STATE APPROPRIATIONS: IMPLICATIONS FOR TUITION AND FINANCIAL AID POLICIES

STATE APPROPRIATIONS: IMPLICATIONS FOR TUITION AND FINANCIAL AID POLICIES
STATE APPROPRIATIONS: IMPLICATIONS FOR TUITION
AND FINANCIAL AID POLICIES
By
Matthew J. Foraker
Copyright © Matthew J. Foraker. 2009
A Dissertation Submitted to the Faculty of
THE CENTER FOR THE STUDY OF HIGHER EDUCATION
In Partial Fulfillment of the Requirements
For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2009
2
THE UNIVERSITY OF ARIZONA
GRADUATE COLLEGE
As members of the Dissertation Committee, we certify that we have read the dissertation
prepared by Matthew J. Foraker entitled: State Appropriations: Implications for Tuition
and Financial Aid Policies and recommend that it be accepted as fulfilling the dissertation
requirement for the Degree of Doctor of Philosophy.
_______________________________________________Date:____4-10-2009__
Dr. John J. Cheslock
_______________________________________________Date:____4-10-2009_
Dr. Gary Rhoades
_______________________________________________Date: ____4-10-2009_
Dr. Cecilia Rios-Aguilar
_______________________________________________Date: ____4-10-2009_
Dr. Richard J. Kroc II
Final approval and acceptance of this dissertation is contingent upon the candidate’s
submission of the final copies of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and
recommend that it be accepted as fulfilling the dissertation requirement.
________________________________________________Date:____4-10-2009_
Dissertation Director: Dr. John J. Cheslock
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an
advanced degree at the University of Arizona and is deposited in the University Library
to be made available to borrowers under the rules of the Library.
Brief quotations from this dissertation are allowable without special permission provided
that accurate acknowledgement of source is made. Requests for permission for extended
quotation from or reproduction of this manuscript in whole or in part may be granted by
the copyright holder.
SIGNED: Matthew J. Foraker
4
ACKNOWLEDGEMENTS
To faculty of the Center for the Study of Higher Education: your commitment to
scholarship has been inspiring and motivating. A special thanks to Gary Rhoades for
conducting excellent and thought provoking classes that encourage students to ask deeper
questions and expand their grasp of the big pictures. Most importantly, thanks to my
adviser, John Cheslock, without whose commitment and unwavering support I could not
have completed this work.
To Errol Anderson, whose willingness to take a stand for my possibilities taught me to
believe in my ability to climb over the wall after tossing my hat.
Finally, to my parents who taught me to value education and my own development and
gave me the confidence to pursue my highest aspirations.
5
TABLE OF CONTENTS
LIST OF ILLUSTRATIONS …………………………………………… 6
LIST OF TABLES ……………………………………………………… 7
ABSTRACT …………………………………………………………….. 8
CHAPTER ONE – INTRODUCTION ………………………………... 9
CHAPTER TWO – LITERATURE REVIEW …………….………… 19
Introduction ……………………………………………………. 19
Historical Overview ……………………………………………. 20
The Costs of Higher Education ………………………………… 30
State Appropriations ……………………………………………. 34
Tuition …………………………………………………………… 37
Enrollment Management ………………………………………. 43
Related Studies ………………………………………………….. 47
Pertinent Theoretical Frameworks …………………………….. 50
Conclusion ……………………………………………………….. 53
CHAPTER THREE – DATA AND METHODOLOGY ……………… 55
Introduction ……………………………………………………… 55
Research Questions ……………………………………………… 55
Methodology ……………………………………………………… 57
Data Sources ……………………………………………………… 65
Variables …………………………………………………………. 67
Sample Populations ……………………………………………… 71
CHAPTER FOUR – RESULTS ..…………..…………………..………. 76
Introduction ……………………………………………………… 76
Institutional Level Analysis – Four Year Institutions …….….. 77
Institutional Level Analysis – Two Year Institutions ………… 84
Student Level Analysis – Four Year Institutions .…………….. 93
Student Level Analysis – Two Year Institutions ……………… 101
Summary of Results.……………………………………………. 104
CHAPTER FIVE – CONCLUSION ………………………….…………`107
Introduction……………………………………………………… 107
Key Findings…………………………………………………….. 107
The Big Picture …………………………………………………. 109
Implications ……………….……………………………………. 111
Future Research ………………………………………………… 113
Conclusion ………………………………………………………. 115
APPENDIX A . IPEDS INSTITUTIONAL LEVEL
DATA SOURCES….…………………….……………………… 117
APPENDIX TABLES …………………………………………………… 122
REFERENCES ………………………………………………………… 152
6
LIST OF ILLUSTRATIONS
Figure 1: Loan and Grant Share of Student Aid…………………………27
7
LIST OF TABLES
Table 1. Average Higher Education Tuition and Fees…………………………..
Table 2. The IPEDS Institution Sample Populations……………………………
Table 3. The NPSAS Sample Populations: Four-Year Institutions…………..….
Table 4. The NPSAS Sample Populations: Two-Year Institutions ………..…....
Table 5. Institutional Level Analysis – Four Year Institution
Descriptive Statistics……………………….. ……………………
Table 6. Institutional Level Analysis – Four Year Regressions:
In-State Tuition State Appropriations Coefficient................ ……
Table 7. Institutional Level Analysis – Four Year Regressions:
Out-of-State Tuition State Appropriations Coefficient………….
Table 8: Institution Level Analysis – Two Year Institution
Descriptive Statistics ………………………. ……………………
Table 9. Institutional Level Analysis – Two Year Regressions…… ………........
Table 10: Institution Level Analysis – Fixed Effects Regression ……………….
Table 11. Student Level Analysis – Descriptive Statistics: Four-Year Institutions
Table 12. Student Level Analysis – Linear Regression: In-State Students
at Four-Year Institutions ………………………….………..……..
Table 13. Student Level Analysis – Linear Regression: Out-of-State
Students at Four-Year Institutions …………………………..……
Table 14. Student Level Analysis – Linear Regression: In-State Students
At Two-Year Institutions …………………………………..….….
Table 15. IPEDS / NPSAS / Fixed Effects Regression
Analysis Comparison …………………………………………….
Appendix Table 1. Institution Level Analysis
Complete Four Year In-State Regression………………………..
Appendix Table 2. Institution Level Analysis
Complete Four Year Out-of-State Regression…..………………
Appendix Table 3. Institution Level Analysis
Complete Two Year In-State Regression……….……………….
Appendix Table 4. Student Level Analysis
Complete Four Year In-State Regression…..……………………
Appendix Table 5. Student Level Analysis
Complete Four Year Out-of-State Regression………….………..
Appendix Table 6. Student Level Analysis
Complete Two Year In-State Regression……………….………..
Appendix Table 7. Descriptive Statistics – IPEDS Institutions………………….
Appendix Table 8. Descriptive Statistics – NPSAS Students ……………………
39
72
74
75
78
80
83
85
87
89
94
98
99
103
108
118
119
120
121
123
125
127
137
8
ABSTRACT
Over the past 30 years the costs of higher education have climbed faster than the
rate the inflation. As these costs have risen, state appropriations for public institutions
have not kept pace.
While not declining in real dollars, as a portion of meeting the
expenses of funding public higher education, state appropriations have been steadily
falling over the past three decades. Not surprisingly, during this period tuition at public
colleges and universities has risen dramatically, leading to concerns about access to
higher education, in particular for students of low income backgrounds.
The literature contains many studies highlighting the increasing costs and tuition
charged by public colleges and universities. Little has been written about the specific
relationship between the level of state appropriations at a particular institution and the
pricing and financial aid policies it then adopts. By analyzing the data for public
institutions in the Integrated Postsecondary Educational Data System (IPEDS) as well as
data for specific students in the National Postsecondary Student Aid Survey (NPSAS) for
five school years spanning 1989-1990 to 2003-2004, this study conducts a quantitative
analysis to create a predictive model capable of forecasting the impact of changes in state
appropriation on institution pricing and financial aid policy. In an environment where the
continued decline of state appropriations as a portion of meeting educational costs is a
real possibility, such forecasting ability may prove invaluable in crafting policies to
insure access to higher education for certain student populations.
9
CHAPTER ONE
INTRODUCTION
In the past few decades, state appropriations to public universities have fallen
dramatically as a portion of institutional budgets, from 60 percent in 1969 to 24 percent
in 1995 (Hearn, 1998; Heller, 2002). In 2008 dollars, from 1978 to 2008 average in-state
tuition at public four-year schools rose from $2225 to $6185 or 279 percent, while instate tuition at two-year public institutions grew from $1040 to $2361 or 227% (US
College Board). In response to the public outcry over the escalation of higher education
prices at rates dramatically faster than the consumer price index, politicians have
convened committees, delivered speeches, and published letters on the importance of the
affordability of a college education. Some have introduced legislation to cap price
increases as well as implement policies that would penalize institutions that raise tuition
above a specified percentage. At the same time, they allocate state appropriation levels
that leave a gap in meeting the costs institutions incur. Views differ regarding the causes
of the escalating costs of higher education. Whether it follows from the advancement of
technology and the need to educate for an increasingly complex knowledge economy
workforce, or from schools competing for reputations of prestige and excellence, or from
other causes, costs are increasing. As state appropriations fail to keep pace, as we will
see, schools do not simply make up the difference with tuition. However, clearly the
level of state appropriations received by an institution plays a role in its tuition price, and
the trends in financial aid programs affect the ability of students to obtain a degree.
10
Higher education is a service oriented endeavor that requires not only expertise in
the subject matter, but individuals capable of providing quality instruction in an
environment conducive to learning. Labor intensive interaction between teacher and
student is often necessary for effective education. Like health care, which requires
human interaction between care giver and patient, the nature of education involves a
“cost disease” (Baumol & Blackman, 1989) relative to operations that can be automated
or mass produced, such as the manufacture of physical products, food, or other fixed
items. Technological advances may have in fact exacerbated the costs of providing
higher education. Skilled professionals must be hired to support increasingly
sophisticated laboratories, classroom equipment, computer centers, libraries, and other
technical support services (Rhoades, 1998). Instead of producing the same result at lower
costs, higher education chases a moving target that grows more expensive to produce.
Contributing to the increase in the cost of higher education is the competition that
exists between institutions of similar type. Many schools, in particular four-year public
flagship research universities, operate in an environment of intense competition for
prestige and a reputation for educational excellence (Ehrenberg, 2001). In an “arms race”
to attract the best students and faculty, schools experience pressure to outspend their
rivals to produce more attractive campuses, better facilities, first rate Web sites, superior
athletic programs, and other enticements to improve their standing in rankings such as
those published by US News and World Report. The pressure to “pursue excellence”
among the nation’s colleges and universities may have become so institutionalized that
even schools substantially down the prestige hierarchy embark on efforts to “climb the
11
ladder.” In what Bowen called the “revenue theory of cost” in his 1980 book, schools
generate as much revenue as possible via fundraising, tuition, bond elections in local
communities, various grants and other donations, or other sources. Once this revenue is
understood, the full amount is allocated towards a prioritized “wish list” until costs meet
the total revenues obtained (Bowen, 1980). According to this theory, only the amount of
revenue generated restricts the costs that can be incurred.
In this study I implicitly examine the influence of three interrelated broad forces,
1) the rising costs of providing higher education, 2) the state appropriations that are not
keeping up with these increasing costs, and 3) the implications this gap has for the pricing
(tuition) set by public institutions and the financial aid made available to students in
meeting those prices. As we will see, both the setting of a tuition price and the level of
state appropriations received by a public university follows a complex set of factors
including local politics, state resources, university budgets, and competing interests for
public support. The financial aid provided to students is also determined by a complex
system that combines Federal grants and loans, state merit and need programs, and
institutional aid. The model produced in the study quantifies the relationships between
variables representing the factors and seeks to explain some of the influences that exist
between them.
The relevance of the study will only increase given the scenarios likely to exist in
the future. The outlook for state appropriations for higher education suggests a
continuation of the trends of the past few decades. In funding public higher education,
states face tremendous pressure for scarce resources from other important and competing
12
community needs including K-12 education, state infrastructure, correctional facilities,
and Medicaid. States often turn to higher education as a budget balancing tool during
recessions and other difficult budget situations. In the face of these stagnating state
appropriations, schools must either prevent costs from rising or increase revenues from
other sources. As we have seen, costs have not been reduced. A natural alternative
revenue source is tuition and fees. However, to simply infer that tuition is replacing state
appropriations in meeting the costs of instruction overlooks the complexity of higher
education pricing and financial aid. The nature of the influence of state appropriations
levels per student at a public institution on its tuition price can be expected to vary based
upon operating environment in which it functions. Institutions in different states may
face different governance structures linked to legislatures and political relationships with
different priorities and agendas. Large research universities with prestigious business
schools, medical schools, or particular departments recognized as the best in their field
have a broader range of choices in response to a change in funding than, for example, a
local community college serving a rather fixed population of nearby students.
Tuition has risen dramatically in the past three decades at both private and public
higher education institutions. While in total dollars the increases have been the greatest
at elite, private schools, in percentage terms the public institutions have raised tuition as
much and in many cases more, including tuition at two-year community colleges. At the
same time that costs and tuition levels have risen, financial aid policies have shifted at the
federal, state, and institutional level. Federal financial aid provided directly to students
has shifted dramatically from a grant focused approach to that of loans which must be
13
repaid, and the federal aid system has significantly relaxed its requirements on student
need, allowing even relatively affluent students to obtain low cost loans for attending
college (Hearn, 1998). States have used tobacco settlements and lotteries to fund state
financial aid programs that emphasize student merit over financial need. Both of these
trends have impacted access and attendance patterns of lower income students, who are
loan averse and disadvantaged against wealthier students in merit based programs. At the
institutional level, the use of tuition discounting and institutional grants has grown
significantly (Redd, 2000). While partially motivated by the desire to provide aid to
students based on need, institutional aid is significantly motivated by enrollment
management considerations, and often students are awarded grants based on non-need
factors such as academic ability, demographics including gender, and indeed, the
student’s ability to pay high tuition (e.g. someone from a wealthy family or an out-ofstate student) (Wilkinson, 2004).
In this study, at both the institutional level and the student level, I examine the
impact of state appropriation support per full time student provided to public higher
education institutions on the list tuition prices they set and the financial aid packages
provided to students. At the institutional level, I use data from the Integrated
Postsecondary Education Data System (IPEDS) to run regressions quantifying the
relationship between state appropriations and tuition prices taking into account the
selectivity of the organization and the higher education governance system of the state in
which they operate. The study separately examines five years during which the NPSAS
studies were conducted, 1989/90, 1992/93. 1995/96, 1999/2000, and 2003/04. For these
14
years, I compare the results with those obtained using student level data from the
National Postsecondary Student Aid Surveys (NPSAS) for students attending the same
set of institutions. The study will clarify our understanding of the role state
appropriations plays in pricing and financial aid policy.
At the level of institutions, the analysis quantifies the impact of state
appropriations on 1) the listed “sticker price” tuition at public universities and community
colleges while controlling for the selectivity of the institution and the governance
structure of the state in which it operates. In-state tuition is distinguished from out-ofstate tuition, and four-year schools are run separately from two-year schools. For the
four-year institutions, I also disaggregate the schools by classification of selectivity to
discern any differences between the different types of schools, anticipating that more
selective institutions may respond differently when changes to their appropriations
occurs.
At the level of the individual students, the analysis quantifies the impact of state
appropriations on 1) the listed tuition price, 2 ) the “net tuition price the student paid after
receiving a grant (not to be repaid) from the school, 3) the institutional grant provided to
the student, and 4) the “unmet need” as reported by the data set, equal to the difference
between the student’s expected family contribution and the price the student actually
incurred. The NPSAS data set with student level data allows the analysis to operate at
more refined level, permitting one to control for student characteristics including
socioeconomic status (using expected family contribution as a proxy), academic merit
(using their cumulative grade point average at the school), and ethnicity and gender.
15
Since the student analysis can control for these factors, I anticipate student level data to
paint a different picture and will explore why these differences may exist.
Purpose of the Study
The purpose of the study is to examine the relationship between state
appropriations and higher education pricing and financial aid policies. Using detailed
quantitative analysis of prices charged and aid awarded to students attending public
colleges and universities across the country, the study will produce a model with the
ability to generate forecasts of the probable pricing and financial aid consequences that
will occur at a public two-year or four-year college or university when it faces changes in
state appropriation levels. This tool will provide additional clarity in the understanding
of the relationships between state appropriations, governance, and the financial
transactions that occur between students and the institutions they attend. I investigate the
theories applicable to the analyses and address how the theories might explain the
findings of the model.
Finally, the study uses existing price and aid response literature
to explore the implications of the findings on access and participation of different student
populations in higher education.
Organization of the Study
The study consists of five chapters. Chapter one provides an introduction to the
study including its purpose, approach, and its contribution to the study of higher
education. Chapter two presents a brief overview of the history of higher education in the
United States before conducting a review of the pertinent literature related to public
support for higher education, its costs, pricing, and the financial aid programs designed to
16
help students enroll. I then present the political, economic, and other organizational
theories applicable to the study and relevant to the interpretation of its findings. Chapter
three provides the details of the methodology of the study and the data sets utilized. I
explain the selection of the sample populations and variables used to produce the
econometric models. Chapter four provides a detailed discussion of the results obtained
and relates them to the frameworks and theories under consideration. Chapter five
provides an overview summarizing the results and the conclusions one can draw from the
study along with the implications they may have for policy makers.
Summary and Significance of This Study
An abundance of literature exists regarding rising costs of providing higher
education and the tuition levels charged to students. Many studies have also documented
the declining share of state appropriations in public university budgets. Descriptive
statistics regarding tuition levels, financial aid awards, state appropriations, and
attendance patterns have been published, but little quantitative work has been performed
in an effort to create predictive models that can forecast the likely impact of changes in
one area (in this case, state appropriation levels) on what occurs in other areas (tuition
prices set and financial aid packages provided). Further, the extent to which this
relationship may vary for different kinds of students (family income, gender, ethnicity,
academic preparation, in-state status) as well as for different kinds of institutions
(selectivity, two-year or four-year, state governance environment) has yet to be explored.
A deeper understanding of the implications of the growing gap between public
support for higher education and the costs of providing it, and by implications I refer to
17
the resulting tuition prices and financial aid packages made available to aspiring students,
is critical to providing policy makers the information necessary to produce legislation that
leads to intended consequences to the extent possible. Without such knowledge, policies
may be set that have unintended and undesirable consequences for access to higher
education and higher education programs in general.
Public support for higher education remains politically viable, and it is likely to
continue to be an important component in assisting the majority of American college
students. However, the mechanisms for financing college attendance are clearly shifting
in focus and intent. Federal assistance has moved from a predominantly need justified
grant orientation towards the use of subsidized (and unsubsidized) federally backed loans
with less strict need qualifications (Hearn, 1998). States have moved towards direct aid
to students based on merit. As these shifts occur and occur differently in different states
across the country, what can be said about the impact of changes in unrestricted state
appropriations for higher education, traditionally the greatest source of financing higher
education? A quantitative analysis producing a model providing predictive capability of
the effect of such changes on pricing and financial aid can lead to the next step in
understanding impact on access and participation in higher education in the United
States.
