Welcome to the Missouri Department of Higher Education Website (MDHE)

Access and Affordability

Access and Affordability
Patterns of Financial Aid and Student Performance for
a Cohort of Missouri College Freshmen

June 2004

Michael Podgursky, Middlebush Professor of Economics and Chairman, Department of Economics, University of Missouri - Columbia
Debra Cheshier, Director, Educational Policy, Planning, and Improvement Center,Missouri Department of Higher Education
John Wittstruck, Senior Research Associate, Educational Policy, Planning, and Improvement Center,Missouri Department of Higher Education
Don Watson, Senior Research Associate, Department of Economics, University of Missouri - Columbia
Ryan Monroe, Research Associate, Department of Economics, University of Missouri - Columbia

Funded by a grant (Grant No. 999) from the Lumina Foundation for Education, Indianapolis,Indiana

Foreword

In recent years, several national and state reports have been released discussing the role and impact of student financial aid as it relates to the affordability of higher education and student participation and performance. These reports have shaped state- and institutional-level conversations about public policies, funding levels for institutional operations and student financial aid, and their effect on the participation and performance of students, particularly those who are first-generation, low-income, minority, and/or working adults.

Some of the reports that have shaped institutional- and state-level discussions about student financial aid includeMissed Opportunities - A New Look at Disadvantaged College Aspirants, released by The Institute for Higher Education Policy in 1997, and Straight Talk About College Costs and Prices, released in 1998 by the National Commission on the Cost of Higher Education.

Other notable reports include the Missouri Coordinating Board for Higher Education's reportof the Missouri Commission on the Affordability of Higher Education,Toward an Affordable Future (1999), that reported:

There has been a dramatic shift in responsibility for financing higher education to students and their families, and this additional burden is increasingly being financed by student loans.
College and university aspirations and public expectations of these institutions have a significant impact on the cost of higher education.
Missourians have access to a diverse system of public and independent higher education with institutions that offer a wide range of choices in terms of tuition and fee levels.
State-level policy makers and the public have a limited understanding of institutional pricing policies and the reasons for the increasing costs of higher education.
No student pays the full cost of a college education. Public and independent colleges and universities subsidize the cost of higher education with different sources of revenue, and the subsidy varies by institution, but all students receive substantial subsidies.

The 22 recommendations offered by the Missouri Commission on the Affordability of Higher Education addressed issues of cost containment, state budget policy, pricing, financial access, and consumer information. Among the financial access recommendations, the commission recommended that the state coordinate student financial aid programs to increase state funding for need-based financial aid for students pursuing the first two years of postsecondary education. The guiding principles for establishing a coordinated, need-based student financial aid program suggested by the commission resulted in the following recommendations:

Financial aid should follow Missouri students;
Eligible institutions should include approved two- and four-year public and independent colleges and universities, vocational, and private career schools;
Aid awards should be at least $2,250 initially, and future increases should be linked to an appropriate index;
An additional $1,000 could be awarded if matched by institutional funds;
Aid should be applied to part-time students on a proportional basis;
Aid programs should address the needs of all students; and
The design of aid programs should minimize the amount of student debt accumulated in the first two years of postsecondary education.

The Lumina Foundation for Education has released several reports that have had an impact on state policy development regarding student financial aid programs and funding levels. Some of these reports includeDesigning a State Student Grant Program: A Framework for Policy Makers, prepared by Jerry Sheehan Davis, Vice President for Research, in September 2001. The Lumina Foundation for Education's January 2002 report,Unequal Opportunity - Disparities in College Access Among the 50 States by Samuel M. Kipp III, Derek V. Price, and Jill K. Wohlford, was the first national report to identify colleges and universities as being admissible and/or affordable with or without student loans.

Missouri is a participating state in the Lumina Foundation-funded Western Interstate Commission on Higher Education (WICHE) project, "Changing Direction: Integrating Higher Education Financial Aid and Financing Policy." The project, also sponsored by the State Higher Education Executive Officers (SHEEO) and the American Council on Education (ACE), provides a forum in which states exchange research, strategies, best practices, and ideas for integrating student financial aid policy and policies for financing institutional operations.

Patrick Callan of The National Center for Public Policy and Higher Education, throughMeasuring Up 2000 andMeasuring Up 2002, the state-by-state report cards, has focused attention on improving the preparation, participation, affordability, completions, benefits, and learning of students. Finally, academic scholars such as Tom Mortensen of Postsecondary Education Opportunity, Donald Heller of Pennsylvania State University, Edward P. St. John of the Indiana Education Policy Center at Indiana University, and others, continue to provide evidence through their research that institutional, state, and federal policies and funding levels for student financial aid are insufficient to meet the financial needs of the nation's postsecondary education students. This is especially true for first-generation, low-income, minority, and working adult students. Through this Lumina Foundation for Education-funded research, Missouri is contributing to the body of research and literature about the role of student financial aid in helping students complete their postsecondary education objectives.

Table of Contents

List of Tables

Table 1: Characteristics of Missouri Public Four-Year Participating Institutions

Table 2: Comparison of Institutional Financial Aid for Public Higher Education Students in 1995-96 National Postsecondary Student Aid Study (NPSAS) and Institutions Participating in the Longitudinal Study

Table 3: Average Need Based and Non-Need Based Financial Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

:Table 3A: Distribution of Need Based and Non-Need Based Financial Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

:Table 4: Average Federal, State, and Institutional Need-Based and Non-Need-Based Gift Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

:Table 4A: Distribution of Federal, State, and Institutional Need Based and Non-Need Based Gift Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table 5: Average Need-Based and Non-Need Based Financial Aid by Race in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table 5A: Distribution of Need-Based and Non-Need Based Financial Aid by Race in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table 6: Average Need-Based and Non-Need Based Financial Aid by ACT Composite Score in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table 7: Distribution of Need-Based and Non-Need-Based Financial Aid by ACT Composite Score in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table 8: Changes in Institutional and Non-Institutional Gift Aid from Year 1 (1997-1998) to Subsequent Years: First-Time Freshmen Enrolled in Participating Institutions

