive the full benefits of tax-based aid (Burd, 1998). However, others argued for the need to incorporate tax-based aid awards in the calculation of institutional aid. One such director noted, “…families that receive $1500 from the federal government are better off than those that don’t. And I don’t think that I can ignore that” (Burd, 1998).6

Despite the awareness that institutions may decrease aid, rather than increase tuition, in response to increases in external aid, Curs and Dar (2011), Long (2004a) and McPherson and Schapiro (1991) are the only papers that explicitly raise this possibility. They document student aid incidence at the school level and reach different conclusions on whether external aid is a substitute for institutional aid. Curs and Dar (2011) report that schools lower institutional aid in response to increases in state need-based aid. Long (2004a) reports that merit-based aid substitutes for institutional aid at private colleges in Georgia in the 1990s. In contrast, McPherson and Schapiro (1991) find that federal grant aid complements institutional aid for private schools in an earlier period for a nationally representative sample. Yet, the use of school-level data prevents these studies from determining which students are impacted by the institutional response. The flexibility of student-level data allows me to add to this work by addressing several related questions. First, do colleges and universities selectively lower institutional grant aid for students that benefit from tax-based aid? Second, how do students who experience these aid declines cope? Due to a likely time delay in benefit receipt of tax-based aid, a reduction in institutional aid may cause students to borrow more in order to offset their short-term unmet need.

In this paper, which is the first to consider the institutional response at the student level, I focus on 4-year colleges and universities. Unfortunately, I am unable to consider 2-year schools separately due to sample size considerations.7 Omitting 2-year schools is a limitation of this study, because 2-year institutions are an important component of the national postsecondary education market. To estimate student-level effects, I exploit policy-induced variation in all three tax-based aid programs, the Hope Tax Credit, the Lifetime Learning Tax Credit and the Tuition Deduction, using data from the National Center on Education Statistics. The analysis sample includes students enrolled at 190 4-year schools during the 1995–96, 1999–2000 and 2003–04 school years.8 Enrollment at the schools in the sample represents roughly 40 percent of students enrolled in 4-year colleges and universities, although schools in the sample are relatively more selective and have larger enrollments than comparable 4-year schools nationally. As a result, the findings here may not reflect the institutional response for 4-year schools generally. I estimate the impact of tax-based aid on institutional grant aid using instrumental variables to address the endogeneity of education spending. I include school fixed-effects to control for unobserved heterogeneity in student aid practices across institutions. This approach insures that identification relies on the comparison of students within a given institution. Across-school variation in institutional aid practice will confound identification to the extent that different schools offer different levels of institutional aid for the same student.

Contrary to the goal of policymakers, who sought to increase postsecondary access for eligible students by lowering the cost of enrollment, I find that the institutional price response substantially counteracts the intended cost savings of tax-based aid. Students appear to increase loans in response to the reduction of institutional aid, suggesting that tax-based aid may fall short of an important federal aid goal to reduce student indebtedness (Burgdorf & Kostka, 2006). These results imply that students eligible for tax-based aid may not be the economic beneficiaries of the programs. To determine the ultimate incidence of tax-based aid, I consider two ways in which institutions might utilize the captured resources. One, that institutions redirect aid toward students who are ineligible for tax-based aid, or two, that institutions channel the resources into other expenditures, such as capital improvements or faculty/staff salaries. Unfortunately, these results are largely uninformative so that the incidence of tax-based aid is uncertain (see Appendix A). However, I offer an important first step in establishing the incidence of tax-based aid by demonstrating that eligible students and their families are not directly benefiting from tax-based aid in the manner envisioned by policymakers. Similar unintended behaviors are found to offset the intention of policies in other contexts, including public health insurance (Cutler & Gruber, 1996) and intergovernmental grants (Baicker and Gordon, 2006, Gordon, 2004, Hines and Thaler, 1995).

The remainder of this paper proceeds as follows. Section 2 provides information on the tax-based aid. In Section 3, I discuss the institutional price response to federal student aid. Section 4 discusses the empirical specifications and results. Section 5 concludes.