The implications have substantial relevance in light of the reality that the current
downward trend in the share of state appropriations in financing public higher education
will continue. Even prior to the financial upheaval in the fall of 2008, states have been
contending with tremendous pressure for support from other important community needs
18
including transportation infrastructure, the public kindergarten through high school
education system, departments of corrections, and Medicaid for a growing population of
seniors. All of these will continue to compete with higher education for scarce public
resources, suggesting that the findings of this study will be important and remain
important into the foreseeable future.
19
CHAPTER TWO
LITERATURE REVIEW
Introduction
The review begins with an overview of the history of public higher education
finance and summarizes the key developments that led to the current system. After
introducing the broad historical background of higher education, and I provide a brief
overview of the period from the industrial revolution and the need for a more skilled
workforce through the world wars, the cold war, and the Higher Education Act of 1965. I
then present the history of higher education finance in greater detail, discussing the major
federal legislation and the financial aid programs produced at the federal, state, and
institutional level. I discuss the more recent trends associated with the costs of higher
education and cover the two fundamental cost theories most often referenced when
discussing the growing expenses associated with providing higher education. I then
address the developments taking place regarding state appropriations for public colleges
and universities and how the trends in costs and state appropriations are impacting tuition
policies. After I explore the trends and the literature discussing tuition in detail, I discuss
the development of the enrollment management function and review the literature that
examines the ways in which both institutions and students are responding to all of the
above. Finally, I explore the economic and institutional theories pertinent to these issues
and provide a conclusion of the review and an overview of its implications for public
higher education in the United States.
20
Historical Overview
Prior to the Civil War, higher education existed as an option only available to a
select few born into wealth and privilege. Most Americans regarded the completion of
high school as ample education necessary to enter the workforce in the nineteenth century
economy that was heavily agricultural, retail, and trades and crafts to apprentices taught
by their employers as their employers were once taught as apprentices a generation
earlier. For the small, elite (and male) privileged set of youth destined to become
lawyers, doctors, politicians, and other civic leaders, prestigious private schools existed
to develop leaders and professionals with a curriculum rooted in philosophy, Latin,
culture, critical thinking, and the classics of western civilization. The private schools at
this time still struggled for enrollment and admitted most who applied and could pay the
fees.
As the nation shifted to an industrial economy and the rise of the corporation,
organizations emerged that had management and control structures requiring managers,
accountants, engineers, scientists, bookkeepers, attorneys, and other professionals
necessary to run increasingly sophisticated manufacturing lines, oil refineries, chemical
plants, and other operations. These positions required workers educated beyond the high
school level. On July 2, 1862, President Abraham Lincoln signed The Morrill Land
Grant Act of 1862 which gave the states federal lands for the establishment of colleges
that not only provided the traditional academic curricula, but also offered formal training
in agriculture, mining, engineering, home economics, and other “practical” subjects
(Wilkinson, 2004). Over time all of the states in the nation built land grant universities
21
that subsidized the cost of educating students with state appropriated funds provided to
schools as unrestricted support per student enrolled. This funding allowed universities to
charge little or no tuition. In some states, such as New York and California, tuition
remained free for state residents well into the twentieth century. Enrollment at public
universities steadily grew as the industrial economy matured and electricity, the
automobile, and the department store became common place.
Large numbers of soldiers returning from World War Two combined with Cold
War pressure for technological superiority to the Soviet Union dramatically increased the
nation’s interest in having as educated a population as possible. On June 22, 1944,
President Franklin D. Roosevelt signed the Servicemen’s Readjustment Act of 1994,
better known as the GI Bill. The landmark legislation, controversial at the time, fueled
unprecedented levels of participation in higher education and altered the perception of
higher education itself, now seen as viable for many if not all instead of a select few. The
Soviet Union’s launch of Sputnik in 1957 intensified the government’s interest in an
educated workforce, particularly in math and science. In 1958 the National Defense
Education Act (NDEA) provided aid to education at all levels, but particularly targeted at
programs seen as benefiting national security, including technical education, math,
science, and also modern foreign languages. The legislation also established the National
Defense Student Loan Program (NDSL), which offered low interest loans to qualified
students.
President Lyndon Johnson’s “War on Poverty” effort noted that college graduates
earned more income than high school graduates. The support of widespread participation
22
in higher education fit into the “Great Society” agenda that included Medicaid, the
Voting Rights Act, and the Elementary and Secondary Education Act (Hannah, 1996).
This combined with the view that an educated population enhanced society led to the
profound and far reaching Higher Education Act of 1965 (Hearn, 1998; Wilkinson,
2004). President Johnson signed the legislation in November of 1965, and it funneled
billions of dollars into the higher education system, its expansion, and financial support
for students to pay the price of attendance. Title IV of the Act established the Federal
Guaranteed Student Loan (GSL), the Equal Opportunity Grant (EOG), and the College
Work Study Program (CWS) (Hearn, 2002). All of these programs sent participation in
college spiraling upward. In 1940, 16 percent of eighteen to twenty-one year olds
attended college; by 1970, the portion had tripled to 48 percent enrolled in college
(Wilkinson, 2004). The nation’s culture shifted further. Not only was higher education
now seen as viable for all, but many came to believe everyone, including ethnic
minorities and women, should have the right to obtain a college degree.
Starting in the 1970s public sentiment began to shift. The financial aid programs
enacted by the Higher Education Act of 1965 provided the families of students the means
to finance higher education, lessening the need for universal low tuition (or in some
states, no tuition at all). In 1968 free market proponent Milton Friedman published an
article suggesting that the laws of supply and demand should apply to higher education,
with students paying the full price of the market determined cost of their education
(Friedman, 1968). The “High Tuition High Aid” model argued that universal low tuition
was inherently inefficient, allowing wealthy students who could easily pay the full cost to
23
enjoy highly subsidized prices. While few endorsed the severity of Friedman’s
suggestion, support grew for what came to be called the “rationalization” of tuition, i.e.
having tuition prices more closely reflect the costs incurred by the institution, with the
belief that removing the “blanket subsidy” for all students led to improved equity and
efficiency (Hearn, Griswold, & Marine, 1996).
In addition to the voices advocating tuition prices more in line with the costs of
the education were sentiments that supported the idea of introducing competition into the
market for higher education. The Higher Education Amendments of 1972 radically
altered the landscape of higher education finance by attaching aid to the student instead
of the institution, based on the argument that empowering students with choice would
foster competition and provide incentive for higher education institutions to improve
quality and prestige to attract better students (Wilkinson, 2004). The amendments
created the Basic Education Opportunity Grant Program (BEOG), which placed the
financial support at the disposal of the enrolling student to use at the institution of choice.
The legislation also renamed the National Defense Student Loan as the National Direct
Student Loan, and the Economic Opportunity Grant was renamed as the Supplemental
Education Opportunity Grant (SEOG). In a prescription for trouble, the law allowed
proprietary (for-profit) schools to use Title IV financial aid programs. The for-profit
institutions became most adept at assisting students in obtaining financial aid. Caught in
the middle without the means to pay tuition but failing to qualify for federal need based
aid, middle income families successfully exerted political pressure to succeed with The
Middle Income Student Assistance Act (MISAA) of 1978. The Act removed income
24
restrictions on federally subsidized student loans (GSL), which sent the use of such loans
soaring from $3.389 billion in 1975-76 to $10.623 billion in 1980-81 (Hearn, 1998). In
1981, income restrictions were restored, but not at levels prior to the MISAA legislation
(Wilkinson, 2004).
Viewing government as “the problem” and the private sector as the solution, the
Reagan administration battled Congress regarding higher education funding and financial
aid, arguing that the Department of Education was not necessary and calling for its
dissolution. Secretary of Education William Bennett asserted that “greedy” higher
education institutions raised tuition knowing students would receive federal aid to pay for
it, an assertion never proven (McPherson & Shapiro, 1991) for public institutions and
refuted for private schools (Wilkinson, 2004). Reagan failed to eliminate financial aid or
the Department of Education, but he succeeded in producing a profound shift in the
culture regarding higher education. The student, not society, received the bulk of the
benefits of a college degree. Accordingly, the students and their families should bear the
bulk of the cost (Hearn, 1998).
In 1980, Congress wanted to honor a man known for his advocacy of higher
education participation and attainment on the part of those requiring financial assistance,
Senator Claiborne Pell of Rhode Island. Pell was largely responsible for the creation of
the Basic Educational Opportunity Grant (BEOG), so the Higher Education Amendments
of 1980 renamed the program as the Pell Grant, a direct financial assistance program
available to all students seeking to enroll with demonstrated financial need. The
legislation also created the Parent Loan for Undergraduate Students (PLUS) that allowed
25
middle income families to borrow $3,000 a year for each dependent child regardless of
the family income (Wolanin, 1998).
The Higher Education Amendments of 1992 sought to rationalize and standardize
the financial aid system as much as possible. The political environment of 1992,
however, lacked the cohesive thrust for education that existed during Johnson’s “Great
Society” (Hannah, 1996) and despite its efforts, the results continued the “alphabet soup”
of overlapping federal and state efforts to promote college enrollment. The legislation
did create a single methodology for the determination of the cost of college attendance
and the ability of a dependent or independent student to pay for it. It mandated for all
students seeking aid to solely use the Free Application for Federal Student Aid (FAFSA),
and it standardized the calculation of student need to a single method called the Federal
Need Analysis. The act also provided a system for the pro-ration of Pell Grants and
Stafford Loans to adjust for attendance not fitting a complete academic year. Additional
name changing took place, with the renaming of the Guaranteed Student Loan Program to
the Federal Stafford Loan in honor of Senator Robert Stafford of Vermont. The Stafford
Loan Program also created an additional loan program for students who did not qualify
for the established, federally subsidized loans. These were called the Stafford
“Unsubsidized” loans and while less generous than the subsidized loan programs, they
still offered attractive loan options for otherwise ineligible students (Hannah, 1996).
In 1997, President Clinton signed the Taxpayer Relief Act which established tax
credits for higher education expenses. It created the HOPE Scholarship Tax Credit
giving up to a $1500 tax credit for each of the first two years of college attendance. The
26
interest on loans for education was made tax deductible, and it created the ability for a
family to establish an educational “individual retirement account” (IRA) for each
dependent child under 18. The law also permitted the withdrawal of funds from existing
IRA’s without penalty if the funds were used for college attendance.
What did all of this mean for the students? Perhaps the single greatest difference
experienced by students and their families over the past 30 years was the federal financial
aid system’s substantial shift from a need-based grant orientation towards a significantly
less need sensitive loan orientation. This change resulted in students graduating with
debt burdens so great that James Hearn wrote of the system graduating “indentured
servants” (Hearn, 1998). Some middle and lower income students were graduating with
four-year degrees and college loans exceeding $40,000, more than their annual salaries
upon graduation. The shift in the federal financial aid system from grants to loans shifted
the share of total aid provided to students to one based more upon loans than upon grants
(Baum & Payea, 2003), as shown in Figure 1.
27
Figure 1. The Percent Share of Total Financial Aid, 1982/83 to 2002/03
Percent Share of Total Aid
65
Percentage
60
55
50
Grants
45
Loans
40
35
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
30
Academic Year
(Graph produced with data from Trends in Student Aid, College Board, 2003)
As the Federal government shifted to a loan orientation for helping students pay
for college, the states, using settlements from tobacco lawsuit settlements and/or revenues
generated from state lotteries, shifted from a need-based focus towards a focus on
rewarding students based on academic credentials and performance while attending
college. In 1993, Georgia created a program called “Helping Outstanding Pupils
Educationally” (HOPE) (Heller, 1997). Florida followed quickly with a program called
Bright Futures. Other states quickly implemented merit based programs. While state aid
based on need did not decline in absolute dollars, state merit aid increased at a
dramatically higher rate. From 1982 to 1998, state need based aid increased 88 percent in
real dollars. In the same time period, state merit aid increased 336 percent, most of it
since 1993 (Heller, 1998).
28
Both the federal aid trend towards loans and the state aid trend towards merit
based rewards tend to leave lower income students, who are less likely to qualify for
merit aid and are loan averse (Heller, 2003) without any means of financial assistance.
Higher education institutions seeking to maintain access as well as maintain diversity in
their classrooms resorted to institutional aid to recruit students. Institutional aid
(especially grants) is the fastest growing segment of financial aid in higher education
(Heller, 1997). Using NPSAS (National Postsecondary Student Aid Study) data for
1989-90 and 1995-96, Heller found that institutional grants to undergraduates climbed
from $386.6 million to $928.4 million at public institutions, the fastest growing type of
financial aid to students (Heller, 1997). Institutional aid to students exists within the
context of enrollment management (which is discussed in depth later in this chapter),
which may (or may not) consider student need a priority. While institutions may target
aid with financial need as one factor, other factors including the academic merit,
ethnicity, gender, or other aspects of the student’s background, (in particular, the
student’s ability to pay as an attractive means to boost net tuition revenue) may be
considered in determining the aid package. Further, universities do not appear to employ
institutional financial aid practices that compensate for shortcomings in state level
financial aid policies. Doyle, Delaney and Naughton (2003) found that instead of using
institutional aid to fill gaps or replace aid reductions caused by state shifts in financial
aid, the schools instead distributed financial assistance in a manner that reinforced the
state position, complying with and furthering the state’s philosophy regarding which
students should receive what amounts of financial assistance. In those states moving
29
towards merit based aid programs, institutions only exacerbate the shortfalls created for
need based aid funding (Doyle, Delaney, & Naughton, 2008).
The debate over the financing of higher education has continued to the present
day. The stratification of institutions in terms of prestige has continued to grow with the
gap widening between private schools and public universities. Economic inequality in
the United States increased during the Bush administration, the period where the
wealthiest ten percent of Americans realized a 24% increase in income while the median
income for the nation fell 8% and income for the lowest ten percent of Americans
dropped 17% (Bureau of Labor Statistics, 2008). Despite the efforts to rationalize the
financial aid system, in particular the Higher Education Amendment of 1992, to this day
we still have a system where college access and choice remains stratified by
socioeconomic status and ethnicity (Thomas & Perna, 2004). Hearn (2001) noted that
instead of a unified and coherent system of programs with clarity and “distinctiveness,”
the system continues to consist of overlapping efforts and objectives. The federal
programs implemented during the Great Society Era have been modified, adjusted, and
renamed. The states have each implemented their own financial assistance programs
with shifting goals and objectives. As a result, we have a confusing array of federal,
state, and institutional programs to promote college enrollment (Perna, et. Al; 2008) that
can overwhelm graduating high school seniors considering higher education, in particular
those who would be the first time college students in their families.
30
The Costs of Higher Education
The costs of providing a higher education have risen dramatically in the last few
decades. Baumol and Blackman (1989) first coined the term “cost disease” to refer to
labor intensive activity that cannot realize the productivity gains made possible by
technology or the streamlining of processes that can occur in other endeavors such as
manufacturing or information technology. According to the “technology theory of
costs,” Baumol and Blackman assert that services fitting this category such as education,
health care, legal representation, and others, will continue to experience rising costs
above the consumer price index (Baumol and Blackman, 1989).
Bowen’s revenue theory of cost offers a different perspective. This view sees
higher education institutions seeking to maximize revenue in the pursuit of the greatest
degree of excellence possible. The increasing costs of higher education develop from the
intense competition for status and prestige, exacerbated by the increasing stratification of
higher education institutions (Trow, 1984) and widely publicized and read rankings of
colleges and universities in the US News and World Report Magazine (Ehrenberg, 2002).
The top private elite universities with massive endowments have the ability to spend
extraordinary sums of money to increase status to attract and recruit top talent for both
faculty and students. Princeton’s endowment per student in 2006 was estimated at $1.9
million, with several other top private universities also having endowments per student
exceeding $1 million (NACUBO Endowment Study, 2006). With respect to the public
universities, the revenue theory implies that institutions will already be setting tuition at
levels which maximize revenue, so an examination of the affect of state appropriation
31
changes on tuition prices should find little or no relationship. The institutions maximize
revenue subject to constraints placed on the student body enrolled in terms of gender,
ethnicity, academic merit, or other institutional imperatives (Bowen, 1980).
Bowen’s theory has drawn a lot of attention and is plausible given the
extraordinary stratification of educational costs in higher education institutions. Schools
at different levels in the hierarchy of prestige spend vastly different amounts teaching
their students (Winston, 1999). While the students at highly selective private universities
may pay far greater levels of tuition, they get what they pay for (Winston, 1999) because
the well endowed institution has the resources to incur educational costs far exceeding
the amount of the tuition. While a community college may charge $500 for a public (and
state subsidized) program that costs $800 per student, an extremely expensive, selective,
and elite private school, while charging $40,000 tuition, may be incurring costs
exceeding $100,000 for each student who enrolls. As noted earlier, at the top of the
hierarchy, private institutions and (increasingly) public flagship research institutions have
endowment incomes per student exceeding the listed tuition price (NACUBO
Endowment Study, 2006). Bowen’s revenue theory is an appealing perspective in an
environment of such competition.
Universities have engaged in an academic prestige “arms race” with a scoreboard
published each year that Ehrenberg (2002) has shown to influence the number of
applications a school receives and the enrollment yields of students admitted. The race
pits institutions into competition not only for top students, but top faculty, and while the
public universities have all but lost any capability to compete with top private institutions
32
for faculty (Alexander, 2001), they certainly compete with each other. In the competition
for faculty, universities can use means other than direct compensation by offering lower
teaching loads or other incentives. Zemsky (1994) coined the phrase of “The Academic
Ratchet” regarding the downward pressure on faculty teaching loads and its contribution
to the escalation of educational costs.
While the revenue theory may not fully explain the rising costs of providing
higher education, it does predict activities that universities have begun to enhance their
revenue streams. Public schools now generate endowments. They conduct alumni
fundraising campaigns or other programs to solicit donations. Public research
universities seek lucrative research grants and business ventures at technology parks. At
the undergraduate level, schools have a greater profit orientation at student recreational
centers, student unions and cafeterias, parking fees, and others (Levy, 1995; Slaughter &
Rhoades, 2004). The key point to be made here for this study is that the revenue theory,
the diversification of revenue streams, and shifting financial aid all suggest that the
relationship between state appropriation levels and tuition at a particular institution, while
intuitively appealing, is anything but straightforward.
Some question the extent to which the revenue theory really applies to higher
education institutions. Getz and Siegfried (1991) listed six competing (with the revenue
theory as well as each other) explanations for the escalation of costs: 1) the cost disease
introduced by Baumol and Blackman, 2) a higher education shift towards more expensive
disciplines, 3) shortages of higher education inputs, 4) faculty and administrators in
charge having inflated desires for quality, 5) poor management in higher education, 6)
33
costs arising from government regulations creating expanded duties for higher education
(Getz & Seigfried,1991). While the desire of the institution for quality may fit into
Bowen’s ideas, the rest are distinctly different and more thematically in line with cost
disease. In a study designed to compare the explanatory effectiveness of the revenue
theory of cost with that of cost disease, Archibald & Feldman (2008) examined unit costs
for a large variety of goods and services in the United States from 1949 to 1995. If cost
disease were the major factor, other products and services also affected by the theory
should experience similar cost increases. If Bowen’s theory were more significant, cost
increases should be quite different from those to whom it applied. They found that:
Cost per student in higher education follows a time path very similar to the
time path of other personal service industries that rely on highly educated
labor. This is entirely consistent with the cost disease explanation of the
rise in cost in higher education. This explanation is based on strong
economy-wide influences that affect industries that tend to experience
lagging productivity growth and that rely on highly educated labor, not on
characteristics of higher education itself. (Archibald & Feldman, 2008)
The findings cast doubt on the universal application of Bowen’s revenue theory to
higher education as a whole, but do not go so far as saying that it does not apply in
certain groups of institutions or certain departments within an institution. A study that
applies the ideas of Archibald and Feldman that also compares the results of different
groups of schools after disaggregating the institutions by selectivity, size of endowment,
or other measures associated with prestige may yields interesting and useful findings, as
would a study that examines the tuition price response and financial aid changes that
occur when institutions experience a change in public funding.