Table 9: Four- and Six-Year Graduation Rates for Students Receiving Different Types of Financial Aid: First-Time Freshmen Enrolled in Participating Institutions

Table 10: Student Loan Debt for Students Graduating Within Six Years by Family Income, Race, and Gender: First-Time Freshmen Enrolled in Participating Institutions

Table 11: Student Loan Debt for Students Graduating or Not Graduating Within Six Years by Accumulated Credit Hours for the Non-Graduates: First-Time Freshmen Enrolled in Participating Institutions

Table 12: 2002-03 Missouri Freshman FAFSA Applicants by Date of Application

Table 13: 2002-2003 Missouri First-Time Freshman FAFSA Applicants by Gender and Age

Table 14: 2002-2003 Missouri Freshman FAFSA Applicants by Education of Parent

Table 15: 2002-2003 Missouri Freshmen FAFSA Applicants and All Taxpayers by Family Adjusted Gross Income

Table 16: 2002-2003 Missouri Freshmen FAFSA Applicants by Expected Family Contribution

Table 17: Mean and Median Estimated Family Contribution (EFC) for 2002-2003 Missouri Freshmen FAFSA Applicants by Family Income

List of Figures

Figure 1: In-State Undergraduate Tuition and Fees, Based on 30 Credit Hours per Year, for Participating Institutions

Figure 2: In-State Undergraduate Full-Time Equivalent Enrollment for Participating Institutions

Figure 3: Average Need-Based and Non-Need Based Financial Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Figure 4: Average Federal, State, and Institutional Need Based and Non-Need Based Gift Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Figure 5: Average Need-Based and Non-Need-Based Financial Aid by Race in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Figure 6: Average Federal, State, and Institutional Need Based and Non-Need Based Gift Aid by ACT Composite Score in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Figure 7: Changes in Total Gift Aid from Year 1 (1997-1998) to Year 2 (1998-1999), Year 3 (1999-2000), and Year 4 (2000-2001): First-Time Freshmen Enrolled in Participating Institutions

Figure 8: Changes in Institutional and Non-Institutional Gift Aid from Year 1 (1997-1998) to Year 2 (1998-1999): First-Time Freshmen Enrolled in Participating Institutions

Figure 9: Changes in Institutional and Non-Institutional Gift Aid from Year 1 (1997-1998) to Year 4 (2000-2001): First-Time Freshmen Enrolled in Participating Institutions

Figure 10: Estimated Effect of Increasing Financial Aid from No Aid to the Average Award on Four- and Six-Year Graduation Rates: First-Time Freshmen Enrolled in Participating Institutions

Figure 11: Estimated Effect of an Additional Dollar of Financial Aid on School Year Labor Market Earnings in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Figure 12: Missouri 2003:3 Earnings for Graduates and Non-Graduates by Type of Financial Aid: First-Time Freshmen Enrolled in Participating Institutions

Figure 13: Populations Analyzed in this Report: All 2002-2003 Undergraduate Missouri FAFSA-Filers and Freshman Enrolled in Missouri Public Higher Education Institutions

Figure 14: 2002-2003 Missouri Freshmen FAFSA Applicants and All Missouri Taxpayers by Adjusted Gross Income

Figure 15: K-12 and Higher Education Enrollment and FAFSA Submissions by Race: Freshmen Who Graduated from a Missouri High School and Enrolled in Missouri Public Two- or Four-Year Institution, Academic Year 2002-2003

Figure 16: FAFSA Submission Rates: Percent of College Freshmen Who Filled Out a FAFSA, Academic Year 2002-2003

Figure 17: First-Time Freshmen FAFSA Completion Rates by Institution: Missouri Public Two- and Four-Year Institutions

Figure 18: Simple and Regression-Adjusted Measures of the Difference in FAFSA Submission Rates between White and Minority Students

List of Appendix Tables

Table A1: Federal, State, and Institutional Financial Aid Collected from Participating Institutions: Definitions for Gift, Loan, and Work Aid and Need and Non-Need Aid

Table A1.1 Institutional student financial aid data file layout used by participating institutions to report data to the Missouri Department of Higher Education for this research project

Table A1.2 Instructions and definitions for student financial aid awarded and reported on the MDHE -14 survey

Table A1.3 Glossary of terms that may be useful when completing the DHE-14 survey

Table A2: Combinations of Financial Aid Packages in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table B1: In-State Undergraduate Full-Time Equivalent Enrollment for Participating Institutions

Table B2: Index of In-State Undergraduate Full-Time Equivalent Enrollment for Participating Institutions

Table B3: In-State Undergraduate Tuition and Fees, Based on 30 Credit Hours per Year, for Participating Institutions

Table B4: Index of In-State Undergraduate Tuition and Fees, Based on 30 Credit Hours per Year, for Participating Institutions

Table B5: Distribution of Family Income from the ACT for Students Who Filled Out the FAFSA and Those Who Did Not: First-Time Freshmen Enrolled in Participating Institutions

Table C1: Frequency Distributions and Four-, Five-, and Six-Year Graduation Rates by Student Characteristics: First-Time Freshmen Enrolled in Participating Institutions

Table C2: Linear Probability Model of Four- and Six-Year Graduation: First-Time Freshmen Enrolled in Participating Institutions

Table C3: A Comparison Between First-Time Freshmen Enrolled in Participating Institutions and First-Time Freshmen Enrolled in Non-Participating Public Four-Year Institutions in Missouri

Table D1: Regression Results for Estimated Effect of an Additional Dollar of Financial Aid on School Year Labor Market Earnings in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Table E1: Who Fills Out a FAFSA? Linear Probability Model Estimates: Dependent Variable = 1 if student completed a FAFSA, 0 otherwise (absolute value of t-statistic in parenthesis)

Summary of Findings and Policy Implications

Summary of Findings

Research on student financial aid and performance has relied heavily on national survey data. However, there is a great deal of variation between states in the financing of higher education and in student financial aid programs. This makes national averages less useful for state education policy-making. Moreover, state-level samples drawn from national surveys are typically too small to reliably draw inferences. This makes it important to develop research strategies and databases that make better use of state and institutional administrative data for financial aid research. This project makes a contribution to that agenda by developing student-level research data files from a variety of state administrative data in Missouri. The following sections summarize our findings.