Section snippets

Tax-based federal student aid

Tax-based aid provides a convenient natural experiment for examining the impact of federal aid on college pricing. Program implementation and changes in program generosity create discrete changes in aid for eligible students over time. In 1997, the Taxpayers’ Relief Act introduced the Hope Tax Credit and the Lifetime Learning Tax Credit. In 2001, The Economic Growth and Tax Relief Reconciliation Act added a third program, the Tuition Deduction. Between 1998 and 2006 these three tax-based aid

Institutional pricing behavior

While most authors agree that the standard profit maximization model does not fit colleges and universities (Clotfelter, 1999, Winston, 1999) there is not a consensus in the literature on the objective function for institutions of higher learning. Despite this uncertainty, it is possible to infer the general institutional price response to tax-based aid. Suppose that schools have an optimal input allocation in equilibrium that includes student enrollment.12

Analysis sample

To explore the institutional price response to tax-based aid, I use data from the National Postsecondary Aid Study (NPSAS) published by the National Center for Education Statistics. These data provide student-level information on financial aid, student and parent characteristics, and institutional detail. Using samples from the 1995–96, 1999–2000 and 2003–04 school years, I include roughly 74,280 undergraduate students aged 18–24 from 190 4-year schools in the primary analysis sample. To

Conclusion

I demonstrate the importance of benefit incidence analysis by showing that the intended cost reductions of tax-based federal student aid are substantially counteracted by reductions in institutional grant aid for a sample of 4-year colleges and universities. I find that students at these schools cope with this reduction of institutional support by increasing student loan amounts. Together, these findings imply that students eligible for tax-based aid are not directly benefiting from the

Acknowledgements

I am grateful to Julie Cullen, Nora Gordon and Roger Gordon for their helpful comments and for providing essential feedback. I benefited from comments from Caroline Hoxby and from other participants at the NBER Education Program Meeting. I am also thankful to Mark Showalter and two anonymous referees for numerous helpful suggestions.

References (65)

  • S. Baum et al.

    Tuition discounting, not just a private college practice

    (2006)
  • S. Baum et al.

    Trends in student aid 2007

    (2007)
  • W. Bennett

    Our greedy colleges

    (February 18, 1997)
  • E. Bettinger et al.

    The role of information and simplification in college decisions: Results from the H&R FAFSA experiment. NBER Working Paper No. 15361

    (2009)
  • S. Budack

    IPEDS finance data comparisons under the 1997 financial accounting. Standards for private, not-for-profit institutes, a concept paper

    (2000)
  • S. Burd

    Some private colleges may cut awards to students receiving. Hope scholarships

    (1998)
  • B. Burgdorf et al.

    Eliminating complexity and inconsistency in federal. Financial aid programs for higher education students, towards a more strategic approach, a national dialogue

    (2006)
  • S. Cameron et al.

    Can tuition policy combat rising wage inequality?

  • S.R. Cellini

    Crowded colleges and college crowd-out: The impact of public subsidies on the two-year college market

    American Economic Journal: Economic Policy

    (2009)
  • C. Clotfelter

    The familiar but curious economics of higher education, an introduction to a symposium

    The Journal of Economic Perspectives

    (1999)
  • J. Currie

    The take-up of social benefits

  • J. Currie et al.

    Saving babies, the efficacy and cost of recent changes in the medicaid eligibility of pregnant women

    Journal of Political Economy

    (1996)
  • Curs, B., Dar, L. (2011). Do institutions respond asymmetrically to changes in state need- and merit-based aid?...
  • D.M. Cutler et al.

    Does public insurance crowd out private insurance?

    Quarterly Journal of Economics

    (1996)
  • G. Dahl et al.

    The impact of family income on child achievement, evidence from the earned income tax credit. NBER Working Paper No. w14599

    (2008)
  • S. Dynarski

    Hope for whom? Financial aid for the middle-class and its impact on college attendance

    National Tax Journal

    (2000)
  • S. Dynarski

    Does aid matter? Measuring the effect of student aid on college attendance and completion

    American Economic Review

    (2003)
  • S. Dynarski

    The new merit aid

  • D. Ellwood et al.

    Who is getting a college education? Family background and the growing gaps in enrollment

  • J. Gruber

    The incidence of payroll taxation. Evidence from Chile

    Journal of Labor Economics

    (1997)
  • F. Hayashi

    Econometrics

    (2000)
  • D. Heller

    Student price response in higher education, an update to Leslie and Brinkman

    The Journal of Higher Education

    (1997)
  • Cited by (0)

    1

    This work was completed when I was a graduate student at UCSD. My current position is in the Office of Tax Analysis at the U.S. Treasury, 1500 Pennsylvania Avenue, NW, Washington, DC 20220, USA.

    View full text