34
State Appropriations
As mentioned earlier, the Morrill Land Grant Act of 1962 provided federal land
(and funds from the sale of federal land) for the purpose of created state supported public
higher education rooted in the practical advancement of knowledge if fields of use to the
economy. States provide funding appropriations, usually based on the number of
students, to offset the cost of the education and permit schools to keep tuition well below
the actual cost of the education. Known as universal low tuition, the practice rose from
the belief that education benefited society as a whole, furnished the national economy
with workers it required, and improved the country’s ability to compete (Heller, 2002).
At all non profit higher education institutions, both private and public, educational costs
equal price plus a subsidy (Winston, 1999). Historically, to keep the price well below the
cost, private schools used endowments and fundraising, and public schools used state
appropriations. While this continues to apply to private institutions, during the last three
decades the financing of public higher education has shifted away from taxpayer
subsidies towards individual students and their families. The conversation has shifted
from education as a public good to be financed predominately at public expense, to
education as an investment on the part of the students in anticipation of enhanced
employment opportunities and income. While Milton Friedman’s desire to see the free
market fully implemented in higher education has not occurred, during the last three
decades the promotion of competition both on the part of students as consumers and on
the part of institutions as suppliers has clearly shifted the production and consumption of
higher education in the direction of the free market.
35
State appropriations to public colleges and universities have not kept pace with
increasing costs (Selingo, 2003). While in absolute dollars, state subsidies have not
fallen, as a share of public institution budgets, state appropriations have declined since
1980 (Hearn, 1998; Heller, 2002 ). In 1969, the state covered 60 percent of the
expenditures share, while tuition and fees covered 15 percent, a spread of 45 percent.
The 1969 difference between state appropriations and tuition and fees was 45.1%, which
peaked at 47.4% in 1978 and has been falling ever since. By 1995, it had fallen to
24.1%, half of its prior value (Heller, 2004).
The conservative shift in the view of the federal government’s role in social
programs that started with the Reagan administration saw the federal government transfer
responsibility for many programs to state or local governments (Weertz & Ronca, 2006).
From 1980 to 1992, the higher education expenditures covered by the federal government
declined from 18 percent to 14 percent (Hossler, 1997). Partially due to political
sentiment, and partially out of necessity, states started to view higher education, now
competing with Medicaid, corrections, and other social services for funding, as a more
“optional” program in tough times of recession and economic scarcity. Seen as a
“balancing wheel” component of the state budget, higher education faced likely
reductions during recessions or budget crisis situations. (Selingo, 2003; Callan, 2002).
One consequence of this perspective is that in many cases the tuition levels set by state
legislatures and higher education institutions hardly follows carefully crafted strategy or
rational policies. Rather, as Mary McKeown (1982) noted, higher education public
institution pricing policies are often “backed into” after addressing budget short falls, and
36
Griswold & Marine (1996) also found that “many states appear to increase tuition in
reaction to situational factors rather than developing a rational policy or even a planned
response to perceived needs.”
Johnstone (1993) observed numerous obstacles to the implementation of a hightuition high aid model. He noted that aid for middle and low income students may not
keep pace with rising tuition, and that the rapidly rising costs of attending college had
political repercussions, an assertion confirmed when Congressmen McKeon tried to
introduced legislation punishing universities if they raised tuition beyond a certain
“affordability index” (McKeon, 2003). Clearly, the tuition levels of public higher
education institutions, in particular the in-state tuition paid by the families of voters, are
influenced by political considerations. McLendon, Hearn, and Hammond (2008)
conducted a panel study of the tuition prices at 162 of the nation’s top public universities
over the period from 1984 to 2002. They found that the setting of tuition levels is a
complex function of many factors including a state’s population and economic climate,
state appropriations and aid policies, higher education governance structures, and
importantly, the state’s political environment. Higher levels of minority representation in
state legislatures were associated with lower tuition prices at the state’s public flagship
institutions (McLendon, Hearn, & Hammond, 2008). Weertz & Ronca (2006) also found
political factors played a key role in the tuition prices at public universities. They found
that the support for higher education in the form of state appropriations per student was
most associated with a) campus commitment to public service and community outreach,
b) the higher education governance system, and c) gubernatorial and legislative support.
37
As state appropriations decline in relevance as a share of meeting educational
costs, simple budget balancing demands that other revenue sources must increase in
relevance, the most obvious replacement being net tuition revenue. At one extreme,
where public support fully covers the institution’s budget, we have schools providing
universal low tuition to all applicants, a system most closely approached during the 1960s
and 1970s with the large expansion of higher education participation. The opposite
extreme, Friedman’s desire for a free market in higher education, would have students
and their families pay the full cost of their education, minus whatever financial aid is
made available to them on a case by case basis. Without getting into the full implications
of removing the “public” from public higher education institutions, the political reality of
voters with children desiring a college education is likely to keep federal and state
support in place for the foreseeable future. The continued slippage in the share of state
appropriations as a portion of university budgets, however, remains likely (Selingo,
2003).
Tuition
In the face of rising costs and the leveling off of state support, the notion that
institutions simply pass the difference off to students in the form of higher tuition
oversimplifies the complex response of public universities and students to the
development. That said, tuition levels have indeed risen at rates far exceeding the
consumer price index and the cost of educating the students (Ehrenberg, 2002). The
stagnation of state appropriations combined with the cultural shift to the idea that
students receive the bulk of the benefits of a higher education has increased scrutiny on
38
universal low tuition. Opponents of universal low tuition noted that wealthy and upper
income students, those most likely to participate in higher education, could easily afford
to pay higher tuition levels and free the taxpayer by reducing the subsidy higher
education receives (Hearn & Longanecker, 1985). Theoretically, raising tuition and also
raising aid for those with financial need results in a more precise allocation of assistance
to those that required it. Wealthy students would pay an amount equal to (or at least
much closer to) the true cost of their education.
Hearn, Griswold & Marine (1996) noted that financial distress in state budgets
often results in tuition levels and policy being “backed into” as well as regional
differences in the state policies. The northeastern and upper Midwestern regions of the
country performed greater “rationalization” tuition than southern and southwestern states.
The setting of tuition levels, in particular in-state tuition, occurs inside of a political
context. As shown in Table 1, more recently the percentage increases in tuition and fees
at public four-year institutions exceed those at private four-year schools.
How do the students and their families react to such tuition increases? During the
environment of heated debate about financing higher education, Hearn & Longanecker
(1985) conducted a study of the enrollment affects of pricing policies. Not surprisingly,
they found that students responded to tuition prices differently depending on the incomes
of their families, with the demand elasticity the greatest (i.e. the sharpest drop in
enrollment when tuition is raised) for low income students. Leslie and Brinkman (1987)
39
Table 1
Average Higher Education Tuition & Fees
(Enrollment Weighted – Constant 2007 dollars)
School Year
Private
Four-Year % Chg
Public
4-Year % Chg
Public
2-Year % Chg
1977-78
$9,172
$2,225
$1,040
1982-83
$9,872
8%
$2,194 – 1%
$1,007 –3%
1987-88
$12,808
30%
$2,699
23%
$1,343 33%
1992-93
$15,416
20%
$3,444
28%
$1,647 23%
1997-98
$17,823
16%
$4,022
17%
$2,026 23%
2002-03
$20,778
17%
$4,715
17%
$1,926 –5%
2007-08
$23,712
31%
$6,185
51%
$2,361 41%
(Source, US College Board, Trends in Higher Education Series, 2007)
reviewed the results of 25 studies between 1967 and 1985 to create an average student
response to increases in tuition, finding that on average student demand drops 0.6 %
when tuition is increased $100. Ten years later Heller (1997) replicated this finding with
results varying between 0.5 and 1.0 percent reductions in enrollment given a $100 rise in
tuition. Heller’s study went further and presented five key findings of critical importance
to the “high tuition, high aid” policy. First, regarding a student response to an increase in
tution, in 1983 dollars, a $100 increase produced a 0.7% drop in enrollment. Second,
student response to a change in financial aid, while more complex because it varies on the
type of aid involved, was weaker than that of a tuition increase. Importantly, even with a
change in aid to neutralize an increase in tuition, keep net price constant, the increase in
the sticker price of tuition caused “sticker shock” and a drop in enrollment. As one
would expect, loans and work study had an even weaker effect than grants (Heller, 1997),
a fact exacerbated by the federal policy of increasing the share of federal financial aid to
40
loans (Hearn, 1998). Third, students from lower income families are more sensitive to
price increases. Only at the higher SES levels does sensitivity to price disappear.
Poorest students react strongest to grants, even stronger than tuition (likely that they
know they will qualify). Fourth, rising tuition in higher education is producing the
“trickle down” mentioned earlier, where meritorious students select schools at lower
levels of selectivity because of price. Many of the poorest students are giving up on
higher education entirely and obtaining unskilled work in the labor market. Naturally
these are aggregate trends with exceptions existing at all levels. Ethnicity and race do
play a role, but one must be cautious of the interplay between these and income. Family
income is the most influential variable (Heller, 1997). Finally, he finds variation across
institutional type with significantly stronger student response to price increases at
community colleges. Price increases are experienced as a portion of the total cost. A
$100 increase at an institution charging $20,000 annual tuition is not the same as a $100
increase at an institution charging $1,000 (Heller, 1997). Summarizing, the model
demoralizes and discourages participation of the lower income students.
Using the 1995-1996 and 2003-2004 National Postsecondary Student Aid Surveys
(NPSAS), Wellman, Desrochers, and Lenihan (2008) found that among dependent
students, participation in higher education of those from families earning less $20,000 fell
from 14.5% to 12.8%, and the impact is not limited to the lower income families. Over
the same period the enrollment of graduating high school seniors from families earning
less than $80,000 fell from 66.8% to only 49.1% (Wellman, Desrochers, and Lenihan,
2008). Gerald & Haycock (2007) found that the portion of Pell grant recipients at more
41
prestigious flagship public schools among all Pell grant recipients fell from 0.83 in 1992
to 0.63 in 2003, with even more dramatic declines for students with minority
backgrounds (Gerald & Haycock, 2007).
Intuition readily anticipates that non-resident (out-of-state) student response
would differ from that of in-state students. Zhang (2007) specifically examined nonresident demand at public universities. Conducting analyses at the national, state, and
institutional level, he found that at the national level, a one percent increase in nonresident tuition reduced non-resident enrollment by 0.90 percent. However, at the state
level no such relationship occurred, and at the institutional level the elasticity was far
smaller, where a percent increase in non-resident tuition only reduced non-resident
enrollment by 0.20 percent. He also found the interesting result that an increase in the
resident (in-state) tuition increased non-resident enrollment, suggesting that in-state
tuition may serve as a proxy for prestige, a result supported when he disaggregated the
schools by type and selectivity. For the Research I Doctoral universities or highly
competitive schools, non-resident student demand exhibited virtually no elasticity with
enrollment falling only 0.04 percent for the former and 0.11 percent for the latter when
non-resident tuition was increased by one percent.
The generation of funding streams other than tuition is relevant to this study. The
National Association of State Universities and Land-Grant Colleges, analyzing IPEDS
revenue data, compared the total revenue percentages at public four-year institutions
between the 1990/91 academic year and the 2000/01 academic year, dividing total
revenue into three sources: 1) state appropriations, 2) tuition, and 3) other revenue. State
42
appropriations fell from 36.1% of total revenues to 29.8%. While tuition revenue
increased, the rise was significantly smaller than the drop in state appropriations,
climbing from 15.3% to 17.7%, an increase of only 2.4% in the face of a 6.3% drop in
state appropriations. The “other revenue” stream increased 3.9% from 48.6% to 52.5%.
I want to underscore the point that public four year institutions now receive over half of
their total revenue from sources other than state appropriations and tuition.
Political factors appear to play a role in the setting of in-state tuition prices.
Studying panel data for 162 public flagship universities over the period from 1984 to
2002, McLendon, Hearn, and Hammond (2008) found that higher levels of minority
representation in state legislatures were associated with lower levels of tuition. They also
confirmed the findings of Koshal and Koshal (2000) that increases in state appropriation
levels decelerated increases in resident tuition. Regarding out-of-state enrollment and
tuition prices, they found that schools enrolling larger shares of out-of-state students
charged higher in-state tuition levels, a finding consistent with Zhang’s study on
nonresident enrollment demand. Strictly focused on nonresident students, Zhang found
an increase of 1% in resident tuition is associated with a 0.88% increase in nonresident
enrollment (Zhang, 2007). In particular for out-of-state students, tuition appears to serve
as a proxy for institutional quality (Zhang, 2005). As institutions become more
prestigious, out-of-state and international student demand rises, and universities charge
these students a substantial premium over the in-state tuition level charged to state
residents. As state-appropriations fall, the political mission of serving state residents can
be expected to have a suppressing effect on using in-state tuition to make up the
43
difference. In the context of the politics of state legislatures governing university tuition
prices, one would expect in-state tuition to react more sluggishly than out-of-state tuition
prices in efforts to increase revenue.
The downward political pressure on the levels of in-state tuition at public
universities does not apply to out-of-state tuition for the simple reason that it does not
apply to residential students and their voting parents. The distinction between the in-state
student and out-of-state student is important to this study. The setting of in-state tuition
levels has been shown to be a function of the political environment and the university
campus’s engagement in the community. The highly inelastic out-of-state student
demand found by Zhang suggests that such students are a ripe source of net tuition
revenue provided the institution has the necessary level of prestige and reputation. The
balance between in-state tuition and enrollment and out-of-state tuition and enrollment
has become important for public universities, and many now have departments dedicated
to a relatively new and increasingly important function in higher education, enrollment
management.
Enrollment Management
During the period of rapid expansion of participation in higher education that
followed the Higher Education Act of 1965, a natural stratification of institutions
occurred as demand far outstripped supply. The prestigious Ivy League and other elite
private schools, bombarded with many more applicants than they could admit, became
highly selective, and an “axis of prestige” developed with expensive and highly selective
private institutions at the top and inexpensive community colleges at the bottom. At the
44
same time that the Higher Education Act of 1972 encouraged the competition of
institutions for students now free to apply their financial aid to the school of their choice,
competition amongst students for admission into the best college possible also flourished.
Barron’s and other publications started publishing books categorizing schools by
selectivity, with the measure serving as a universally understood proxy for status,
prestige, the quality of the education, and the prospects of the students upon graduation.
In 1993, the magazine US News & World Report started publishing detailed rankings of
institutions, a publication devoured by high school students and their parents as they
reached college age. In a matter of less than twenty years, an institution’s place in the
prestige hierarchy began to significantly affect its operating environment in terms of
applications received, percentages of admitted students choosing to enroll, and ability to
recruit top faculty (Ehrenberg, 2001). The formulas used by the magazine to rank the
institutions rewarded selectivity and the costs expended per student with profound
implications. Summarizing, the increased stratification of institutions and the
introduction of competition on the part of both suppliers and customers complicated the
“market” for higher education. The need to manage this growing complexity transformed
university admissions departments from the rather simple gate keeping function to a
diversified set of operations including marketing, recruitment, financial aid strategies,
retention, and others fitting into a new set of functions called enrollment management.
Higher education pricing and financial aid now occur in the context of enrollment
management, an increasingly sophisticated system that addresses “the student flow”
(Kroc & Hansen, 2005) from initial contact with students as early as their junior (or even
45
sophomore) years in high school through recruiting, enrollment, retention, graduation,
placement, and relations with alumni (Hossler & Anderson, 2004). Enrollment
management recognizes that just as students compete for admission into selective
schools, institutions must compete for desirable students, where desirable usually refers
to academic talent but can also apply to the goals of an appropriate balance of gender,
ethnicity, socioeconomic status, or other factors such as region of the country where the
student was raised. Enrollment management departments develop expensive marketing
programs as elaborate as those created by Fortune 500 companies. Recruiting operations
seek to maximize the pool of applications to provide the institutions the greatest possible
choice of students to admit. Sophisticated computer models based on a growing body of
empirical institutional research provide detailed simulations that forecast enrollment
volume, statistics regarding student body composition, and expected revenue.
Enrollment management is the natural product of the shift of higher education
towards a competitive environment on the part of both producer and consumer.
Addressing the competitive environment developing among institutions seeking to enroll
desirable students, Matthew Quick wrote of the “taste for blood” and the concept of
“eating another institution for lunch” to refer to one school’s stealing attractive prospects
from another school (Quick, 2006). Others (Hossler, 2004; Duffy & Goldberg, 1998;
Jaschik, 2006) note the concerns associated with the use of financial aid to “buy students”
and the possibility that it draws both students and institutions are drawn away from the
true mission of higher education. Clearly expressing anguish, in a letter to the New York
Times titled, “To All the Girls I’ve Rejected,” Jennifer Britz of Kenyon University fully
46
acknowledged that she rejected female applicants with credentials superior to male
applicants who were accepted (Britz, 2006). In a study examining how a public
university optimizing its admissions might respond to a shock in its funding, Nagler
(2008) distinguished how a school could admit affluent students known as “inferior
goods” (as judged by academic merit) but still maximizing the school’s overall interest
because of their ability to pay significantly higher tuition and fees.
Going further than the modest enhancement of financial aid packages to “tip”
student enrollment yields is the aggressive use of tuition discounts specifically designed
to “buy” certain students. This is distinct from the practice of using discounts to fill
empty classroom seats. The policy of filling vacant seat capacity with discounted prices
has been practiced successfully by the airline industry for decades, and when an airline
fills an otherwise empty chair on a plane, revenue is increased since the cost of the
additional passenger is negligible. However, when a discount is used to purchase one
student with desired characteristics over another student, the result can lead to institutions
undercutting each other to where losses can occur. Kenneth Redd (2000) attacked the use
of tuition discounting to obtain desirable students as leading to decreases in net tuition
revenue to where the schools incur operating losses (Redd, 2000). While institutions may
restrain themselves from such deep discounting, they do face trade-offs with respect to
merit based institutional aid, in particular when it displaces funds that would have gone to
students on basis of need. Ehrenberg, Zhang, and Levin (2006) studied the relationship
between institution awarded National Merit Scholarships and lower-income student
enrollment as quantified by those qualifying for Pell Grants. They found that on average
47
an increase of ten students receiving the merit awards was associated with a decrease of
four students receiving the Pell Grants (Ehrenberg, Zhang, & Levin, 2006).
Related Studies
This section presents the results of some other studies relevant to the setting of
tuition levels, financial aid, and the generation of revenue from sources other than tuition.