I. Understanding Institutional Financial Aid

Institutional financial aid data are typically not available to state higher education agencies. For this research, student-level financial aid data were collected directly from higher education institutions and merged with various state higher education agency files. The resulting database provides a complete longitudinal record of aid and performance data for a very large cohort of Missouri freshmen, including federal, state, and institutional financial aid data, unemployment insurance (UI) wage records, and detailed data on student participation and performance at six public four-year higher education institutions in Missouri. These six institutions account for 48 percent of public four-year higher education enrollments in the state. These data were used to analyze patterns of financial aid to students and the effect of aid packages on student performance and the likelihood of graduation. Major findings are:

Institutional financial aid is a very important source of support for low-income students in Missouri. Institutional aid made up 44 percent of total gift aid for students with family incomes less than $25,000. The vast majority of this institutional aid was not need-based.
Gift aid is strongly and positively associated with student ACT composite scores. The positive relationship is due largely to a state merit scholarship program tied to ACT performance and institutional gift aid.
Students who receive gift aid in their freshman year are likely to receive gift aid in subsequent years. However, the size of awards varies considerably from year-to-year: for some students, gift aid falls and for others it increases. On balance, more students lose than gain gift aid over time.
The major source of year-to-year change in gift aid is institutional. Students are much more likely to experience a decline in institutional gift aid between their freshman and subsequent years than they are to experience a gain
Students who receive gift aid are more likely to graduate in four or six years than students who do not receive gift aid. The same is true for work aid. However, this relationship does not hold for loan aid.
One mechanism by which aid may raise graduation rates is by reducing a student's need to work during the school year. Students who receive gift aid have lower labor market earnings during the academic year.
Most students graduate with debt. Sixty percent of all students who graduated within six years had loans, a proportion that rises to 82 percent for African-American graduates. Average debt for all graduates was $13,633, compared to $18,162 for African-American graduates. Part of this gap can be explained by family income differences. However, even controlling for family income, African-American graduates have higher levels of debt.

II. Who Fills Out a FAFSA?

The first step in securing financial aid for most college students involves completing a Free Application for Federal Student Aid (FAFSA) form. Data from these forms provides a measure of the demand for financial aid and provides some insight as to the characteristics of low-income students who desire to attend a higher education institution. In order to better understand the demand for financial aid and the extent to which students, particularly minority students, are making use of existing avenues of financial support, the researchers analyzed data on 53,807 Missouri FAFSA completers who planned to enroll as first-time freshmen in a Missouri public two- or four-year higher education institution in the fall of 2002. The findings for these first-time freshmen are:

Roughly two-thirds of Missouri high school graduates fill out a FAFSA
The vast majority of freshman FAFSA applicants are under 25 years in age (80 percent); however, seven percent are 35 or older.
Females account for 59 percent of freshmen FAFSA applications.
First-generation college students account for the majority (51.5 percent) of FAFSA applicants.
FAFSA applicants come from households with higher incomes than the typical Missouri household. In Fall 1997, 49 percent of FAFSA applicants came from families with Adjusted Gross Incomes (AGI) of $35,000 or less. However, in 1997, 59 percent of all Missouri households reported AGIs of $35,000 or less.
Most freshman in Missouri public higher education institutions fill out a FAFSA: 75.9 percent of first-time freshmen in public four-year institutions and 66.0 percent of first-time freshmen in public two-year institutions complete a FAFSA.
Among enrolled freshman, African-American students are much more likely than white students to fill out a FAFSA. In public four year institutions, 93 percent of African-American versus 75 percent of white students fill out FAFSA's.

Policy Implications

The findings of this research have implications for a number of public policy issues related to the distribution of institutional, state, and federal student financial aid. Some of these are specific to the state of Missouri; however, many undoubtedly apply to other states' and to the federal student financial aid programs.

While institutional, state, and national reports are replete with discussions of concern about the cost and price of higher education, it is noteworthy that this study found that 23.5 percent of the freshmen enrolling in fall 1997 received no financial aid during their freshman year (Table A2). This suggests that the issue of cost and price of higher education, while important, may not affect all incoming freshmen.