Public university tuition prices and financial aid programs can vary widely from state to
state. The higher education governance structures adopted by state legislatures can
influence tuition prices. Some regions of the country, having greater densities of private
colleges and universities, may have different political landscapes and views of what is
acceptable regarding tuition and financial aid. I also explore the efforts of public
universities to solicit private donations with fundraising programs, including the
increasingly prevalent practice of building endowments.
The most centralized governing boards, those with consolidated authority over
public postsecondary institutions, have shown greater adoption of tuition
“rationalization” (raising tuition to levels more closely matching the actual costs of the
education) while states with coordinating boards lacking budget authority were less
aggressive in raising tuition (Hearn, Griswold, and Marine, 1996). However, the
significance of differences between strong and weak governing boards is declining
because most states have abandoned the weak governance structure. In their study of the
influence of governance structures on the management and performance of higher
education institutions, Knott and Payne (2004) discovered that since 1990, eleven states
had shifted to stronger governing systems, bringing the total to 37 states with strongly
48
regulated boards, with six states having moderately regulated systems and only five states
with minimally regulated structures.
Koshal & Koshal (2000) directly examined the relationship between state
appropriation levels and tuition by analyzing a cross section of 1990 data for four-year
institutions in 47 states. They found tuition to be a function of state appropriations,
median family income of the state, the percentage of out-of-state students at the school,
and the region of the country. The finding regarding the region of the country is
consistent with the findings of increased tuition rationalization at the northeast and
northern Midwest institutions, which have increased tuition substantially more than
schools in the south and southwestern regions of the country (Hearn, Griswold, and
Marine, 1996).
In addition to setting the level of state appropriations provided to public higher
education institutions, the states also set financial aid policies for the assistance provided
to students. State policies shape the extent to which state resources are provided directly
to institutions or directly to students, and they also influence the criteria for determining
the financial aid packages students receive. These policies differ widely from state to
state (Doyle, Delaney, and Naughton, 2008). As state policies have shifted from a need
based focus on financial aid towards a merit based approach (Heller, 1997), what has
occurred with the institutional aid at their public colleges and universities? Anticipating
that perhaps the schools might use institutional aid to compensate for shifts in state aid to
students, Doyle, Delaney, and Naughton (2008) specifically examined the relationship of
the institutional aid packages awarded to students and the financial aid policies of the
49
states. They found that instead of compensating for shortfalls or surges due to changes in
state programs, the institutions tended to reinforce or “comply” with the state’s emphasis
(Doyle, Delaney, and Naughton, 2008). Such practices by the institutions only amplify
the reallocation of financial aid resources from a need based orientation towards one that
is merit based.
One obvious choice for enhancing revenue besides increasing tuition is to pursue
fundraising efforts with alumni or other community organizations. Many public
universities now pursue endowments that can become as significant as those possessed by
some private universities. Special projects bearing the name of a famous alumnus, a prior
president, or other dignitary can solicit substantial donations. Cheslock and Gianneschi
(2007) examined the ability of institutions to generate revenue with alumni fundraising
and other voluntary gift programs. Their results reinforced the importance of the prestige
hierarchy (as measured by selectivity), where the most selective institutions received
approximately 1.8 times the gift revenue received by moderately selective schools and
over 3 times the amount given to less selective institutions (Cheslock & Gianneschi,
2008). In fact, the degree of inequality along the stratification of selectivity is greater for
fund raising than for state appropriations. While a nationally recognized flagship like the
University Wisconsin at Madison successfully raised $1.5 billion in a fundraising
campaign (University of Wisconsin Foundation, 2004), lesser known schools lack the
name recognition or alumni populations to support such efforts. Examining four-year
public institutions in 2004, Cheslock & Gianneschi (2008) found that a $1000 loss in
state appropriations (per full time equivalent student) is associated with a $45 reduction
50
in private donations to the school. At the minimum, this casts considerable doubt on the
likelihood of public universities replacing state appropriations with voluntary donations
or fundraising programs.
Pertinent Theoretical Frameworks
Several theoretical frameworks are relevant to this study and the interpretation of
its findings. The basics of economics and business administration tell us that higher
education institutions must balance their budgets over time if they wish to survive. While
the complex dynamics of setting tuition prices and determining financial aid packages for
students produce the complicated set a relationships this study investigates, one simple
fact underlies the details. As state appropriations decline as a share of the budget
required to fund higher education, either other sources must fill the gap or the size of the
budget must be reduced. That noted, the facts clearly show that the former (finding
additional funding) is far more prevalent than the latter (reducing budgets). I have
already discussed the two competing and most often discussed theories regarding the
costs of higher education, the “cost disease” presented by Baumol and Blackman (1978)
and Bowen’s (1980) revenue theory of cost. When examining the relationship between
state appropriations and both list and net tuition prices experienced by students, the
former theory would anticipate a negative relationship, with the level of state
appropriations helping to alleviate the need for higher tuition prices and revenues.
Bowen’s revenue theory offers a distinctly different perspective. If institutions
aggressively seek to maximize revenue via all methods each optimized in its own right,
then universities are already setting in-state and out-of-state pricing and packages to
51
generate the largest revenue possible. In this scenario, little or no relationship
whatsoever could exist between the state appropriations made available to an institution
and its pricing and financial aid policies.
Looking at higher education institutions from the lens of an economist, Garvin
(1980) described the university as a utility maximizing organization that seeks to increase
its prestige subject to a resource constraint. Not a corporation focused solely on profits,
public universities seek to improve their standing on the hierarchy of status and
reputation for excellence. Financial performance remains imperative, but does so as a
means to improve the school’s position compared to its competition for both students and
faculty. Noting the exception of communities served by a single community college with
few or no comparable alternatives in proximity, most higher education institutions
compete for enrollment and prestige, but they do not all compete with each other. Rather,
schools compete only with similar institutions in terms of status, geographical location,
and selectivity. Nagler (2008) examined the implications of an institution that optimizes
its perceived utility function on its response to funding shocks they may experience as the
result of a recession or other economic calamity. What does a funding shock predict
when viewed from the perspective of maximizing a function (prestige) other than profit
alone? To accomplish this Nager distinguished students between those that provide
prestige and those that provide revenue, since both must be considered. In such a
situation, he produced a model that drew the intuitively reasonable conclusion that
schools may admit “inferior goods” (referring to their academic credentials) should their
52
ability to pay out-of-state or full in-state tuition serve to sufficiently increase net tuition
revenues in a response to the shock (Nager, 2008).
Institutional Theory (Dimaggio & Powell, 1983) suggests that institutions of
similar type (selectivity, governance) will develop highly similar practices. Their
discussion of normative and mimetic institutional isomorphism provides sound
explanations for the rise of relatively consistent practices and policies across institutions
of similar type. As goes Harvard one can expect Stanford, Princeton, and other top
privates to follow or at least seriously consider. The theory applies to public flagship
institutions and community colleges as well. Such a view would predict minor variations
in their response to changes in state appropriation levels per student on their pricing and
aid practices. Institutional isomorphism tells us to expect the pricing and financial aid
policies of similar institutions to exhibit similar behavior (Dimaggio & Powell, 1983).
One could infer from this institutional theory that, for example, highly selective
institutions respond in one manner while non-selective institutions react differently.
Different behavior might also be found between schools located in different geographic
regions, with findings for those in the northeast New England region different, perhaps
substantially, from those in the South or Southwest. I also anticipate findings to differ
considerably between two-year and four-year institutions.
To the extent institutions are already maximizing prestige, they have tapped all
resources they have identified, and a change in appropriations would have little impact on
tuition. In this situation, tuition levels are influenced predominately priorities and factors
(politics, student demand, etc.) that have little or nothing to do with state appropriation
53
levels. Basic economic theory would anticipate a negative coefficient for state
appropriations per student when regressing listed tuition price, net tuition paid by
students, or the level of unmet need. A drop in state appropriations, ceteris paribus, can
be expected to associate with higher levels of list and net tuition and unmet need. A drop
in state appropriations per student, ceteris paribus would also be associated with lower
levels of institutional aid, since the smaller amount of state support reduces an
institution’s ability to provide such grants.
Conclusion
As expressed by Art Hauptman (2001), the key financial components of public
higher education are the state appropriations provided to institutions, the tuition levels set
by those institutions, and the financial aid mechanisms in place to help aspiring students
acquire the resources to pay the price. I have presented the literature discussing the
declining levels of state appropriations as a portion of higher education budgets and the
factors behind the decline. I have also discussed the rising tuition prices that have
occurred in the last three decades and the shifting policies in financial aid during the
same period. The economic conditions currently facing the United States suggest that all
of these trends will continue. The questions of this study do not ask whether state
appropriations will continue to decline or at what rate. Instead it seeks to clarify what
data from the past suggests we can expect to happen to the in-state and out-of-state
tuition levels of an institution given a particular change in its state appropriation level of
support. More significantly, the study investigates how a change in state appropriations
54
translates to the tuition price and financial aid experienced by the individual students as
they attempt to enroll at a public school.
As demonstrated in this section, very little research has addressed this question.
Furthermore, existing theories provide conflicting predictions, so we can not expect a
clear answer on theoretical grounds. It is clearly an empirical question, and the next
chapter will present the empirical strategy that will be utilized to answer it.
55
CHAPTER THREE
DATA AND METHODOLOGY
Introduction
This chapter begins with the research questions that address the fundamental
problem of the study, the relationship between state appropriations and higher education
pricing and financial aid policies. After presenting the research questions I discuss the
methodology of the study in detail, which includes the data sources, key variables, and
the equations and econometric models used to conduct the analysis. After a thorough
treatment of the methodology, I provide additional details on the data sources and the
variables. I conclude the chapter by explaining the specific sample populations selected
from the data sets and why those samples were chosen.
Research Questions
To clarify the relationship between state appropriations, tuition levels, and
financial aid practices at public higher education institutions, I explore the following
fundamental questions:
1. Given a cross section of institutional data from four-year and two-year public
colleges and universities, how does the list price of tuition vary as state
appropriations per student in public funding changes?
a. Does the relationship between tuition and state appropriations change over
the five years examined for the 1989/90 to the 2003/04 period?
b. For four-year institutions, does this relationship vary between more
selective, moderately selective, and less selective schools?
56
c. What is the difference in the relationships between state appropriations
and in-state tuition as opposed to out-of-state tuition?
d. What happens to these relationships when the five year data sets are
constructed as panel data to control for institutional fixed effects?
2. When examining student-level data from a cross section of college students
throughout the country, how do the list tuition, net tuition, institutional grant, and
their unmet need vary as the state appropriation level per student changes?
a. Does the relationship between tuition and state appropriations change over
the five years examined for the 1989/90 to the 2003/04 period?
b. For four year institutions, does this relationship vary between more
selective, moderately selective, and less selective schools?
c. How is the analysis different for students with in-state status from that of
students with out-of-state status?
d. What happens to the relationship between state appropriations and the
items listed in question two when the student characteristics (ethnicity,
gender, academic merit, etc. to be discussed in detail later) are taken into
account, i.e. controlled for?
57
Methodology
Overview
This study combines the analysis of institutional level data for the public higher
education institutions in the United States and student level data for the students
attending those same institutions. For each of five separate academic years spanning the
period from 1989/90 to 2003/04, I examine data for higher education institutions across
the country from the Integrated Postsecondary Educational Data System (from now on
referred to as IPEDS) used and maintained by the United States Department of
Education. All higher education institutions in the United States are required to have
their admissions or enrollment management departments collect and maintain data
regarding numerous student information including enrollment levels, pricing, financial
aid, demographics, and many other items. Institutions must report statistics into the
IPEDS system, and detailed reports and data sets can be downloaded from the IPEDS
Peer Analysis tool available on the Internet. Independently, for the same years I examine
data at the student level from the National Postsecondary Student Aid Study (from now
on referred to as NPSAS) also development and maintained by the United States
Department of Education. The NPSAS data are collected with carefully designed surveys
of a representative sample of students attending college throughout the United States.
The five academic years selected for this study (1989/90, 1992/93, 1995/96, 1999/2000,
and 2003/04) are those in which the NPSAS studies were conducted. I will present more
about both the IPEDS data and the NPSAS data in the section on Data Sources.
58
The institutional level and student level analyses share data. For each, the same
data and values are used for institutional enrollment and total state appropriations. These
two numbers play the important role in the study by determining the level of state
appropriations an institution receives on a per full time equivalent student basis. At the
institutional level, the tuition is the value reported by the school to the IPEDS system. At
the student level, the tuition is that reported for the student in the NPSAS data. The
student level analysis allows for many additional variables associated with the student
including ethnicity, gender, cumulative grade point average, and expected family
contribution. The student level regressions control for these factors in determining the
relationship between state appropriations and tuition and fees.
Throughout the study, I separate the groups being analyzed into four distinct
attendance scenarios consisting of 1) in-state attendance at a four-year school, 2) out-ofstate attendance at a four-year school, 3) in-state attendance at a two-year school, and 4)
out-of-state attendance at a two-year school. For each of the four attendance situations, I
examine the descriptive statistics of selected variables over the five years to identify
trends that may be occurring. After examining trends in the descriptive statistics, I
estimate ordinary least squares regressions for each individual year. This is done for the
same five years for each group for both the individual and student level data sets. To
prevent inflation from impacting the relationships under investigation, I convert all
financial variable values to 2004 dollars using the consumer price index published by the
US Department of Labor Bureau of Labor Statistics.
59
Institution Level Analysis
For both four year and two year institutions, I separately run ordinary least
squares linear regressions for the in-state and out-of-state listed prices for tuition and fees
as reported by the institution to the IPEDS data system. I control for three factors known
to influence tuition pricing at public institutions, their selectivity, the institution’s
governance system, and which region of the country in which the institution resides. The
econometric model consists of the following equations:
In-State Tuition = β0 + β1(SA) + β2(G1) + β3(G2) + β4(S1) + β5(S2) + β6(S3) +
β7(S4) + β8(NE) + β9(ME) + β10(GL) + β11(SE) + β12(SW) +
β13(R) + Err
Out-of-State Tuition = β0 + β1(SA) + β2(G1) + β3(G2) + β4(S1) + β5(S2) + β6(S3)
+ β7(S4) + β8(NE) + β9(ME) + β10(GL) + β11(SE) + + β12(SW) +
β13(R) + Err
SA = Total unrestricted state appropriations per FTE student
G1 = Dummy variable for strong governance
G2 = Dummy variable for moderate governance
S1 = Dummy variable for most selective
S2 = Dummy variable for more selective
S3 = Dummy variable for selective
S4 = Dummy variable for less selective
NE = Dummy variable for New England Region
ME = Dummy variable for Mid-East Region
GL = Dummy variable for Great Lakes Region
PL = Dummy variable for Plains Region
SE = Dummy variable for Southeast Region
SW = Dummy variable for Southwest Region
R = Dummy variable for Rocky Mountain Region
Err = Error Term
The selectivity of a school serves as a powerful proxy for its status and prestige
and is clearly a factor in the institution’s tuition levels and its financial aid practices
60
(Ehrenberg, 2002). The model relies on the selectivity ranking published by the 2005
Barron’s selectivity classifications. Barron’s produces a selectivity measure for
institutions based upon its admission rate and the average test scores and high school rank
of the students that enroll. It then groups the schools into nine categories ranging from
“Most Selective” through “Non-selective.” The Barron’s groupings occur in much
smaller sizes as they become more selective. To obtain groupings more suitable for this
study (less disparity in sample sizes) I create five selectivity categories that combine
some of the more selective Barron’s categories (this is shown in the variables section of
this chapter) to produce the following groups 1) most selective, 2) more selective, 3)
selective, 4) less selective, and 5) non-selective. For all descriptive statistics and
regressions, an institution’s selectivity is addressed with five dummy variables, one for
each selectivity category. As the regressions require the omission of a category, I omit
the “non-selective” dummy variable from the regressions. While the statistics retain the
variables above, when I split the institution sample into groups by selectivity, I further
reduce the grouping to just three categories to obtain even greater equality in the sample
sizes and simplify the analysis. The three groups used to run separate analyses are more
selective (the most and more selective categories combined), middle selective (the
selective category unchanged) and less selective (the less and non-selective categories
combined). This applies simply to the groupings of institutions on which the analysis is
performed. Selectivity is not a factor for two-year institutions, so it is not part of the
model for two-year schools.
61
Studies have shown (Hearn & Griswold, 1994; Griswold & Marine, 1996; Knott
& Payne, 2004) that the governance structure of the state influences the degree to they
pursue policy innovations, including tuition rationalization policies. For this reason, I
control for governance structure as well using dummy variables. The governance
structures for the study are utilize those published by Knott & Payne (2004) and consist
of 1) strongly regulated, 2) moderately regulated, and 3 ) weakly regulated. In their 2004
study, Knott and Payne started with the classification scheme of higher education
governance systems from the State Postsecondary Education Structures Handbook,
(1991, 1994, 1997) that distinguished board structures as consolidated board,
coordinating boards, or planning agencies. Recognizing that coordinating boards could
exercise different degrees of regulatory authority and seeking a classification to better
measure the degree of regulation, they created three alternate classifications: highly
regulated (governing board or coordinating board with strong regulation powers),
moderately regulated (coordinating boards with some regulation), and weakly regulated
(coordinating boards with weak regulation power, advisory boards, or planning agency).
I use these classifications, and the states fitting each are listed in the variables section of
these chapter. I create dummy variables for each group, and for the regressions, the
omitted variable is the one indicating “weakly regulated.”
In addition to selectivity and governance structure, the geographic region of a
state (and embedded within it, the dominant political sentiments of the state legislatures
for that state) has been shown to influence the pricing and financial aid policies of its
public higher education institutions (Hearn, Griswold, & Marine, 1996). For the same
62
reasons described above, I control for the region of the country using dummy variables
for each region and setting the value to one if the school is in that region. The regions
match those used in the IPEDS data and consist of: 1) New England, 2) Mid-East, 3)
Great Lakes, 4 ) Southeast, 5) Plains, 6) Rocky Mountains, 7) Southwest, and 8 ) Far
West. The specific states in each region are listed in the variables section of this chapter.
For the regressions, the omitted dummy variable is the one for “Far West.”
The majority of institutions are present for all of the years, providing a panel data
set on which fixed effects regression can be performed. The fixed effects regression will
refine the analysis by controlling for institutional level factors that remain unchanged
across the five years that are studied. Also, the fixed effects model introduces dummy
variables to indicate the specific year to which the data belong. The coefficients of these
dummy variables will provide robust results on the nature of tuition increases over time.
Student Level Analysis
The student level analysis follows the same methodology as the institutional level
analysis but includes considerably more variables that are available for the students in the
data, namely the characteristics of ethnicity, gender, socioeconomic status (as represented
by the expected family contribution) and academic preparation or merit (as represented
by the cumulative grade point average of the student at the institution). The student level
study uses the same data and values for the institution’s state appropriations per student,
governance, and region of the country. For all other data, the student level analysis
utilizes the NPSAS variables attached to each student, including the tuition price. The
student level analysis allows the study to examine the student specific information
63
regarding the net tuition paid, the institutional grant the student received, and the
student’s level of unmet need. It also allows the model to control for student
characteristics including ethnicity and gender, and importantly, the student’s ability to
pay. As discussed in chapter two, the ability to pay is calculated from the information
collected by the standardized Free Application for Federal Student Aid (FAFSA) and
known as the Expected Family Contribution (EFC). The student’s ability to pay is
clearly an important variable in the analysis of the financial aid awarded. The
econometric model consists of the following equations:
LIST TUITION = β0 + β1(SA) + β2(G1) + β3(G2) + β4(S1) + β5(S2) + β6(S3) +
β7(S4) + β8(NE) + β10(ME) + β11(GL) + β12(P) + β13(SE)
+ β14(SW) + β15(R) + β16(W) + β17(AA) + β18(H) + β19(A)
+ β20(F) + β21(GPA) + β22(EFC) + Err
NET TUITION = β0 + β1(SA) + β2(G1) + β3(G2) + β4(S1) + β5(S2) + β6(S3) +
β7(S4) + β8(NE) + β10(ME) + β11(GL) + β12(P) + β13(SE)
+ β14(SW) + β15(R) + β16(W) + β17(AA) + β18(H) + β19(A)
+ β20(F) + β21(GPA) + β22(EFC) + Err
INST GRANT = β0 + β1(SA) + β2(G1) + β3(G2) + β4(S1) + β5(S2) + β6(S3) +
β7(S4) + β8(NE) + β10(ME) + β11(GL) + β12(P) + β13(SE)
+ β14(SW) + β15(R) + β16(W) + β17(AA) + β18(H) + β19(A)
+ β20(F) + β21(GPA) + β22(EFC) + Err
UNMET NEED = β0 + β1(SA) + β2(G1) + β3(G2) + β4(S1) + β5(S2) + β6(S3) +
β7(S4) + β8(NE) + β10(ME) + β11(GL) + β12(P) + β13(SE)
+ β14(SW) + β15(R) + β16(W) + β17(AA) + β18(H) + β19(A)
+ β20(F) + β21(GPA) + β22(EFC) + Err
The Dependent Variables:
List Tuition – the reported list tuition and fees prior to the application of financial aid
64
Net Tuition – the reported net tuition paid by the student after receiving financial aid
(NOT including loans).
Institutional Grant – the reported institutional grant provided to the student
(does not have to be repaid by the student).
Unmet Need – the calculated difference between the student’s ability to pay (as
Measured by the expected family contribution (EFC) and the total charges
incurred by the student after receiving financial aid that is NOT a loan.
The Independent Variables:
SA – the calculated state appropriations per full time equivalent student received by the
institution.
G1 –
G2 –
S1 –
S2 –
S3 –
S4 –
NE –
ME –
GL –
P–
SE –
SW–
R–
AA –
H–
A–
F–
Strongly regulated governance structure
Moderately regulated governance structure
Most selective institution category
More selective institution category
Selective institution category
Less selective institution category
New England region
Mid East region
Great Lakes region
Plains region
Southeast region
Southwest region
Rocky Mountain region
African American
Hispanic
Asian
Female
GPA – The student’s cumulative grade point average at the NPSAS institution
EFC – The student’s reported expected family contribution
Err – Error Term
The purpose of the econometric model is to quantify the relationship between
each of the dependent variables and the state appropriations per student received by the
institution. The other variables are included as control variables to account for other
factors that may be influencing the values of the dependent variables.
65
The NPSAS survey incorporates weighting and sample strata to insure that
minority populations are sampled in sufficient quantity to obtain meaningful results. This
weighting can influence both the descriptive statistics and the results of the regression, so
I run the analysis both with and without the incorporation of the sample design. I run the
analysis first without any weighting or sample design, then with the weighting variable
alone, and finally with the full sample design taken into consideration. For all
regressions, the analysis is run on those students for which the data is complete, resulting
in slightly different sample sizes.
Data Sources
Institutional Level Analysis
For the institutional level data, this study uses data from the United States
Department of Education’s National Center for Education Statistics Integrated
Postsecondary Education Data System (IPEDS) for the schools years 1989-1990, 19921993, 1995-1996, 1999-2000, and 2003-2004. Federal law requires all higher education
institutions to report detailed data regarding their operations including student enrollment
numbers for full and part-time students at each grade level. In addition to enrollment
data, schools must report detailed financial information that include prices for tuition,
fees, room and board, estimates for other expenses and with the detail necessary to
distinguish in-district, in-state, and out-of-state pricing. Importantly, all institutions must
report the amount of state appropriations they receive each year, a value fundamental to
this study. IPEDS data are available to the public, and for this study the pertinent data
sets were downloaded as spreadsheets from the IPEDS Peer Analysis System Web site
66
hosted by the National Center for Education Statistics. The specific data sets downloaded
are listed in detail in Appendix A. After the spreadsheets were stored on the local hard
drive, the statistical software STATA 10.0 was utilized to perform the analyses.
The IPEDS data set does not include the selectivity of an institution. Throughout
the study, for both the institutional level and student level analyses, four-year institutions
were assigned a selectivity value based upon their designation in the 2005 Barron’s
Profiles of American Colleges.
As noted in the previous section, for the level of regulation exerted by the governing
boards of the states in which the institutions reside, it was also necessary to use another source. I
use the categories as published by Knott & Payne (2004) that classifies states as 1)
minimally regulated, 2) moderately regulated, and 3) highly regulated. As I discussed in
the literature review, the state governance system was found to be significant in several
studies on tuition rationalization. Since so many states have adopted stronger governance
arrangements in recent years (Knott & Payne, 2004), I anticipate a decline in variation
explained by governance structure but consider its inclusion desirable for consistency
with other studies and confirmation of the decline. The list of states in each category is in
the variables section of this chapter.
Student Level Analysis
For the student level analysis, I use the United States Department of Education’s
National Center for Education Statistics National Postsecondary Student Aid Study
(NPSAS) for the same years as above, 1989-1990, 1992-1993, 1995-1996, 1999-2000,
and 2003-2004. The NPSAS study is designed to select a nationally representative
67
sample of postsecondary education institutions. NPSAS obtains and fortifies its data by
using multiple sources including national databases, computer assisted telephone methods
(CATI), institutional records, and student interviews.
The student level study utilizes the same measures for institution selectivity and
state governance structure as that described for the institutional level.
Variables
Variables in both the Institution and the Student Level Analyses
The level of state appropriations received by an institution plays a central and
critical role in the entire study and warrants a detailed explanation. The value is
produced entirely with data from the IPEDS system. I directly read the value of total
state appropriations received by each institution from the IPEDS financial data set that
contains this variable for each institution. To calculate the state appropriations received
per full time equivalent student, it is necessary to produce a variable that expresses the
total enrollment of full time equivalent students. The total enrollment at the school is
calculated to full time equivalent students using the following formula:
Total FTE students = (total full time undergraduate students) +
0.333 x (total part time undergraduate students) +
2 x (total full time graduate students +
0.666 (total part time graduate students)
The formula has been used by Cheslock and Giannaschi (2008) and in other
studies (Bowen, 1980; Brinkman, 1990) to account for the additional expenses associated
with graduate students and reduced costs for part time students. The total appropriations
received at the institutions are divided by the total full time equivalent enrollment to
generate the variable used in the regression, state appropriations received per student.
68
State appropriations per student = (Total state appropriations)
(Total calculated FTE)
The selectivity dummy variables are generated from the classifications assigned to
the schools in the 2005 Barron’s publication. The Barron’s publication places greater
ranking sensitivity of its grouping for the more selective schools (it splits the top 7% of
schools into three groups while the bottom 12% are one group). To produce groups
better suited to this study, I collapse the more selective Barron’s groups as shown below.
The percentage specified in parentheses indicates the portion of schools that fall into the
category.
Selectivity Dummy Variables:
Select1 = Most competitive + Highly competitive plus + Highly competitive (7%)
Select2 = Very competitive plus + Very competitive + competitive plus (20%)
Select3 = Competitive (43%)
Select4 = Less competitive (18%)
Select5 = Non competitive (12%)
For the governance classifications, three dummy variables are created.
G1 – Highly Regulated
Alabama, Arizona, Arkansas, Connecticut, Colorado, Indiana, Florida, Georgia, Idaho,
Illinois, Iowa, Kansas, Kentucky, Louisiana, Maine, Maryland, Massachusetts,
Mississippi, Missouri, Montana, Nevada, New Hampshire, North Carolina, Ohio,
Oklahoma, Oregon, Rhode Island, South Carolina, South Dakota, Tennessee, Utah,
Wisconsin, Wyoming
G2 – Moderately Regulated
Nebraska, New Jersey, New York, Texas, Virginia, Washington
G3 – Minimally Regulated
California, Delaware, Michigan, New Mexico, Pennsylvania.
I adopt the regional classifications used by the IPEDS data. This system classifies the
states as follows:
69
The New England Region – Connecticut, Maine, Massachusetts, New Hampshire, Rhode
Island and Vermont.
The Mid-East Region – Delaware, District of Columbia, Maryland, New Jersey, New
York, Pennsylvania
The Great Lakes Region – Illinois, Indiana, Michigan, Ohio, Wisconsin
The Plains Region – Iowa, Kansas, Minnesota, Montana, Nebraska, North Dakota, South
Dakota
The Southeast – Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi,
Tennessee, Virginia, West Virginia
The Southwest – Arizona, New Mexico, Oklahoma, Texas
The Rockies – Colorado, Idaho, Montana, Utah, Wyoming
The Far West – California, Alaska, Hawaii, Nevada, Oregon, Washington
Institutional Level Analysis Variables
For the institution level regressions, the primary dependent variable is the listed
price for tuition and fees as reported in IPEDS with the primary independent variable of
state appropriations per student. Four-year institutions are run separately from two-year
institutions, and in-state prices are distinguished from out-of-state prices.
IPEDS TUITION: The listed tuition prices are directly available from the IPEDS
institutional characteristics data sets and are available for in-district, in-state, and out-ofstate status. The specific variable names accessed in each data set are listed in Appendix
A, which includes all IPEDS data sets utilized with the variable names read from each. It
is worth noting that IPEDS collects tuition data in a variety of ways including charges by
credit hour, charges for different programs, graduate versus undergraduate charges,
including tuition associated fees (or not including them), and so on. For this study the
variable selected is that which represent full time full year annual tuition and fees for an
undergraduate student.
70
Student Level Analysis Variables
The student level analysis adopts the same institutions and institutional
characteristics utilized in the institutional level analysis. The same data and calculation
for state appropriations per student is used as is the criteria for including the institution
(and therefore, the students attending that institution) in the study. In addition to these,
the student level data allow for far more precise regressions controlling for many student
characteristics and financial aid information. Four measures are run separately as the
dependent variable (all as reported by NPSAS), listed tuition price, net price paid by the
student, institutional grants, and the student’s unmet need. The primary independent
variable is the state appropriations per student received by the school, but many other
independent variables are included as controlling variables including the student’s
ethnicity, gender, academic merit, and socioeconomic status. This allows for a more
precise regression of the relationships.
NPSAS TUITION: Like the approach for the IPEDS data, the variables used for
listed tuition price is the variable corresponding to the full year full time undergraduate
tuition. Reported on individual students, the variable is the same whether the student is
in-state or out-of-state, a status that is reported with the jurisdiction variable of the data
set. In instances where the tuition prices reported by the institution are different from the
tuition prices reported for individual students at that same institution, this analysis uses
the student data. Why the prices would be different is outside the scope of this work, and
efforts to understand this continue.
71
NET TUITION is the list tuition reported for the student minus the institutional
grant the student received. In this study, the student’s unmet need is the difference
between a student’s calculated expected family contribution and the costs actually
incurred after financial aid not including loans. The intent behind the variable is to
measure the degree of financial burden, including the debt incurred, associated with
paying for college. I anticipate this result to be more statistically significant and
compelling than the analysis of tuition levels.
GPA: The student’s cumulative grade point average refers to that of the student at
the institution in which the student is enrolled for the NPSAS sample. This is NOT the
student’s high school grade point average prior to admission. For this study, its value is
used as a proxy for academic merit.
INSTITUTIONAL GRANT – This variable is reported directly in the NPSAS
data.
STUDENT CHARACTERISTICS – The NPSAS data contain the student specific
characteristics of ethnicity, gender, age, dependency status, attendance pattern, and
importantly, a great deal of financial information. For this study, the composite expected
family contribution (EFC) variable is used. The composite variable is the most robust in
the NPSAS data set, with considerable effort on the part of NPSAS management to
impute values when necessary.
72
Sample Populations
Institution Level Analysis
The sample is restricted to public non-profit institutions that enroll undergraduate
students. The sample was created by starting with the full universe of public institutions
that reported to IPEDS and then dropping institutions that did not receive any state
appropriations, institutions that are labeled as less-than two-year institutions, and
observations that were listed as district offices, administration headquarters, or other
overhead associated entities that do not have students. Other cases were dropped due to
missing data on institutional enrollment or selectivity. All private institutions including
proprietary schools are excluded. The descriptive statistics for the variables used in the
institutional level analysis are provided in Appendix Table 7.
TABLE 2.
The IPEDS Institution Sample Populations
Academic Year
2003 – 2004
1999 – 2000
1995 – 1996
1992 – 1993
1989 – 1990
Four Year Institutions
432
463
460
468
453
Two Year Institutions
900
941
797
864
858
Student Level Analysis
For the student level analysis with the NPSAS data, the study merges the studentlevel NPSAS data with institutional data from IPEDS and other sources. To maximize
the comparability of the two analyses, the student sample is restricted to those at the same
73
institutions in the IPEDS study. The students are then restricted further to undergraduates
that are dependent, full-time, full year students who attended a single institution during
the year of the NPSAS sample.
The NPSAS Student Sample Populations
The NPSAS population consists of full time full year dependent undergraduates
attending institutions included in the IPEDS institution population. Table Three shows
the sample populations for in-state and out-of-state students at the institutions for each
academic year. It also shows the sample sizes for the institutions disaggregated into the
three selectivity subgroups. The descriptive statistics of the key variables used in the
analysis are provided in Appendix Table 8.
74
Table 3
The NPSAS Sample Population Sizes
Four Year Institutions
(N = Number of Students)
1989 – 1990
All schools
More Selective
Middle Selective
Less Selective
In-State Students
3352
1177
1714
461
Out-of-State Students
395
179
159
57
1992 – 1993
All schools
More Selective
Middle Selective
Less Selective
In-State Students
5433
2517
2180
736
Out-of-State Students
1233
633
425
175
1995 – 1996
All schools
More Selective
Middle Selective
Less Selective
In-State Students
5533
2349
2222
961
Out-of-State Students
1161
613
435
113
1999 – 2000
All schools
More Selective
Middle Selective
Less Selective
In-State Students
5863
2577
2464
822
Out-of-State Students
1039
600
365
74
2003 – 2004
All schools
More Selective
Middle Selective
Less Selective
In-State Students
7669
2880
3679
1101
Out-of-State Students
1061
527
421
111
75
Selectivity is not a factor at community colleges, so no disaggregation of schools
by selectivity is performed. The NPSAS samples in the earlier years have small
representation of two-year schools. While I include the analysis of two-year out of state
students for sake of completeness, the sample sizes are extremely small and the results
found for this group should be interpreted with caution. Table Four provides the sample
sizes for each year.
Table 4
NPSAS Sample Population Sizes
Two-Year Institutions
(N = Number of Students)
Year
1989 – 1990
1992 – 1993
1995 – 1996
1999 – 2000
2003 – 2004
In-State Students
267
398
489
693
3156
Out-of-State Students
39
37
28
102
386
76
CHAPTER FOUR
RESULTS
Introduction
I present the results in two parts, the first part consisting of the institutional level
analysis using the IPEDS data. The second part consists of the student level data
obtained from the NPSAS data. For both sections, I first present the descriptive statistics
and discuss the trends and other key findings. Then I show the results of the ordinary
least squares regressions. Throughout the study, the analysis is conducted separately for
the four different attendance scenarios: in-state and out-of-state status at four-year
schools, and in-state and out-of-state status at the two-year schools. All financial
numbers are expressed in 2004 dollars as calculated by the Consumer Price index
published by the Bureau of Labor Statistics. After presenting the results for the
institutional level and student level models, I discuss the similarities and differences
found between the two and explain the factors that might be causing them.
Recall from chapter three that the selectivity of the four-year institutions is based
on the Barron’s classifications but collapsed to five categories. For more desirable
sample sizes, the schools were further combined to produce three populations, more
selective (the first two combined), middle selective (the third group left as is), and less
selective (the final two combined).
77
Institution Level Analysis – Four-Year Institutions
Table Five provides the summary statistics on the both the in-state and out-ofstate tuition charged at four-year institutions over the five academic years. It shows that
the data examined contain a high degree of consistency from year to year. The sample
populations consist of essentially the same set of institutions, varying from 432 to 468
schools. The selectivity categories are also very consistent in terms of size and data
relationships.
The data strongly demonstrate the positive relationship between school
selectivity and tuition price for both in-state and out-of-state pricing. Average in-state
tuition at the more selective schools exceeded that for the middle selective schools by
approximately 20%, and difference in tuition from the middle selective and the less
selective is about 12%. Selectivity plays a stronger but also apparently stable role with
out-of-state tuition. For out-of-state tuition, the more selective institutions on average
charge tuition at about 25% more than those in the middle selective category, which
charge about 10% more than those in the less selective group. For all four-year
institutions, in-state tuition prices far outpaced the rate of inflation, climbing from an
average of $2456 to $4443 or 81% over the fourteen year period. Out-of-state tuition
rose from an average of $6462 to $11,211, an increase of 73.5%.
The data also confirm the literature regarding the stagnation of state
appropriations received per student over the period of the study. The data for the total
sample show average state appropriations varying from a high of $6727 in 1989/90 to a
78
TABLE 5. Institutional Level Analysis – Four Year Institution Descriptive Statistics
Integrated Postsecondary Education Data System (IPEDS)
Sample Size
Mean / Std. Deviation
All
Institutions
More
Selective
Middle
Selective
Less
Selective
1989 – 1990
In-State Tuition
Out-of State Tuition
State Approp. Per Student
(N=453)
2456 / 883
6462 / 2110
6727 / 2859
(N = 120)
2822/ 1020
7861/ 2385
8375 / 3111
(N = 196)
2427 / 840
6180 / 1767
6370 / 2641
(N = 134)
2163 / 679
5655 / 1682
5828 / 2260
1992 – 1993
In-State Tuition
Out-of State Tuition
State Approp. Per Student
(N=468)
2935 / 1059
7689 / 2565
5798 / 2409
(N = 128)
3441/ 1229
9344 / 2873
7156 / 2201
(N = 202)
2881/ 972
7396 / 2178
5519 / 2200
(N = 134)
2517 / 787
6616 / 1964
5020 / 1786
1995 – 1996
In-State Tuition
Out-of State Tuition
State Approp. Per Student
(N=460)
3278 / 1147
8759 / 2723
5923 / 2237
(N = 124)
3929 / 1350
10549 / 2798
7047 / 2557
(N = 201)
3172 / 1020
8382 / 2331
5683 / 2028
(N = 134)
2819 / check
7689 / 2182
5275 / 1799
1999 – 2000
In-State Tuition
Out-of State Tuition
State Approp. Per Student
(N=463)
3483 / 1154
9464 / 2685
6581 / 2448
(N = 127)
4159/ 1335
11446 / 2943
7804 / 2926
(N = 201)
3368 / 1011
9038 / 2170
6234 / 2089
(N = 134)
3019 / 852
8233 / 2028
5939 / 2029
2003 – 2004
In-State Tuition
Out-of State Tuition
State Approp. Per Student
(N=432)
4443 / 1495
11211 / 3407
5595 / 2323
(N = 114)
5334 / 1869
13791 / 3820
6671 / 2708
(N = 187)
4302 / 1217
10793 / 2586
5312 / 2172
(N = 127)
3848 / 1100
9603 / 2731
5120 / 1811
79
low of $5598 in 2003/04 with levels rising and falling. For all four year institutions, a
trend line over period of the study declines with a slope of $40 (negative) per year in state
appropriations received per student. The level of state appropriations received per
student is also stratified by selectivity, with more selective institutions on average
receiving about 25% more than the middle selective schools, but the difference in state
appropriations received by middle selective schools and the less selective is rather small,
only 4% in 2003/04. Summarizing, the key conclusions to be drawn regarding tuition
from the descriptive statistics are that both in-state and out-of-state tuition are rising far
faster than the rate of inflation and that the stratification of pricing along the selectivity of
the institution is increasing in actual dollars, growing in a way that approximately
maintains the differences as ratios or expressed as percentages. State appropriations have
stagnated but remain stratified with the more selective schools receiving more funding
per student, although the difference between the middle selective and the less selective
schools is diminishing, only 4% between them in 2003/04.