For those freshmen that do receive aid, the number of different combinations of student financial aid make it difficult for students and parents to know how the mix of student financial aid programs can help them finance a college education (Table A2). Institutional, state, and federal student financial aid policies and programs need to be integrated and streamlined if the distribution of financial aid is to become more efficient and understandable to students and their families, and is to be an effective strategy in promoting both access to and successful participation in higher education.
Institutional gift aid is important to how students are able to finance their college education. Institutional gift aid, however, is invisible to many - including policy makers. The current lack of publicly available information on institutional financial aid is especially problematic for the more than 50 percent of students who reported themselves as first generation college students (Table 14). Increasing funding for state-wide need-based programs versus institutional programs may provide first-generation students with a better understanding of financial aid that is available to them.
Arbitrary cut off dates for applying for financial aid and submitting the FAFSA need to be examined. For example, in Missouri only 44 percent of the FAFSA filers considered for state need-based student financial aid complete and submit the FAFSA by the March 31 state cut-off date for state student financial aid (Table 12). This cut-off date should be reconsidered in light of the 56 percent of students completing and submitting a FAFSA after that cut-off date.
Nearly 58 percent of those completing a FAFSA report adjusted gross incomes (AGI) of $50,000 or less compared to 72 percent of state residents reporting adjusted gross incomes of $50,000 or less (Table 15). This suggests that there may be additional middle and low-income residents who could file a FASFA and benefit from further education. Early awareness and outreach efforts targeted to middle and low income students could increase the number and proportion of these students completing a FAFSA and potentially enrolling in one of the state's higher education institutions.
Slightly more than 18 percent of first-time freshman FAFSA filers are over the age of 25 (Table 13). This suggests college attendance decisions are not the exclusive domain of students enrolling directly out of high school. Older adults enrolling as freshman have different challenges and needs than recent high school graduates, and place different demands on institutions and states for course delivery, program offerings, and access to student financial aid. State public policies and institutional practices need to focus more attention on the adult student, particularly when nearly one out of five new freshmen is over the age of 25.
The largest proportion of freshmen (12 percent) receiving aid are awarded institutional non-need based gift aid. Gift aid is important to student success as demonstrated by its relationship to increased graduation rates and reduced time to degree completion for those students receiving gift aid (Figure 10). More freshmen lose their gift aid during their subsequent academic careers than those who receive more gift aid (Table 8). Some of this loss is a result of the academic performance criteria institutions require students to maintain to retain their non-need based gift aid. These institutional policies should be examined to ensure they are in the best interests of the student, the institution, and the state.
Debt accumulated from student loans is of significant concern as noted in various studies and reported in the mass media. Sixty percent of baccalaureate degree recipients form the entering freshman class of fall 1997 graduated with an average of $13,633 in loan debt (Table 10). In addition, many students who left school without a degree but were near degree completion (90 to 120 credit hours) had accumulated $10,811 of debt (Table 11). Institutional retention strategies and practices as well as retention and student information strategies promulgated by the state need to focus attention on the importance of degree completion and on students who borrow to finance their higher education. Institutions, the state, and the federal government need to work together to ensure that students are sufficiently financially literate to understand how the accumulation of so much debt, particularly without completing a degree, can lead to financial hardships in the future for these students.
Race and income differences have well documented influences on student outcomes. While a higher proportion of African-American students complete a FAFSA than white students (Figure 18), African-American students accumulate more loan debt upon graduation than white students - $18,162 for African-Americans compared to $13,046 for white students (Table 10). African-American students receive 14 percent of the gift aid disbursed by institutions while comprising only 7.5 percent of student enrollments. White students receive nearly 79 percent of all gift aid but comprise slightly more than 86 percent of enrolled students.

Low-income students (incomes less than $25,000) comprised just over 19 percent of the fall 1997 freshman class but received 29.5 percent of all gift aid (Table 3A) These low-income students also accumulated an average debt of $17,452 by the time they graduated (Table 10). Although middle income students (incomes of $25,000 but less than $75,000) comprised slightly more than 37 percent of the fall 1997 freshmen and received nearly 42 percent of all gift aid distributed to fall 1997 freshmen (Table 3A) had accumulated debt of $18,585 by the time they graduated.

These findings suggest that access to and success in higher education is not an issue of race as much as it is an economic issue. Institutional, state, and federal student financial aid policies need to be integrated and connected to ensure that both low- and middle-income students have access to non-loan grants. Throughout the findings of this study, it is clear that middle-income students face many of the same challenges in financing their higher education as low-income students do.
Unfortunately, this study is not able to say anything about students who do not complete a FAFSA or enroll in a public college or university. Joint research initiatives by states departments of elementary and secondary education and higher education are needed to help each learn more about who are the students that, for what ever reason, do not go on to some further education and training after high school. Missouri is essentially a no growth state with enrollments in higher education not increasing as they have and are project to in other states. For example, the participating schools in this study only increased enrollments by 14 percent between fall 1997 and fall 2003 (Tables B1 and B2). If the state's system of higher education is to increase the level of educational attainment of Missouri citizens, the state must put in place policies and strategies to promote increased participation in higher education.
This study and the research presented in this report demonstrate limited causal relationships between the impact of various forms of student financial aid on access to and successful participation in higher education and selected student outcomes. It is essential, however, that this kind of research be continued and broadened to include more types of institutions within Missouri's system of higher education to more fully explain how student financial aid impacts student participation and the outcomes of higher education. Only through research involving institutional, state, and federal financial aid data can institutions, the state, and the federal government learn about where and how financial aid policies may be working at cross purposes, and how student aid programs might be better integrated and aligned to best meet the financial needs of students. Such research is vital to knowing the role student financial aid has in financing higher education to ensure that financing policies related to state support for higher education, institutional tuition rates, and financial aid are better aligned to ensure that all students have financial access to postsecondary education.

Section I: Introduction

Since the early 1970s, college prices have increased faster than inflation and during this same period the average income for families in the lowest income quintile showed a slight decline. These two trends have created considerable concern among families about paying for a college education. Adding to these concerns is the reduced purchasing power of the federal Pell grant. In 1979-1980, the maximum Pell grant covered 77 percent of the average tuition costs, fees, and on-campus room and board at a public four-year college. By 2002-2003, the maximum grant covered only 41 percent of these prices (King, 2003). From 1975 to 1985, federal funding for higher education decreased by 27 percent while from 1985 to 2000, it increased by 21 percent (National Center for Education Statistics, 2000a). In addition, both public and private institutions have experienced a decrease in the percentage of their current funding from government sources, and as a result have come to rely increasingly on tuition revenue to compensate for the decline in funding. The dynamics of college costs, higher education funding at both the state and federal level, and declining family income are creating an increasing challenge for individuals to continue their education beyond high school.

What effects have these changes had on patterns of financial aid and degree completion? Research on student financial aid and performance has relied heavily on national survey data. But the long time lags between the collection and release of these national data limit their usefulness to current policy debates. While state administrative data are timelier and offer many advantages over national data, most states' administrative data lack institutional financial aid data. In order to address this limitation, the researchers constructed a longitudinal student data file from federal, state, and institutional financial aid data, unemployment insurance wage records, and detailed data on student higher education performance at six public four-year higher education institutions in Missouri. These six institutions account for 48 percent of public four-year higher education enrollments in the state. These data were used to analyze patterns of financial aid to students and the effect of aid packages on student performance, and they provide a very useful complement to research based on national survey data.