Table Six shows the results of the ordinary least squares regressions performed
on the institutional data for in-state tuition pricing. The data are consistent with the
findings in the literature (Koshal & Koshal, 2000; Hossler et al., 1997) regarding the
suppressing influence of state appropriations on in-state tuition levels. The findings do
indicate a negative relationship between state appropriations and in-state tuition levels, as
negative coefficients were found in almost every regression. For the total sample the
coefficient ranged from – 0.043 (a $100 drop in state appropriations per student is
80
Table 6: Institutional Level Analysis – Four Year Institution Regression
Integrated Postsecondary Education Data System (IPEDS)
In-State List Tuition dependent on State Appropriations
1989 – 1990
All Institutions
More Selective
Middle Selective
Less Selective
B
- 0.043
- 0.046
- 0.018
- 0.080
Std. Dev.
0.013
0.027
0.020
0.018
t value
- 3.13
- 1.71
- 0.90
- 4.30
P>|t|
0.001
0.091
0.371
0.000
1992 – 1993
All Institutions
More Selective
Middle Selective
Less Selective
- 0.031
- 0.031
- 0.007
- 0.082
0.016
0.031
0.024
0.025
- 1.96
- 0.99
- 0.28
- 3.29
0.051
0.324
0.783
0.001
1995 – 1996
All Institutions
More Selective
Middle Selective
Less Selective
- 0.026
- 0.023
- 0.013
- 0.076
0.017
0.035
0.024
0.023
- 1.59
- 0.67
- 0.56
- 3.31
0.112
0.505
0.579
0.001
1999 – 2000
All Institutions
More Selective
Middle Selective
Less Selective
- 0.015
- 0.008
0.002
- 0.079
0.016
0.031
0.023
0.024
- 0.97
- 0.26
0.10
- 3.33
0.331
0.792
0.923
0.001
2003 – 2004
All Institutions
More Selective
Middle Selective
Less Selective
- 0.014
- 0.032
- 0.001
- 0.042
0.025
0.061
0.029
0.043
- 0.058
- 0.52
- 0.04
- 0.98
0.565
0.602
0.967
0.327
81
associated with a $4.30 increase in tuition) in 1989/90 to – 0.014 (a tuition increase of
only $1.40). Interestingly, the magnitude of the coefficients for each year falls rather
steadily as follows: - 0.043, - 0.031, - 0.026, - 0.015, - 0.014. What also occurs is a
steady decline in the significance of the results, starting at a highly significant p value of
0.001 for 1989/90, then 0.051, 0.112, 0.331, and 0.565 respectively for the following
years. One interpretation of these two trends is that the relationship between state
appropriations and in-state tuition across institutions inside of a given year used to exist
as discussed in Koshan & Koshan (2000) but has since essentially disappeared. Another
explanation is that while state appropriations remain a critical funding stream, as they
decline in significance as a part of institutional budgets, appropriations are playing less of
a role (when comparing institutions) in determining tuition levels.
Examining the institutions within the different levels of selectivity shows that the
less selective schools for every year exhibit a stronger relationship between state
appropriations and tuition prices. With the exception of 2003/04, at the less selective
schools a $100 increase in tuition is associated with a tuition increase of approximately
$8.00, double the value of the other schools and with very significant p values of 0.001 or
0.000. While the less selective schools show the strongest response, the middle selective
schools show the least. The data show no virtually no relationship at all between state
appropriations and tuition at the schools in the middle selective category for any of the
years.
Table Seven shows the regression coefficients for out-of-state tuition at four-year
institutions. The findings show a positive relationship between state appropriations and
82
tuition levels. I believe this is the result from a positive relationship between prestige,
state appropriations, and out-of-state tuition. The out-of-state tuition regressions over the
five years showed a high level consistency in the coefficient for the state appropriations
variable. For all five years, it remained within the range of 0.047 (2003/04) to 0.071
(1992/93), steadily declining from 1992/93 to its value in 2003/04. Like the results for
in-state tuition, one clear trend is the decline in the statistical significance of the
coefficients. Over the five years the significance of the relationship steadily decreased,
with the associated p values climbing as follows: 0.073, 0.092, 0.172, 0.250, and 0.437.
This suggests that for both in-state and out-of-state tuition, the association with state
appropriations (across institutions in a single year) is falling during the period of the
study.
When looking at the results with the schools disaggregated by selectivity, no
consistent picture emerges that remains constant for the five years. In the first two years,
the strength of relationship between state appropriations and tuition correlated with
selectivity with the most selective institutions have the highest coefficient, followed by
the middle selective and then the less selective schools. However, this situation reverses
in 1995/96 and by 2003/04, the least selective institutions had the largest coefficient, one
that suggests a $19.10 increase in tuition is associated with a $100 increase in state
appropriations.
83
TABLE 7: Institutional Level Analysis – Four Year Institution Regression
Integrated Postsecondary Education Data System (IPEDS)
Out-of-State List Tuition dependence on State Appropriations
B
Std. Dev.
t value
P>|t|
1989 – 1990
All Institutions
More Selective
Middle Selective
Less Selective
0.055
0.071
0.062
0.044
0.031
0.059
0.045
0.058
1.80
1.19
1.37
0.76
0.073
0.235
0.172
0.446
1992 – 1993
All Institutions
More Selective
Middle Selective
Less Selective
0.071
0.081
0.072
0.044
0.042
0.074
0.062
0.081
1.69
1.09
1.17
0.55
0.092
0.277
0.245
0.586
1995 – 1996
All Institutions
More Selective
Middle Selective
Less Selective
0.062
0.038
0.043
0.119
0.045
0.082
0.063
0.085
1.37
0.47
0.68
1.40
0.172
0.641
0.498
0.165
1999 – 2000
All Institutions
More Selective
Middle Selective
Less Selective
0.053
0.027
0.085
0.013
0.046
0.084
0.069
0.090
1.15
0.32
1.23
0.14
0.250
0.753
0.221
0.886
2003 – 2004
All Institutions
More Selective
Middle Selective
Less Selective
0.047
0.049
0.009
0.191
0.060
0.117
0.084
0.122
0.78
0.42
0.10
1.56
0.437
0.674
0.917
0.122
84
While the relationships between tuition and state appropriations were weak, the
regressions did explain a considerable amount of the variation in tuition levels. The full
regression analyses showing the coefficients for all of the independent variables are
provided in Appendix Table 1 (in-state tuition) and Appendix Table 2 (out-of-state
tuition). Summarizing, the adjusted R squared values ranged from 0.4397 to 0.6018,
suggesting that the variables explained approximately half of the variation in tuition
prices. The results confirm the literature that the New England and Mid East regions
have significantly higher tuition prices followed by the Great Lakes and Plains regions,
with the Southeast and Southwest regions having the lowest tuition. Not surprisingly, the
dummy variable associated with a school’s being in the most selective category had the
largest of the selectivity variable coefficients, ranging between $1200 and $1400 until
2003/04, where this coefficient jumped to $2143. The coefficients for the selectivity
variables decrease as one drops from a higher to a lower category. The variable for
strongly regulated governance systems added between $255 (1999/00) and $677
(1995/96) to in-state tuition, a finding also consistent with the literature that ties stronger
regulation to increased tuition rationalization.
.
Institution Level Analysis – Two-Year Institutions
Table Eight contains the descriptive statistics for the two-year institutions which
also distinguish in-district tuition from in-state and out-of-state tuition, and for all years
where in-district tuition was available (the IPEDS data for 1999/2000 did not contain an
in-district tuition variable) it was lower than in-state tuition by approximately $250 to
85
$300 per year (a substantial 20 - 30%). As with the four-year schools, tuition is
increasing at rates far exceeding inflation, rising 57.7%, 50.0%, and 37.6% for in-district,
in-state, and out-of-state respectively over the fourteen year period. State appropriations
have also stagnated for the two-year schools, rising and falling but remaining in the
vicinity of $4450 per student.
TABLE 8: Institution Level Analysis – Two Year Institution
Descriptive Statistics: Tuition and State Appropriations
Year and Sample Size
Mean / Standard Error
1989 – 1990 (N = 858)
1992 – 1993 (N = 864)
1995 – 1996 (N = 797)
1999 – 2000 (N = 941)
2003 – 2004 (N = 900)
In-District
Tuition
In_State
Tuition
Out-of-State
Tuition
State
Approp / FTE
1221 / 721
1438 / 784
1591 / 803
NA
1925 / 884
1546 / 1096
1752 / 1088
1811 / 1014
1609 / 1045
2319 / 1293
3843 / 882
4431 / 2080
4597 / 2275
4247 / 2313
5287 / 2484
4456 / 2335
3862 / 4921
4437 / 2367
5297 / 9494
4481 / 10691
While the sample sizes of the two-year schools were larger than the four-year
institution samples, ranging from 797 to 941, the two-year schools show a far higher
degree of variation in their tuition levels. While the samples of four-year schools had
standard deviations of about a third of the mean values, the samples of the two-year
institutions have standard deviations closer to a half of the mean values. The degree of
variation among two-year schools is even greater for the level of state appropriations they
receive, and the data suggest the variation is increasing. For the year 1999/2000, state
appropriations per full time equivalent student averaged $5297 with a standard deviation
of $9494. For 2003/2004, the mean funding per student fell to $4481 while the standard
86
deviation rose to $10,691. This degree of variation suggests that the universe of two-year
colleges consists of substantially different financing and pricing environments across the
country. These differences could also be due to more variation in calculating the full
time student equivalent enrollment at the two-year institutions.
Table Nine shows the results of the regression for the two-year institutions
including in-district, in-state, and out-of-state tuition. For all dependent variables, the
coefficients of the independent state appropriations per student variable are very small
and most have minimal statistical significance. Given the relatively small tuition levels at
two-year institutions and their reliance on local appropriations as well as state
appropriations, these results are not surprising.
87
TABLE 9: Institution Level Analysis – Two Year Institution Regressions
State Appropriation Coefficient on Tuition Prices
Tuition Price on Appropriations
B
Std. Dev.
t value
P>|t|
1989 – 1990
(N = 858)
In-District
In-State
Out-of-State
0.014
- 0.053
0.032
0.008
0.011
0.026
1.76
- 4.96
1.26
0.078
0.000
0.210
1992 – 1993
(N = 864)
In-District
In-State
Out-of-State
0.005
- 0.008
0.011
0.004
0.005
0.012
1.47
- 1.75
0.89
0.142
0.080
0.371
1995 – 1996
(N = 797)
In-District
In-State
Out-of-State
0.013
- 0.034
- 0.001
0.007
0.010
0.030
1.73
- 3.40
-0.04
0.084
0.001
0.969
1999 – 2000
(N = 941)
In-District
In-State
Out-of-State
NA
0.001
- 0.013
NA
0.002
0.007
NA
0.36
- 1.74
NA
0.717
0.083
2003 – 2004
(N = 900)
In-District
In-State
Out-of-State
0.004
0.002
0.001
0.002
0.003
0.007
2.18
0.56
0.13
0.030
0.579
0.897
The complete regression results for two-year institutions are shown in Appendix
Table 3. As with the four-year institution regressions, even though the relationship
between tuition and state appropriations was weak, the regressions explained a good deal
of the variation in price (slightly more than the four-year regressions) with adjusted R
squared values over 0.6 for every year but 2003/04, where the value was 0.5254.
Although not in the same proportions, the results are approximately consistent with the
four-year schools regarding the region of the country with the exception that the Mid East
and Plains regions both surpass New England in tuition prices. While not a central focus
88
of the study, the results for governance are quite interesting as the sign for the moderately
and strongly regulated governance is negative for the two-year institutions.
Summarizing, for purposes of this study, the key point from the institutional level
regressions above is that within a given year, state appropriations play a small role,
pennies on the dollar, in explaining the variation of in-state tuition across institutions.
The in-state analysis found a negative relationship that steadily decreases in magnitude
from – 0.043 to – 0.014. For out-of-state tuition, the relationship is positive, suggesting
that institutions with higher state appropriations may have more prestige and charge
higher prices to nonresident students. For both in-state and out-of-state tuition, the
results show declining statistical significance.
The above cross-section regressions focused on examination variation across
institutions at a point in time. In other words, the results were essentially identified by
comparing tuition levels for institutions in high state appropriations states with tuition
levels for institutions in low state appropriations states. But it would be more interesting
to examine how the tuition of individual institutions varied over time as their levels of
state appropriations were altered. For this reason, the study conducts a fixed effects
institutional level regression that simultaneously includes data for all five years. As
shown in Table 10, starting with the four-year institutions, the fixed effects results match
the negative relationship between state appropriations and in-state tuition, but show a
stronger link, suggesting that a $100 decrease in state-appropriations is associated with a
$7.10 increase in in-state tuition, seven cents on the dollar. When examining the results
89
Table 10
Institutional Level Analysis
Fixed Effects Regression
Four Year Institutions
All Institutions
In-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.071
580
1087
1619
2860
Out-of-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.069
1446
2855
4257
7041
Std. Dev.
t value
P>|t|
0.014
32.3
32.8
37.9
77.8
- 4.95
17.97
33.12
42.71
77.79
0.000
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.033
74.8
76.0
87.8
85.1
- 2.09
19.33
37.57
48.48
82.71
0.036
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.030
82.7
81.8
91.8
91.0
- 2.28
9.59
17.33
22.13
39.58
0.023
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.061
164
166
186
185
- 1.51
11.00
20.65
27.86
47.64
0.133
0.000
0.000
0.000
0.000
More Selective Institutions
In-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.068
774
1418
1619
3602
Out-of-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.091
1802
3431
5190
8803
90
Table 10
Institutional Level Analysis
Fixed Effects Regression on Five Year Panel Data (continued)
Four Year Institutions
Middle Selective Institutions
In-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.096
550
1026
1554
2739
Out-of-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.151
1411
2779
4201
6853
Std. Dev.
t value
P>|t|
0.019
39.5
40.1
46.7
44.92
- 5.08
13.90
25.59
33.25
60.98
0.000
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.045
95.0
96.4
112
108
- 3.34
14.85
28.84
37.40
63,50
0.001
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.021
43.4
44.5
53.8
51.1
- 2.12
10.64
19.83
25.68
46.88
0.035
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.059
120
123
147
142
1.12
10.02
19.88
23.94
40.80
0.262
0.000
0.000
0.000
0.000
Less Selective Institutions
In-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.045
462
881
1354
2395
Out-of-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
0.067
1207
2452
3511
5785
91
Table 10
Institutional Level Analysis
Fixed Effects Regression on Five Year Panel Data (continued)
Two Year Institutions
In-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
- 0.001
278
511
869
1116
Out-of-state Tuition
B
State Appropriations
Year 9293
Year 9596
Year 9900
Year 0304
0.003
833
1326
1873
2816
Std. Dev.
t value
P>|t|
0.003
20.8
21.5
21.3
21.2
- 0.22
13.37
23.75
40.83
52.56
0.822
0.000
0.000
0.000
0.000
Std. Dev.
t value
P>|t|
0.007
74.8
76.0
87.8
85.1
0.47
19.33
37.57
48.48
82.71
0.637
0.000
0.000
0.000
0.000
92
for institutions at the different levels of selectivity, the middle selectivity group had the
largest response, a $9.60 increase compared to $6.80 for the more selective and $4.50 for
the less selective. Curiously, the fixed effects results for out-of-state tuition paint the
opposite picture of the single year regressions, showing that a $100 decrease in state
appropriations is associated with a $6.90 increase in tuition (the opposite direction),
almost identical to the seven cents on a dollar result for in-state tuition. In the fixed
effects analysis, the increasing coefficients for the year indicator variables are very strong
and reinforce the rapid rise in tuition prices over the period of the study.
For the two-year schools, the fixed effect model matched the results of the prior
regressions, also finding no significant relationship between state appropriations and
either the in-state or the out-of-state tuition prices. The coefficients for the state
appropriations variables are extremely small and have negligible statistical significance,
with p values exceeding 0.6 for both in-state and out-of-state regressions. As with the
four-year schools, the increasing coefficients for the year indicator variables indicate the
rise in tuition prices over time.
93
Student Level Analysis - Four-Year Institutions
One would anticipate slight differences to exist between the institution level data
and the student level data since they come from different data sets and are reporting
summary statistics on different populations (institutions vs. students). However, the two
should have a degree of correspondence, and they do in the case of the descriptive
statistics. With the exception of 1989/90, the institutional level average in-state tuition at
the four-year institutions tracks with the student level average in-state tuition at four-year
schools, with the student data consistently averaging in-state tuition prices approximately
$300 – $600 higher than that reported by the institutions. The out-of-state comparison
differs in that the students reported substantially lower tuition prices for 1989/90 ($4339
compared to $6462 for the institutions), about the same in 1992/93 ($7495 compared to
$7689 for the institutions), and then reporting higher prices by about $1200 for the last
three years. As shown in Table 11, the student level data also show the state
appropriations per student not only stagnating but declining. Like tuition, state funding at
an institution for in-state students and out-of-state students are relatively close within a
given year. The in-state tuition, out-of-state tuition, and state appropriations per student
means are consistently higher for the student level data. This result reflects the greater
representation of students from high-tuition and high-appropriations schools in the
NPSAS samples.
94
TABLE 11: Student Level Analysis – Descriptive Statistics: Four Year Institutions
Sample Size
Mean / Std. Deviation
All
Institutions
More
Selective
Middle
Selective
Less
Selective
1989 – 1990
In-State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
(N=3570)
1675 / 1400
149 / 703
1720 / 4196
7227 / 3052
(N = 1270)
1869 / 1889
187 / 757
1855 / 4582
8658 / 3049
(N = 1795)
1563 / 978
123 / 683
1614 / 3920
6462 / 2810
(N = 505)
1541 / 991
136 / 607
1734 / 4051
6023 / 2293
(N = 431)
4339 / 2678
222 / 987
2378 / 6276
7108 / 2540
(N = 201)
5139 / 3038
150 / 981
2747 / 6563
7950 / 2606
(N = 162)
3608 / 1994
267 / 998
2207 / 6180
6582 / 1967
(N = 68)
3400 / 1952
363 / 969
1510 / 5410
5480 / 2512
(N = 5916)
3204 / 1529
346 / 1152
5449 / 2706
6568 / 2535
(N = 2831)
3496 / 1799
379 / 1255
5689 / 2869
7640 / 2708
(N = 2284)
3006 / 1193
340 /1086
5241 / 2365
5685 / 1932
(N = 801)
2719 / 1029
247 / 923
5372 / 3080
5222 / 1643
(N = 1374)
7495 / 4370
1069 / 2895
6426/ 5065
6181 / 2374
(N = 707)
9008/ 4729
1116 / 3032
7892 / 5434
6731 / 2577
(N = 467)
6154 / 3325
1036 / 2765
5117 / 4194
5759 / 2402
(N = 200)
4694 / 2237
958 / 2643
3736 / 3234
4963 / 1538
Out-of State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
1992 – 1993
In-State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
Out-of State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
95
TABLE 11: Student Level Descriptive Statistics (continued)
Sample Size
Mean / Std. Deviation
All
Institutions
More
Selective
Middle
Selective
Less
Selective
1995 – 1996
In-State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
(N = 5695)
3584 / 1282
444 / 1338
3390 / 4028
6851 / 2490
(N = 2406)
3967 / 1408
483 / 1521
3394 / 4178
7861 / 2682
(N = 2307)
3350 / 1061
423 / 1212
3381 / 3965
6299 / 2134
(N = 979)
3164 / 1167
396 / 1103
3389 / 3779
5583 / 1639
(N = 1209)
10042 / 3990
1398 / 3287
8644 / 4988
6719 / 2518
(N = 642)
11435/ 4330
1441 / 3455
9993 / 5293
7373 / 2847
(N = 453)
8522 / 2797
1502 / 3309
7020 / 4290
6263 / 1944
(N = 113)
10581 / 14939
878 / 2249
7647 / 3728
5191 / 1251
(N = 5844)
4001 / 1756
565 / 1603
3496 / 3889
7111 / 2553
(N = 2572)
4372 / 1857
699 / 1874
3472 / 3935
8120 / 2813
(N = 2452)
3633 / 1307
471 / 1387
3445 / 3828
6455 / 2104
(N = 820)
3877 / 2248
411 / 1141
3723 / 3912
5766 / 1334
(N = 1035)
10095 / 4719
2090 / 4190
6068 / 6642
6900 / 2358
(N = 599)
11830 / 4676
2054 / 4429
6615 / 6856
7353 / 2544
(N = 362)
7693 / 3573
2011 / 3673
5184 / 6327
6422 / 2001
(N = 74)
7618 / 3658
2736 / 4510
5880 / 5951
5553 / 1091
Out-of-State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
1999 – 2000
In-State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
Out-of State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
96
TABLE 11: Student Level Descriptive Statistics (continued)
Sample Size
Mean / Std. Deviation
All
Institutions
More
Selective
Middle
Selective
Less
Selective
2003 – 2004
In-State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
(N = 7669)
4739 / 1542
761 / 2104
4077 / 4205
6138 / 2412
(N = 2880)
5013 / 1640
926 / 2456
3981 / 4331
7073 / 2749
(N = 3679)
4753 / 1483
659 / 1917
4108 / 4163
5726 / 2090
(N = 1101)
3958 / 1169
683 / 1650
4223 / 4009
5110 / 1534
(N = 1061)
12777 / 5025
2231 / 4267
8660 / 8041
6299 / 2378
(N = 527)
15641 / 4105
2241 / 4510
9799 / 8969
6797 / 2783
(N = 421)
10609 / 4479
1881 / 3908
7766 / 7138
5962 / 1868
(N = 111)
8353 / 2982
3684 / 4273
7046 / 6033
5363 / 1561
Out-of State Students
Tuition
Institutional Grant
Unmet Need
State App per FTE
97
Turning to the ordinary least squares regressions, Table 12 shows the results for
the students paying in-state tuition, and Table 13 shows the results for the students paying
out-of-state tuition. For both in-state and out-of-state listed tuition, the results do not
reveal a consistent relationship between tuition and state appropriations, showing
relationships of different size and sign for the different years. The extreme variation in
the findings across years is difficult to explain and certainly suggests that researchers
should be cautious when examining institution-level variables using only one year of
NPSAS.
98
Table 12
Student Level Analysis: Linear Regression
In-State Students at Four-Year Institutions
State Appropriations per student Coefficient
1989 – 1990
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.051
0.054
- 0.003
0.035
Std. Dev.
0.009
0.010
0.005
0.029
t value
5.77
5.24
- 0.59
1.22
P>|t|
0.000
0.000
0.555
0.222
1992 – 1993
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
- 0.039
- 0.031
- 0.007
- 0.044
Std. Dev.
0.007
0.010
0.008
0.026
t value
- 5.27
- 3.09
- 0.98
- 1.68
P>|t|
0.000
0.002
0.326
0.094
1995 – 1996
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.037
0.025
0.012
0.038
Std. Dev.
0.006
0.009
0.008
0.021
t value
6.63
2.66
1.47
1.84
P>|t|
0.000
0.008
0.142
0.066
1999 – 2000
List Tuition
Net
Institutional Grant
Unmet Need
B
- 0.006
- 0.042
0.036
- 0.004
Std. Dev.
0.009
0.013
0.009
0.019
t value
- 0.64
- 3.38
3.84
- 0.19
P>|t|
0.524
0.001
0.000
0.846
2003 – 2004
List Tuition
Net
Institutional Grant
Unmet Need
B
- 0.019
- 0.025
0.006
0.001
Std. Dev.
0.006
0.011
0.011
0.017
t value
- 3.17
- 2.16
0.52
0.05
P>|t|
0.002
0.031
0.602
0.961
99
TABLE 13
Student Level Analysis: Linear Regression
Out-of-State Students at Four Year Institutions
State Appropriation Coefficients
1989 – 1990
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
- 0.099
- 0.036
- 0.063
0.043
Std. Dev.
0.054
0.059
0.024
0.136
t value
- 1.83
- 0.61
- 2.58
0.32
P>|t|
0.068
0.542
0.010
0.751
1992 – 1993
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.068
- 0.070
0.139
0.170
Std. Dev.
0.048
0.060
0.041
0.089
t value
1.41
- 1.16
3.38
1.91
P>|t|
0.159
0.245
0.001
0.057
1995 – 1996
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.067
0.025
0.086
- 0.047
Std. Dev.
0.045
0.009
0.044
0.073
t value
1.49
2.66
1.96
- 0.65
P>|t|
0.138
0.008
0.051
0.518
1999 – 2000
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
- 0.015
- 0.110
0.121
- 0.094
Std. Dev.
0.016
0.055
0.060
0.074
t value
- 0.97
- 1.99
2.00
- 1.26
P>|t|
0.331
0.047
0.046
0.209
2003 – 2004
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
- 0.136
- 0.210
0.074
- 0.176
Std. Dev.
0.053
0.079
0.062
0.087
t value
- 2.55
- 2.67
1.21
- 2.03
P>|t|
0.011
0.008
0.227
0.043
100
Because the net tuition paid by students falls by the amount of institutional grants
they received, one might expect a more negative relationship between net tuition and
state appropriations if institutions indeed use these funds for institutional aid. However,
as would be anticipated by the discussion in the above paragraph, the results do not
support this conclusion and yield contradictory findings from year to year as those
obtained for list tuition. For in-state students the coefficients for state appropriations with
net tuition as the dependent variable have magnitudes below 0.05 (except for 1989/90
where it is 0.054) that change in sign from year to year. The results are more consistent
for out-of-state students, as the sign is negative for four of the five years. The results are
especially strong for the last two years with coefficients of – 0.110 and – 0.210.
Turning to the dependent variable for institutional grants awarded to students, one
would anticipate positive coefficients for state appropriations if these funds indeed
support institutional aid to students. For the in-state students, with the exception of one
year the results yield low statistical significance suggesting little relationship (which is
different from an inconsistent relationship that has statistical significance) between the
state appropriations institutions receive and the institutional grants provided to students.
The most probable explanation is that the state appropriations received by schools simply
do not influence the institutional aid provided to in-state students. For the out-of-state
students, with the exception of the first year, the results are again more consistent,
suggesting that a $100 increase in state appropriations is associated with between a $7.40
(2003/04) and a $13.90 (1995/96) increase in institutional grants to out-of-state students.
101
In this study, the student’s unmet need is the difference between a student’s
calculated expected family contribution and the costs actually incurred after financial aid
not including loans. Before conducting the study, I expected this result to be more
statistically significant and compelling than the analysis of tuition levels. However, the
same inconsistencies and lack of statistical significance found occur with unmet need.
The full regression analysis of tuition for in-state students at four-year institutions
is provided in Appendix Table 4. The same regression for out-of-state students is
provided in Appendix B Table 5. The results for the other independent variables, while
also demonstrating considerably more volatility than the institutional level data, are more
consistent across time than the results for state appropriations. The variables associated
with the student characteristics of ethnicity and gender remain rather small and
insignificant in terms of influencing tuition price. The general themes of higher tuition in
the New England and Mid East regions also occur in the student data.
Student Level Analysis – Two Year Institutions
The student level regression for those attending two-year institutions is shown in
Table 14. The results are for students classified as paying in-state tuition. The results
reflect the same lack of consistency from year to year and demonstrate little or no
relationship between the dependent variables and state appropriations The NPSAS
sample populations for students paying out-of-state tuition at two-year institutions were
so small that I dropped the analysis of these samples from the study.
The complete student level regression of those at two-year institutions paying instate tuition is provided in Appendix Table 6. The results are similar to those found for
102
the four-year students in terms of ethnicity and gender, i.e. no particular relationship with
the average tuition paid by the students. The results for governance match the negative
relationship found at the institutional level. For two-year institutions, strongly regulated
governing boards are associated with lower tuition levels. Though with different
proportions, the theme of higher tuition in New England and the Mid East and lower
tuition in the Southeast and Southwest also occurred.
103
Table 14
Student Level Analysis: Linear Regression
In-State Students at Two Year Institutions
State Appropriations per student Coefficient
1989 – 1990
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
- 0.001
- 0.013
0.012
- 0.121
Std. Dev.
0.007
0.012
0.007
0.088
t value
- 0.12
-1.10
1.60
-1.38
P>|t|
0.907
0.273
0.110
0.168
1992 – 1993
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.003
0.004
0.001
0.350
Std. Dev.
0.004
0.006
0.004
0.179
t value
0.80
0.62
- 0.14
1.96
P>|t|
0.423
0.536
0.887
0.053
1995 – 1996
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
- 0.008
- 0.015
0.024
0.237
Std. Dev.
0.018
0.028
0.024
0.099
t value
- 0.51
- 0.54
0.99
2.41
P>|t|
0.611
0.586
0.325
0.016
1999 – 2000
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.061
- 0.146
0.005
- 0.166
Std. Dev.
0.018
0.127
0.010
0.059
t value
3.24
- 1.15
0.47
- 2.81
P>|t|
0.001
0.254
0.638
0.005
2003 – 2004
List Tuition
Net Tuition
Institutional Grant
Unmet Need
B
0.044
0.114
- 0.070
-0.024
Std. Dev.
0.009
0.024
0.022
0.050
t value
4.49
4.81
- 3.19
- 0.48
P>|t|
0.000
0.000
0.001
0.632
104
Summary of Results
This study examined the relationship between state appropriations and tuition
pricing and financial aid. The analysis was conducted on institutional level data from the
Integrated Postsecondary Educational Data System (IPEDS) and student level data from
the National Postsecondary Student Aid Study (NPSAS) for five academic years
spanning the period from 1989/90 to 2003/04. I explored the following fundamental
questions.
The Institutional Level Analysis
1. Given a cross section of four-year and two-year public colleges and universities,
how does the list price of tuition vary as state appropriations per student in public
funding changes?
The institutional level analysis conducted for each year found a negative but
statistically weak relationship between in-state tuition levels and the state appropriations
per full time equivalent student received by the schools. Further, the relationship grows
weaker over the period of the study, suggesting the higher education institutions are
decoupling state appropriations from their tuition policies. For the four-year institutions,
results did vary by selectivity, with the less selective schools reacting more strongly to
changes in state appropriations. For out-of-state tuition, a positive relationship was
found, suggesting that higher state appropriation levels may correlate with higher
institutional prestige and the ability to charge higher out-of-state tuition. For the twoyear institutions, neither the in-district, in-state or out-of-state tuition levels demonstrated
any relationship with state appropriations.
105
The institutional level fixed effects regression of the study is perhaps the most
robust analysis performed in this study, analyzing panel data combining the information
from all five years. This analysis implicitly controls for all institution-level
characteristics that are constant over time, by relying purely on variation over time for
individual institutions. It confirmed the negative relationship between state
appropriations and in-state tuition but found a stronger correlation of about seven cents
on the dollar, or that a $100 decrease in state appropriations is associated with a $7
increase in in-state tuition. It found a similar relationship for out-of-state students, which
contradicts the out-of-state findings from individual year OLS regressions. The fixed
effects results suggest that a drop of $100 in state appropriations per student is associated
with a tuition increase of approximately $7.00, both in-state and out-of-state.
Student Level Analysis
2. When looking at a cross section of college students throughout the country, how do
the list tuition, net tuition, institutional grant, and their unmet need vary as the state
appropriation level per student changes?
The NPSAS data produced puzzling inconsistencies in the results of the single
year OLS regressions, yielding coefficients for the variable relationships that varied in
size and sign from year to year. These inconsistencies occurred for all four of the
dependent variables. While greater volatility can be expected since the NPSAS samples
consist of different students, the findings are difficult to explain.
The student level analysis of the NPSAS data show a small and weakening
relationship between in-state tuition and state appropriation levels when controlling for
106
the additional factors. The changing signs and significance of the coefficients over the
years suggest that the level of state appropriations received by institutions does not play a
role in the tuition or institutional aid across students in a given academic year. This holds
for both in-state and out-of-state students.
107
CHAPTER FIVE
CONCLUSION
Introduction
This chapter begins with a description of the key findings and a discussion of
what these results are saying about what is happening in the finance of higher education,
and what it all means for both institutions and students. I will address some of the
implications the study may have for higher education policy makers and the leaders of
higher education institutions. The chapter concludes with an exploration of interesting
possibilities for future research that would build on the work performed here.
The Key Findings
As shown in Table 15, the cross sectional institutional level analysis using IPEDS
data confirmed the literature and existing understanding that state appropriations has a
suppressing effect on the in-state tuition charged at public colleges and universities.
However, when looking at this relationship over the fourteen year period of the study,
this relationship is declining as the share of state appropriations of total institutional
budgets declines. For out-of-state tuition, the influence of state appropriations may have
already ceased suppressing tuition entirely, for a positive relationship was found.
Schools receiving more state support charged higher out-of-state tuition, but that
relationship also weakened over time. Amidst all of the complexity of higher education
finance, the cross sectional IPEDS analysis points to a rather simple conclusion. As state
108
appropriations become a smaller part of the budget of higher education institutions, so
does its influence on tuition prices and quite possibly everything else.
Table 15
Summary of Results – Tuition Regressions
Four Year Institutions
State Appropriation Coefficients
In-State Analysis
IPEDS
NPSAS
Out-of-State Analysis
IPEDS
NPSAS
1989 – 1990
1992 – 1993
1995 – 1996
1999 – 2000
2003 – 2004
- 0.043
- 0.031
- 0.026
- 0.015
- 0.014
0.055
0.071
0.062
0.053
0.047
Fixed Effects
- 0.071
0.051
- 0.039
0.037
- 0.006
- 0.019
- 0.099
0.068
0.067
- 0.015
- 0.136
- 0.069
The student level analysis using the NPSAS data tells a story for each of the five
years examined, but unfortunately the stories vary dramatically over time. For both instate and out-of-state tuition, for some years state appropriations appears to suppress
tuition for some years, and in other years the opposite occurs. I had hoped to find
compelling results regarding the relationship between state appropriations and
institutional grants provided to students, the net tuition they paid after receiving financial
aid, and the unmet need faced by students attending college. The NPSAS data did not
produce results with the consistency or the magnitudes to support such findings.
The most compelling analysis conducted in this study is the fixed effects
regression performed on the panel data set provided by the five years of the IPEDS
109
institutional level data. Because the fixed effects regression accounts for changes over
time and controls for institutional traits that remain unchanged, it examines how an
individual institution’s tuition changes over time with its level of state appropriations.
Interestingly, the fixed effects model produced remarkably similar results for both instate and out-of-state tuition in terms of their relationship with the state appropriations
received by institutions. For each, the model suggests an association of seven cents on
the dollar. A drop in state appropriations of one dollar increases both in-state and out-ofstate tuition by seven cents.
In one area, all of the analyses agree. The institutional level for the individual
years, the fixed effects regression using panel data, and the student level analyses all
concur that no relationship exists between the level of state appropriations received and
the tuition prices set at two-year institutions.
The Big Picture
I have examined the data for the relationship between state appropriations and
tuition prices at higher education institutions in many different ways in an attempt to
quantify as precisely as possible any association that may exist between the two at
institutions across the country. The fixed effects analysis found that on average, a one
dollar drop in state appropriations per student was associated with a seven cent increase
in tuition. This result occurred for both in-state and out-of-state tuition at the four-year
institutions..
The institutional level analysis of the individual years for in-state tuition at fouryear institutions also finds a negative relationship between state appropriations and
110
tuition prices, but one that is smaller than the results of the fixed effects analysis. The
single year OLS regressions also yield a coefficient steadily shrinking in size over the
five years, and steadily shrinking in statistical significance as well. These results are
consistent with (but do not prove) the view that a stronger relationship between state
appropriations and in-state tuition at four-year institutions may have existed in the past,
but it is now small and declining.
While the fixed effects analysis over time found a negative relationship between
out-of-state tuition at four-year institutions and state appropriations, the single year OLS
institutional level regressions found a positive relationship, suggesting that high state
appropriations institutions charge higher out-of-state tuition. It is feasible that the
schools receiving greater amounts of state appropriations have more prestige and can
charge higher tuition premiums for nonresident students. Further, the student level
analyses of the NPSAS data shows that institutional aid is higher (and the net tuition is
lower) at the schools with high state appropriation levels. Higher state appropriations
may allow four-year institutions to use institutional aid to attract out-of-state students
with desirable characteristics.
All of the analyses conducted suggest that no relationship exists between the state
appropriations received by two-year institutions and their tuition prices. Since two-year
institutions are focused on low cost programs and also subsidized by local community
funding, this result is not surprising. Since community colleges are not selective, it is
unlikely that they have enrollment management practices seeking to optimize net tuition
revenue or other measures associated with the students who enroll.
111
Implications
The continuing decline in the share of state appropriations in higher education
budgets has caused universities to raise tuition and develop alternative revenue streams to
meet the increasing costs of educating students. The study suggests that while over time
state appropriations are a factor when considering tuition levels, they play a small role in
conjunction with many other factors that include the region of the country, the politics of
the state legislature, the school’s selectivity, and the demand that exists among students.
State appropriations remain a critical element of university budgets, but the regressions
run in this study find little relationship within a given year between the list or net tuition
students experience and the appropriations provided to the institution they attend. The
role of state appropriations in setting tuition levels appears to be one that occurs over
time.