Literature Review

There is a large empirical literature on the effects of tuition and financial aid on student enrollment, academic progress, and degree completion. One strand of this literature has examined the effect of tuition on student enrollment. Often cited surveys of this literature are Leslie and Brinkman (1987) and Heller (1997). Not surprisingly, researchers have generally found that the demand curve for higher education has a negative slope - as price goes up, enrollment or applications go down. There is some variation among studies in the estimated price response and most of the surveyed literature is based on grouped data. These studies generally find that the demand curve is fairly inelastic, e.g., a 10 percent increase in tuition reduces enrollment by less than 10 percent.

More relevant for this report is whether there are differences between high- and low-income students in enrollment or completion in response to tuition increases. Other things being equal, one might expect a different response from low-income families. Even if the payoff to higher education is the same for students from high- and low-income families, the latter may lack the ability to finance investments in higher education. Thus, human capital theorists would describe these low-income students as "credit constrained." Heller (1997) concludes from his survey of the literature that the tuition responsiveness of low-income students is indeed greater. He also finds that student financial aid generally raises the likelihood of college attendance for low-income students. More recent work by Cameron and Heckman (2001) finds less evidence of credit constraints. The problem, according to these researchers, is that high income families make substantial investments in the education of their children. As a result of these investments, children from higher income families tend to have cognitive skills and affective behaviors that make them more "college ready." The smaller investment low income families make over the lifetime of their children may be a factor contributing to low participation in college by low income students. Cameron and Heckman find little evidence that "credit constraints" can account for racial and ethnic gaps in college attendance.

On the question of the distribution and effect of institutional financial aid, the research literature is much slimmer, largely because detailed institutional financial aid data are not available in most national longitudinal student data files (e.g., High School and Beyond, National Longitudinal Study of the Class of Year 1972, National Educational Longitudinal Study of 1988), or in state administrative data. Further, variables that might affect the tendency to enroll or succeed in college (e.g., cognitive skills, motivation) may be correlated with the receipt of financial aid. In some studies the controls for student ability are limited or absent. In principle, "random assignment" experimental studies might be conducted to obtain an estimate of the effect of aid. For example, where an applicant pool of equally worthy and talented students exceeds the available scholarship funds, it would be possible, hypothetically at least, to conduct a lottery to distribute the aid and analyze the behavior of recipients and non-recipients. However, perhaps due largely to ethical concerns, no random assignment studies of financial aid were found. This has led some economists to search for "natural experiments" in which administrative quirks or other environmental phenomena lead some subgroups of approximately equally able and worthy students to get more financial aid than others.¹ The findings in this newer econometric literature with respect to financial aid are mixed (Linsenmeier, Rosen, and Rouse, 2002; Kane, 2003; Bettinger, 2004).

A recent paper by Turner (2004) highlights the divergence between higher education attendance and degree completion. Unfortunately, many discussions of "access" fail to note this very important distinction. Turner shows that while the probability of higher education attendance has increased substantially since the 1970s, the probability of baccalaureate degree completion has increased at a much slower rate. For example, in 1970, among high school graduates aged 23, 23 percent had completed a baccalaureate degree and 51 percent attended college for some period of time following graduation. By 1999, the college attendance rate had grown substantially to 67 percent while the percent completing baccalaureate degrees only grew to 24 percent. More students do, in fact, earn baccalaureate degrees beyond age 23, and the age at degree completion is rising. Turner's work highlights the importance of analyzing the effect of student financial aid not only on higher education attendance, but also on retention and degree completion.

While both institutional and survey data are relied upon in this research literature, perhaps the most widely used is the National Postsecondary Student Aid Study (NPSAS).² NPSAS-based studies of institutional student financial aid include U.S. Department of Education (2003), King (2002), and Davis (2003). NPSAS is unique in that it includes not only student-level data on academic success, but also detailed freshman year data on financial aid, including institutional financial aid. While NPSAS has been a valuable tool for research on student aid and higher education performance, this project, based on longitudinal student data from institutions, can make a number of useful contributions to the research literature and policy discussion beyond those provided by NPSAS.

First, the NPSAS survey is designed to draw a nationally representative sample of roughly 12,000 postsecondary students (the longitudinal file analyzed for this study is only one-half this size). However, the process of sampling introduces sampling error. This is particularly problematic when conducting disaggregated analysis by state or particular types of institutions (e.g., selective four-year). By contrast, administrative institutional and financial aid data include the entire universe of students at the participating institutions. Of course, the tradeoff is that while Missouri 's administrative data in this study are representative of Missouri students, they may not be as representative of students in other states.

A second factor is timeliness. The NPSAS has been conducted at five-year intervals, with the most recent surveys completed in 1995-1996 and 1999-2000. There is a relatively long period of time, however, before the data are released and the issuance of reports on higher education topics. For example, the National Center for Education Statistics only recently published a report on institutional financial aid based on the 1995-1996 NPSAS (U. S. Department of Education, 2003). However, as is well known, tuition rates have increased sharply since the mid-1990s. Patterns of student aid from the mid-1990s may provide an outdated statistical portrait of the current situation faced by students. By contrast, state administrative data are available for analysis in a much timelier manner.

Finally, longitudinal institutional data offer important advantages in studying the dynamics of financial aid. While the base (freshman) year financial aid data in NPSAS are extensive, relatively little financial aid data are collected beyond the freshman year, a problem noted by researchers who have used these data (Heller, 2003). Most studies using state administrative data lack data on institutional merit aid (Bettinger, 2004). By contrast, data collected directly from institutions provide accurate information on the level and types of financial aid received each year that a student is enrolled. Building up such a database, in effect, permits states to "follow the mobile student" as recommended in a recent report (Ewell, Schild, and Paulson, 2003) and also provides insights on the use of both need- and merit-based aid by institutions.

Missouri Longitudinal Database

The longitudinal dataset used in this research project was constructed using data from the Enhanced Missouri Student Achievement Study (EMSAS), the Free Application for Federal Student Aid (FAFSA), student financial aid data from the participating institutions, and Unemployment Insurance (UI) wage record files. The institutional financial aid data were requested from the participating institutions and merged with existing data specifically for this research project. The EMSAS data are student-level data on fall enrollment, term-by-term academic progress, and student degree completion. The EMSAS data also include information on student demographics and academic performance in high school. The Missouri Department of Higher Education (MDHE) collects these data each year from Missouri's 33 public two- and four-year higher education institutions. The FAFSA, collected by the Missouri Department of Higher Education and used for the administration of the state's grant and scholarship programs, provides family income and other demographic data on students who apply for most types of financial aid. The student financial aid data represent all of the federal, state, and institutional financial aid that students received at the institutions participating in this study (for details see Appendix Table A1). For a listing of the combinations of student financial aid packages that make research on student financial aid difficult and complex, see Appendix Table A2.

The process of constructing the longitudinal dataset began with all of the students from Missouri high schools who were enrolled as first-time freshmen in fall 1997 in one of the six public four-year institutions participating in this research project. The participating institutions are Missouri Western State College, Southeast Missouri State University,and the University of Missouri - Columbia,University of Missouri - Kansas City, University of Missouri - Rolla, and University of Missouri - St. Louis.³ These 6,375 first-time students were tracked from their freshman year in 1997 through the 2002-2003 academic year to determine the number of credit hours completed, cumulative grade point average, degrees received over that time period, and earnings extracted from the UI data base.

Table 1 presents data on the six Missouri institutions that participated in this study (details are in Appendix B). Missouri Western State College, an open enrollment baccalaureate degree-granting institution, located northeast of Kansas City, draws students from the northeast and north central part of the state, and also enrolls students from Kansas. Southeast Missouri State University, a moderately selective baccalaureate and master's degree-granting institution, is located in Cape Girardeau in the Bootheel region of Missouri. Students in the southeastern and south central regions attend classes at Southeast Missouri State University. The University of Missouri system consists of four selective enrollment doctoral degree-granting and research institutions. The main campus is located in Columbia; the others are located in Kansas City, St. Louis, and Rolla.

The University of Missouri - Columbia campus is centrally located and attracts students from across the state. A Doctoral Research University - Extensive and offering a large number of undergraduate programs, the Columbia campus provides education to a number of students seeking undergraduate and post-baccalaureate degrees. The University of Missouri - Kansas City and the University of Missouri - St. Louis campuses serve the two large population centers in Missouri. The University of Missouri - Rolla, located along the I-44 corridor between St. Louis and Joplin,has long been recognized as a top engineering school and, as such, draws top high school graduates from across the state and region.

With the diversity in both location and mission, the six institutions enroll a representative cross-section of Missouri public higher education students. These participating intuitions account for 48 percent of public four year higher education enrollments in the state. There are seven public four-year institutions in Missouri that did not participate in this study. A comparison between first time freshmen enrolled in participating institutions and first time freshmen enrolled in non-participating public four-year institutions is presented in Appendix Table C 3. The demographic and socioeconomic characteristics of participant and non-participant institutions are very similar. However, since the participant sample includes several selective institutions, the ACT and high-school rank are somewhat higher than in the non-participant group.

Table 1: Characteristics of Missouri Public Four-Year Participating Institutions

Institution	Carnegie Classification	2003-2004 In-state Undergraduate FTE	1996 Cohort Six-Year Graduation Rate^a	2003-2004 Tuition and Fees^b
Missouri Western State College	Baccalaureate Colleges-General	3,657	28.8	$4,464
Southeast Missouri State University	Master's Colleges and Universities I	6,064	49.9	$4,575
Univ. of Missouri - Columbia	Doctoral Research University - Extensive	15,957	65.1	$6,558
Univ. of Missouri - Kansas City	Doctoral Research University - Intensive	4,566	45.1	$6,725
Univ. of Missouri - Rolla	Doctoral Research University - Intensive	3,010	55.3	$6,839
Univ. of Missouri - St. Louis	Doctoral Research University - Intensive	7,217	36.1	$6,866

^a Graduating from the student's home institution
^b Costs reflect a typical full-time undergraduate student taking 30 hours per year.
Source: NCES Graduation Rate Survey.

Figures 1 and 2 present data on in-state tuition and in-state undergraduate FTE enrollment trends for the six participating institutions. Undergraduate in-state tuition and fees at the four University of Missouri campuses are nearly identical and roughly 50 to 70 percent higher than at the two regional institutions. As with many higher education institutions, in-state tuition rose continually during the late 1990s with a sharp acceleration over the last several years, largely in response to state budget withholdings and reductions in core appropriations. In spite of the tuition increases, in-state undergraduate FTE enrollment increased sharply at the University of Missouri - Columbia, the University of Missouri - Kansas City, and at Southeast Missouri State University. Enrollment was relatively flat at the three other institutions.

Figure 1: In-State Undergraduate Tuition and Fees, Based on 30 Credit Hours per Year, for Participating Institutions

Figure 2: In-State Undergraduate Full-Time Equivalent Enrollment for Participating Institutions (Not available)

Finally, since NPSAS plays such an important role in higher education research, it is useful to compare the longitudinal sample with NPSAS. Table 2 provides descriptive statistics comparing financial aid for base-year freshmen (1997-1998) with average freshmen in the 1995-1996 NPSAS. In the six participating institutions, 53 percent of students received institutional aid, compared to just 23 percent in NPSAS. The average aid award in Missouri schools, in dollars and as a percent of tuition, was comparable to NPSAS. These data suggest that, as compared to the NPSAS sample, relatively more aid passes through institutions in Missouri than in institutions in other states. This is consistent with Heller's (2003) finding that there is wide variation among the states in the magnitude of institutional financial aid awarded.

	Missouri Sample (1997-1998)			NPSAS Public Higher Education Sample (Four-Year Institutions, 1995-1996)
	Percent Receiving Institutional Gift Aid	Average Amount of Institutional Gift Aid	Institutional Gift Aid as Percent of Tuition	Percent Receiving Institutional Gift Aid	Average Amount of Institutional Gift Aid	Institutional Gift Aid as Percent of Tuition
All Schools	53.3%	$2,407	74.6%	23.8%	$2,501	76.6%
Less Selective	35.0%	$2,059	83.7%	23.2%	$2,211	75.0%
Very Selective	64.0%	$2,519	69.3%	25.6%	$3,372	81.2%

^a Data are for first-time, full-time freshmen only. The total Missouri sample is 5,879 students from the six four-year participating institutions in the longitudinal student data set. NPSAS data are from the U. S. Department of Education (2003).

Section II: Patterns of Institutional and Non-Institutional Financial Aid

Income

Figure 3 and Table 3 address who receives institutional and non-institutional financial aid and the extent to which need-based aid reaches poor students. Students are grouped by family income into four broad classes: income less than $25,000, income of $25,000 to $74,999, income of $75,000 or higher, and students with no FAFSA.⁴

Students from families with incomes of $25,000 or less (1997 dollars) received $3,046 on average in gift aid.⁵ Interestingly, roughly half of this gift aid was not need-based. In addition, these low-income students received on average $1,556 in loan aid, roughly two-thirds of which was need-based, and approximately $800 in work and other aid. Thus, a good deal of non-need-based aid, particularly gift aid, was awarded to low-income students. On average, total gift aid fell as family income rose and the share of need-based aid of any type decreased sharply for students from higher-income families. Interestingly, the "no FAFSA" group also received substantial gift aid. This is primarily due to the award of the state Bright Flight scholarship for students who scored 30 or above on the ACT. Receipt of this merit-based scholarship does not require students to file a FAFSA.⁶

A common assumption on the part of financial aid officers is that students who do not file a FAFSA are from relatively high-income families and typically receive non-need-based aid.⁷These data suggest that many low-income students also receive significant amounts of non-need-based aid.

Figure 3: Average Need-Based and Non-Need Based Financial Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions^a

^a Financial aid data are averaged over all students (recipients and non-recipients).

Table 3:Average Need Based and Non-Need Based Financial Aid by Family Income in Academic Year 1997-1998:First-Time Freshmen Enrolled in Participating Institutions

		Number of Students	Need Aid			Non-Need Aid			Total
		Number of Students	Percent Receiving Aid	Average Aid per Recipient	Average Aid per Student	Percent Receiving Aid	Average Aid per Recipient	Average Aid per Student	Percent Receiving Aid	Average Aid per Recipient	Average Aid per Student
Gift	Income < $25,000	1,231	64%	$2,547	$1,633	45%	$3,112	$1,413	83%	$3,798	$3,138*
	Income $25,000-$74,999	2,374	39%	$1,851	$721	52%	$2,903	$1,521	74%	$3,123	$2,307*
	Income $75,000 or more	921	5%	$1,506	$78	51%	$2,784	$1,409	56%	$2,738	$1,531*
	Income Missing	1,849	7%	$1,525	$112	37%	$2,936	$1,099	46%	$2,742	$1,259*
Loan	Income < $25,000	1,231	45%	$2,421	$1,097	14%	$3,171	$459	49%	$3,156	$1,556
	Income $25,000-$74,999	2,374	48%	$2,299	$1,105	29%	$3,314	$977	62%	$3,372	$2,082
	Income $75,000 or more	921	15%	$2,127	$328	44%	$3,918	$1,710	50%	$4,045	$2,038
	Income Missing	1,849	7%	$2,128	$159	7%	$2,419	$170	13%	$2,588	$329
Work	Income < $25,000	1,231	16%	$1,109	$183	5%	$1,001	$50	19%	$1,238	$233
	Income $25,000-$74,999	2,374	16%	$1,108	$178	4%	$1,127	$45	18%	$1,225	$223
	Income $75,000 or more	921	4%	$839	$36	2%	$1,172	$25	6%	$983	$62
	Income Missing	1,849	1%	$884	$12	2%	$1,162	$29	3%	$1,180	$41
Other	Income < $25,000	1,231	-	-	-	-	-	-	28%	$1,609	$455
	Income $25,000-$74,999	2,374	-	-	-	-	-	-	30%	$1,641	$491
	Income $75,000 or more	921	-	-	-	-	-	-	19%	$1,307	$254
	Income Missing	1,849	-	-	-	-	-	-	14%	$1,259	$181
Total	Income < $25,000	1,231	71%	$4,131	$2,913	56%	$3,449	$1,922	89%	$6,062	$5,382
	Income $25,000-$74,999	2,374	61%	$3,287	$2,005	66%	$3,867	$2,543	91%	$5,603	$5,103
	Income $75,000 or more	921	19%	$2,371	$443	72%	$4,348	$3,144	81%	$4,809	$3,885
	Income Missing	1,849	11%	$2,577	$283	42%	$3,077	$1,298	56%	$3,255	$1,810

* Total average institutional gift aid per student does not equal the sum of need and non-need institutional gift aid because some institutional gift aid could not be categorized as need or non-need.

Table 3A presents another way to array these data. This table compares the proportion of students by income class to the proportion of the total financial aid awarded. For example, students who filed a FAFSA from families with less than $25,000 in income comprised 19.3 percent of all students. These students received 50 percent of need-based gift aid and 30 percent of overall gift aid. These data demonstrate how financial aid varies with income. Gift and loan aid are most highly targeted to low income students. However, a good deal of loan aid goes to students in the middle income ($50-$75,000) range.

Table 3A: Distribution of Need Based and Non-Need Based Financial Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Family Income		Number of Students	Percent of Students
Family Income		Number of Students	Percent of Students	Percent of Total Need Aid Dollars	Percent of Total Non-Need Aid Dollars	Percent of Total Aid Dollars
Gift	Income < $25,000	1,231	19.3%	50.2%	20.0%	29.5%
	Income $25,000-$74,999	2,374	37.2%	42.8%	41.6%	41.9%
	Income $75,000 or more	921	14.4%	1.8%	15.0%	10.8%
	Income Missing	1,849	29.0%	5.2%	23.4%	17.8%
	Total	6,375	100.0%	100.0%	100.0%	100.0%
Loan	Income < $25,000	1,231	19.3%	29.6%	11.8/%	20.5%
	Income $25,000-$74,999	2,374	37.2%	57.4%	48.6%	52.9%
	Income $75,000 or more	921	14.4%	6.6%	33.0%	20.1%
	Income Missing	1,849	29.0%	6.4%	6.6%	6.5%
	Total	6,375	100.0%	100.0%	100.0%	100.0%
Work	Income < $25,000	1,231	19.3%	32.0%	25.1%	30.2%
	Income $25,000-$74,999	2,374	37.2%	60.1%	43.6%	55.8%
	Income $75,000 or more	921	14.4%	4.7%	9.4%	6.0%
	Income Missing	1,849	29.0%	3.2%	21.9%	8.0%
	Total	6,375	100.0%	100.0%	100.0%	100.0%
Other	Income < $25,000	1,231	19.3%	-	-	24.4%
	Income $25,000-$74,999	2,374	37.2%	-	-	50.9%
	Income $75,000 or more	921	14.4%	-	-	10.2%
	Income Missing	1,849	29.0%	-	-	14.6%
	Total	6,375	100.0%	-	-	100.0%
Total	Income < $25,000	1,231	19.3%	38.7%	17.3%	25.8%
	Income $25,000-$74,999	2,374	37.2%	51.3%	44.1%	47.2%
	Income $75,000 or more	921	14.4%	4.4%	21.1%	13.9%
	Income Missing	1,849	29.0%	5.6%	17.5%	13.0%
	Total	6,375	100.0%	100.0%	100.0%	100.0%

Figure 4 and Table 4 provide a detailed breakdown of gift aid by source. It is important to note that a substantial share of gift aid was institutional, and the vast majority of this institutional gift aid was not need-based. The average award was relatively similar for high-income and non-FAFSA students as for low-income families. Of course, a consequence of this fairly uniform distribution of institutional aid was that low-income students were substantial recipients - 36 percent of gift aid to students with family incomes under $25,000 was non-need-based institutional aid. If need-based aid is added, then institutional gift aid rose to 46 percent of total gift aid for low-income students. In short, the fact that aid is labeled "non-need" does not mean low-income students do not qualify for this aid, do not receive this type of aid or substantially benefit from it.

Table 4A compares proportions of students to proportions of aid and provides a summary comparison of the extent to which these various types of aid accrue to low-income students. Federal aid stands out clearly in this regard. The 19.3 percent of students who filed a FAFSA and came from a family with $25,000 or less in family income accounted for 50 percent of all federal aid dollars. By contrast, the 43 percent of students who did not file a FAFSA or filed a FAFSA and came from families with $75,000 or more income accounted for only six percent of federal aid dollars. The progressivity of federal aid contrasts sharply from the state's. The poorest 19 percent of students received 19 percent of state aid. The 39 percent of students from families with incomes from $25,000 to $75,000 accounted for 40 percent of state aid, while 14 percent of students from families with more than $75,000 in income accounted for 16 percent of financial aid dollars. Institutional aid is somewhat more progressive (i.e., favorable to low income students) than state aid, but less progressive than federal aid.

Figure 4: Average Federal, State, and Institutional Need Based and Non-Need Based Gift Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions^a

^aFinancial aid data are averaged over all students (recipients and non-recipients).

Table 4: Average Federal, State, and Institutional Need-Based and Non-Need-Based Gift Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

		Number of Students	Need Aid			Non-Need Aid			Total
		Number of Students	Percent Receiving Aid	Average Aid per Recipient	Average Aid per Student	Percent Receiving Aid	Average Aid per Recipient	Average Aid per Student	Percent Receiving Aid	Average Aid per Recipient	Average Aid per Student
Federal	Income < $25,000	1,231	62%	$2,305	$1,423	-	-	-	62%	$2,305	$1,423
	Income $25,000-$74,999	2,374	27%	$1,625	$439	-	-	-	27%	$1,625	$439
	Income $75,000 or more	921	2%	$1,728	$26	-	-	-	2%	$1,728	$26
	Income Missing	1,849	5%	$1,488	$77	-	-	-	5%	$1,488	$77
State**	Income < $25,000	1,231	1%	$1,341	$20	15%	$2,056	$307	16%	$2,002	$327
	Income $25,000-$74,999	2,374	1%	$1,378	$7	17%	$2,011	$347	18%	$1,998	$354
	Income $75,000 or more	921	0%	-	-	19%	$2,037	$378	19%	$2,037	$378
	Income Missing	1,849	0%	$1,125	$1	14%	$2,001	$284	14%	$1,994	$285
Institutional	Income < $25,000	1,231	13%	$1,459	$190	43%	$2,578	$1,106	54%	$2,550	$1,388*
	Income $25,000-$74,999	2,374	18%	$1,569	$276	51%	$2,322	$1,174	63%	$2,417	$1,514*
	Income $75,000 or more	921	4%	$1,265	$52	48%	$2,162	$1,031	52%	$2,148	$1,127*
	Income Missing	1,849	2%	$1,353	$34	35%	$2,313	$815	40%	$2,224	$897*

* Total average institutional gift aid per student does not equal the sum of need and non-need institutional gift aid because some institutional gift aid could not be categorized as need or non-need.
** After 1997, Missouri established a need-based financial aid program that is not reflected in this table.

Table 4A: Distribution of Federal, State, and Institutional Need Based and Non-Need Based Gift Aid by Family Income in Academic Year 1997-1998: First-Time Freshmen Enrolled in Participating Institutions

Family Income		Number of Students	Percent of Students

Access and Affordability

Access and Affordability Patterns of Financial Aid and Student Performance for a Cohort of Missouri College Freshmen

Access and Affordability
Patterns of Financial Aid and Student Performance for
a Cohort of Missouri College Freshmen