As noted in chapter two, basic economic theory and the need to balance
institutional budgets predicts a negative relationship between state appropriations and
tuition. Both the fixed effects regressions and the descriptive statistics support this
relationship, and in particular, that the relationship occurs over time. When isolated to a
single year, however, the results suggest that variation in state appropriation levels has
little or no influence on tuition levels. Further, what little influence was found for instate tuition is decreasing from year to year. Overall, the relationship found is relatively
weak and may be decreasing over time. Bowen’s revenue theory suggest institutions
charge what they can irrespective of state appropriations, and it is possible that tuition
112
prices are set within governance structures and other political and institutional
considerations that place little emphasis on state appropriations.
The single year OLS regressions suggest that the influence of state appropriations
on tuition is weakening. By declining as a factor, the funds lose their desired intent of
suppressing tuition and enhancing access to higher education. If the trends identified in
this study continue, the role of state appropriations in subsidizing higher education will
fall further, and tuition will continue to rise at rates far exceeding the rate of inflation.
Students from low income backgrounds, in particular first generation students less
familiar with the financial aid system and more prone to “sticker shock” are likely to
view education as increasingly out of reach.
The trends noted in the descriptive statistics point to the continuing stratification
of institutions in terms of selectivity and prestige, with the differences between the state
appropriations and tuition of the more selective schools expanding away from that of the
less selective schools. As this trend continues, public schools at the top of the prestige
hierarchy can become as inaccessible to lower income populations as the traditionally
expensive elite private institutions. The drift away from substantial state subsidization
and low tuition prices significantly increases the financial component of the decision to
pursue higher education.
An unintended implication of the study involves the inconsistencies found in the
NPSAS data and the warning that caution should be used when using student level data to
infer results regarding institutions.
113
Future Research
The perplexing inconsistencies found in the NPSAS data suggest that further
research is warranted to clarify the understanding of its variables and their values in a
way that can apply to institutional level studies. It would be useful to see how the
NPSAS tuition data for students compares to the tuition values institutions report to
IPEDS for the same academic years. Even when the student populations were restricted
to full time dependent students attending only one four-year institution in a year, this
study found encountered highly inconsistent data. Research identifying and explaining
the phenomena behind these inconsistencies could prove valuable to those using NPSAS
data in their research.
Another interesting study would be to conduct an analysis similar to this one for
the private not-for-profit institutions replacing state appropriations received per full time
student with endowment income per full time student. In particular, I would anticipate
that schools with generous endowment income are in the position to discount tuition or
offer institutional grants to the students meritorious enough to be admitted. I think a
stronger correlation between endowment income and institutional aid exists than what
this study could find between state appropriations and institutional aid. Such a study
should also note the distinction between list tuition, which is unlikely to have a
compelling story, and what the students actually pay, which may prove very interesting.
Several of the highly endowed, prestigious elite institutions have recently adopted
policies that offer extraordinary financial aid even to students with middle and upper
114
middle class backgrounds. A study analyzing this for all private non-profit higher
institutions would be an interesting contribution to the literature.
For those interested in deepening the understanding of the consequences of
increasing stratification, future research could keep the more precise disaggregation of
the selectivity categories utilized by the Barron’s taxonomy and conduct an analysis
similar to this one. This may provide insight into the confusing differences found here
where the more selective and less selective schools showed a negative relationship
between state appropriations and tuition, while the middle selective group showed a
positive one. As mentioned, the selectivity of a public four-year institution strongly
influences the nature of its enrollment management context. It would also be interesting
to disaggregate four-year institutions by other classifications, such as the region of the
country and a measure I did not consider, the size of the institutions or their Carnegie
classification.
Further, this study examined the cross section of institutions and students one year
at a time to explore the relationship between state appropriation levels and tuition and
financial aid. I found that as one moves from one institution (or student) to others, there
is very little association between the appropriation levels and the tuition prices. The
fixed effects regression, which takes into account changes over time, found the opposite
relationship between out-of-state tuition and state appropriations. A study that creates
variables for year to year changes in state appropriations and the year to year changes in
tuition pricing may find a stronger association between them. Such a study could also
115
perform the analysis with the tuition change lagging the appropriation change by one or
more years.
The findings of the study also point to the value of further research in the
alternative revenue “other” that is growing faster than tuition. Quantifying the amounts
received via fundraising, research grants, room and board, bookstores (most schools pay
75% or less of retail for textbooks and sell at full price), attractive and cash hungry
student unions, and the myriad of other revenue enhancing activities schools have
discovered could provide an interesting contribution to the understanding of higher
education finance. Conducting these studies for the different groupings mentioned in the
previous paragraph could yield valuable insights.
Conclusion
The United States and virtually every country will continue their commitment to
provide substantial funding and resources for the education and higher education of its
citizens. The higher education system will not be allowed to fail. The task at hand is
more accurately viewed as a challenge of seeking optimization. What is the best higher
education system and what is the best way to allocate resources for it? The question has
no easy answer. Saddled with competing and compelling obligations to Medicaid,
corrections, K-12 education, and other demands, state appropriations for higher education
are like to continue to stagnate or decline. Sadly, the recent history in the United States
has not been encouraging. The country has drifted far a field from the vision of
Johnson’s Great Society. Reagan spoke of “trickle down” economics, and in some
116
respects trickle down forces have been implemented. The federal government has
squeezed the states, and the squeeze has trickled down to state support for higher
education.
State legislatures must wrestle with their relationship with higher education.
What role do states want in the college education of their residents? As appropriations
continue to decline in university budgets, state relevance to the system approaches a
tipping point where universities would forego (or only accept without strings) minimally
significant financial support in exchange for the autonomy from state legislatures. The
liberation of this autonomy is likely to be an illusion, for the benefactors filling the void
are sure to voice their sentiments. Who are the likely future benefactors of higher
education? If the spoken goals of the Obama administration prevail in federal education
policy, students armed with generous federal financial aid will line up at higher education
institutions, and their objectives are almost certain to focus on employment. The other
benefactors are likely to be the future employers of those students, and the mission of
higher education will be the manufacturing of skilled employees trained in the context of
economic utility. Knowledge for its own sake has ironically once again become a luxury
reserved for the elite and well connected few attending a small number of higher
education institutions utterly inaccessible to the general population. The 1600s have
emerged in the 21st Century.
117
APPENDIX A
IPEDS INSTITUTIONAL LEVEL DATA SOURCES
IPEDS – Institution level data is accessed from the online IPEDS STATA data sets.
2003-2004
Directory
Enrollment
Finance
Institutional Characteristics
DATA SET
hd2004_data_stata.csv
ef2004a_data_stata.csv
f0304_f1a_data_stata.csv
ic2003_ay_data_stata.csv
Variables
Name, Level, Control, Sector
Enrollment populations
State Appropriations
Tuition and Fees
1999-2000
Directory
Enrollment
Finance
Institutional Characteristics
ic99_hd_data_stata.csv
ef99_anr_data_stata.csv
f9900_f1_data_stata.csv
ip1999_ay_data_stata.csv
Name, Level, Control, Sector
Enrollment populations
State Appropriations
Tuition and Fees
1995-1996
Directory
Enrollment
Finance
Institutional Characteristics
ic9596_a_data_stata.csv
ef95_anr_data_stata.csv
f9596_a_data_stata.csv
ic9596_b_data_stata.csv
Name, Level, Control, Sector
Enrollment populations
State Appropriations
Tuition and Fees
1992-1993
Directory
Enrollment
Finance
Institutional Characteristics
ic1992_a_data_stata.csv
ef1992_a_data_stata.csv
f1992_a_data_stata.csv
ic1992_b_data_ stata.csv
Name, Level, Control, Sector
Enrollment populations
State Appropriations
Tuition and Fees
1989-1990
Directory
Enrollment
Finance
Institutional Characteristics
ic90hd_data_stata.csv
ef1989_a_data_stata.csv
f8990_a_data_stata.csv
ic1989_b_data_ stata.csv
Name, Level, Control, Sector
Enrollment populations
State Appropriations
Tuition and Fees
118
APPENDIX TABLE 1
Institution Level Analysis – Four Year Institution Regression Results
Dependent Variable: In-State Tuition
2004 Dollars
Number of Institutions
Adjusted R-Squared
1989/90
1992/93
1995/96
1999/00
2003/04
453
0.4397
468
0.5700
460
0.6018
463
0.5749
432
0.4446
- 0.015 / 0.016
255 / 128
60 / 113
1397 / 178
773 / 137
259 / 120
- 55 / 137
1924 / 198
1865 / 151
1427 / 158
711 / 176
244 / 153
- 155 /169
245 / 207
2441 / 203
- 0.014 / 0.025
362 / 223
- 85 / 168
2143 / 276
1138 / 197
434 / 174
62 / 201
1560 / 339
1354 / 301
1235 / 309
539 / 315
- 181 / 287
- 754 / 309
- 506 / 364
3555 / 343
Coefficient / Standard Error
State Appropriations
Govern1
Govern2
Most Selective
More Selective
Middle Selective
Less Selective
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Constant / Intercept
- 0.043 / 0.013
387 / 111
- 102 / 99
1176 / 159
498 / 120
238 / 124
0 / 119
1291 / 184
1164 / 132
1280 / 146
597 / 162
597 / 141
- 151 / 157
507 / 186
1741 / 186
- 0.031 / 0.016 - 0.026 / 0.017
590 / 117
677 / 124
274 / 104
453 / 109
1179 / 163
1349 / 171
554 / 125
651 / 130
203 / 109
195 / 114
- 105 / 124
- 107 / 130
2151 / 118
2088 / 191
1445 / 139
1492 / 145
1141 / 149
1152 / 153
568 / 167
525 / 172
342 / 147
179 / 149
- 533 / 161
- 671 / 167
307 / 194
169 / 200
2074 / 192
2389 / 194
=
119
APPENDIX TABLE 2
Institution Level Analysis – Four Year Institution Regression Results
Dependent Variable: Out-of-State Tuition
2004 Dollars
Number of Institutions
Adjusted R-Squared
1989/90
1992/93
1995/96
1999/00
2003/04
453
0.4560
468
0.4940
460
0.4787
463
0.4463
432
0.3650
0.062 / 0.045
2517 / 338
1529 / 297
4071 / 465
2475 / 354
824 / 309
507 / 354
2190 / 518
- 977 / 394
537 / 415
- 1489 / 466
- 1465 / 404
- 2023 / 453
436 / 541
7177 / 528
0.070 / 0.041
1515 / 342
964 / 302
4525 / 474
2851 / 365
934 / 322
454 / 366
1843 / 530
- 1396 / 404
561 /424
- 1757 / 474
- 1261 / 413
- 2228 / 451
265 / 555
7952 / 551
0.047 / 0.060
968 / 544
175 / 409
6712 / 674
4230 / 480
1697 / 424
1083 / 490
951 / 827
- 1298 / 733
1210 / 752
- 1598 / 767
- 874 / 699
- 2656 / 752
- 506 / 364
9290 / 836
Coefficient / Standard Error
State Appropriations
Govern1
Govern2
Most Selective
More Selective
Middle Selective
Less Selective
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Constant / Intercept
0.055 / 0.031
1689 / 262
- 102 / 99
3361 / 373
2006 / 283
731 / 246
511 / 280
1293 / 432
- 1064 / 311
998 / 342
- 1354 / 382
- 784 / 331
- 1139 / 369
34 / 437
5114 / 439
0.071 / 0.042
2456 / 308
1010 / 273
3955 / 427
2406 / 328
885 / 286
469 / 326
2911 / 495
- 1003 / 364
781 / 393
- 1197 / 439
- 830 / 385
- 1723 / 423
466 / 509
5840 / 505
=
120
APPENDIX TABLE 3
Institution Level Analysis – Two Year Institution Regression Results
Dependent Variable: In-State Tuition
2004 Dollars
Number of Institutions
Adjusted R-Squared
1989/90
1992/93
1995/96
1999/00
2003/04
858
0.6310
864
0.6127
797
0.6296
941
0.6115
900
0.5254
- 0.001 / 0.002
- 847 / 87
- 3 / 77
1002 / 144
1715 / 106
2035 / 100
644 / 144
- 250 / 96
- 133 /97
427 / 155
1421 / 90
0.001 / 0.003
- 1119 / 108
- 405 / 97
828 / 179
2113 / 128
1898 / 123
839 / 140
- 246 / 119
285 / 125
113 / 186
1970 / 111
Coefficient / Standard Error
State Appropriations
Govern1
Govern2
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Constant / Intercept
- 0.053 / 0.011
- 620 / 82
- 347 / 73
903 / 129
1901 / 98
2023 / 94
724 / 110
105 / 93
428 / 157
453 / 186
1212 / 96
- 0.008 / 0.004 - 0.034 / 0.010
- 562 / 83
- 516 / 81
- 264 / 74
- 215 / 69
1360 / 134
1305 / 127
2243 / 98
2243 / 93
1845 / 95
1468 / 94
786 / 111
631 / 103
80 / 92
- 71 / 88
380 / 96
302 / 90
466 / 147
397 / 133
1164 / 87
1500 / 91
121
APPENDIX TABLE 4
Student Level Analysis - In-State Tuition Regression
Students Attending Four-Year Institutions
2004 Dollars
1989/90
1992/93
1995/96
Student Observations
Adjusted R-squared
3352
0.1877
5433
0.3215
5695
0.5320
Coefficient / Standard Error
State Appropriations
Expected Family Contribution
Cumulative Grade Point Average
Strongly Regulated
Moderately Regulated
Most Selective
More Selective
Middle Selective
Less Selective
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Female
White
African American
Asian
Hispanic
Constant / Intercept
0.051 / 0.008
0.000 / 0.001
- 0.187 / 2.96
- 226 / 66
- 167 / 74
514 / 167
- 408 / 152
- 367 / 149
- 491 / 157
1091 / 105
552 / 79
621 / 75
150 / 89
48 / 80
- 133 / 85
832 / 94
- 40 / 40
12 / 272
472 / 283
285 / 288
183 / 288
1223 / 324
- 0.039 / 0.007
0.002 / 0.001
1.385 / 0.260
889 / 65
79 / 55
622 / 82
325 / 75
- 86 / 71
- 353 / 86
2541 / 115
1516 / 72
1349 / 78
922 / 83
572 / 76
- 561 / 78
- 923 / 107
- 38 / 120
- 448 / 93
- 262 / 109
- 38 / 120
- 613 / 115
2388 / 157
- 0.037 / 0.005
0.004 / 0.001
1.498 / 0.189
980 / 42
716 / 42
788 / 73
390 / 66
183 / 64
- 183 / 69
3033 / 86
1749 / 50
1307 / 50
690 / 66
153 / 54
- 473 / 62
650 / 74
- 53 / 24
11 / 112
99 / 120
156 / 119
- 204 / 122
1598 / 149
122
APPENDIX TABLE 4 (continued)
Student Level Analysis – In-State Tuition Regression
Students Attending Four-Year Institutions
2004 Dollars
1999/00
2003/04
Student Observations
Adjusted R-squared
5844
0.2848
7669
0.4349
Coefficient / Standard Error
State Appropriations
Expected Family Contribution
Cumulative Grade Point Average
Strongly Regulated
Moderately Regulated
Most Selective
More Selective
Middle Selective
Less Selective
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Female
White
African American
Asian
Hispanic
Constant / Intercept
- 0.006 / .009
0.005 / .002
0.812 / 0.316
494 / 74
222 / 65
902 / 110
209 / 100
- 237 / 96
178 / 111
2396 / 127
1647 / 83
913 / 86
449 / 101
- 150 / 84
- 790 / 91
- 129 / 130
74 / 40
257 / 108
135 / 131
114 / 131
-42 / 131
2869 / 185
- 0.019 / .006
0.004 / .001
0.643 / 0.211
-252 / 59
-122 / 41
1112 / 76
225 / 64
95 / 62
-474 / 76
1571 / 88
1780 / 71
1279 / 73
508 / 81
-468 / 69
-492 / 76
-605 / 86
-5.21 / 27
76 / 66
323 / 80
41 / 91
47 / 84
3953 / 129
123
APPENDIX TABLE 5
Student Level Analysis – Out-of-State Tuition Regression
Students Attending Four-Year Institutions
2004 Dollars
1989/90
1992/93
1995/96
Student Observations
Adjusted R-squared
395
0.4003
1233
0.3191
1209
0.3714
Coefficient / Standard Error
State Appropriations
Expected Family Contribution
Cumulative Grade Point Average
Strongly Regulated
Moderately Regulated
Most Selective
More Selective
Selective
Less Selective
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Female
White
African American
Asian
Hispanic
Constant / Intercept
- 0.104 / 0.050
0.002 / 0.005
23.36 / 16.75
946 / 412
236 / 431
6972 / 988
4824 / 881
3671 / 899
3416 / 960
1334 / 541
- 860 / 492
699 / 423
441 / 514
444 / 451
340 / 523
4179 / 471
- 177 / 221
384 / 1596
2065 / 1637
855 / 1650
- 298 / 1780
- 1549 / 1899
0.068 / 0.048
0.010 / 0.006
- 1.27 / 1.72
2216 / 425
708 / 354
4068 / 571
2583 / 513
949 / 494
- 792 / 609
7592 / 706
2913 / 631
1306 / 587
292 / 607
1243 / 565
57 / 597
2931 / 711
15 / 202
- 432 / 485
- 569 / 597
1269 / 617
- 559 / 737
3733 / 1026
0.067 / 0.045
0.026 / 0.006
0.209 / 1.49
1672 / 403
865 / 340
1584 / 520
1208 / 486
- 837 / 458
- 974 / 566
6674 / 672
1521 / 583
566 / 542
- 143 / 620
44 / 559
- 2102 / 586
518 / 663
22 / 185
- 804 / 603
- 193 / 687
174 / 673
- 1249 / 789
8627 / 1027
124
APPENDIX TABLE 5 (continued)
Student Level Analysis – Out-of-State Tuition Regression
Students Attending Four-Year Institutions
2004 Dollars
1999/00
2003/04
Student Observations
Adjusted R-squared
1035
0.3679
1061
0.4892
Coefficient / Standard Error
State Appropriations
Expected Family Contribution
Cumulative Grade Point Average
Strongly Regulated
Moderately Regulated
Most Selective
More Selective
Selective
Less Selective
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Female
White
African American
Asian
Hispanic
Constant / Intercept
- 0.109 / 0.055
0.048 / 0.009
0.089 / 1.96
654 / 502
368 / 401
4802 / 833
2551 / 775
- 465 / 775
358 / 919
5674 / 663
806 / 611
60 / 571
- 980 / 613
6.35 / 543
- 1399 / 607
1638 / 713
153 / 243
301 / 627
- 859 / 723
1633 / 759
71 / 880
7556 / 1211
- 0.136 / .053
0.015 / .007
3.647 / 1.863
- 140 / 815
- 565 / 374
9925 / 717
8816 / 605
4373 / 609
2936 / 773
3566 / 668
3190 / 637
2986 / 616
- 2127 / 602
471 / 556
-1338 / 637
2241 / 673
- 3 / 233
- 1507 / 633
634 / 698
- 488 / 745
- 218 / 837
6075 / 1209
125
APPENDIX TABLE 6
Student Level Analysis - In-State Tuition Regression
Students Attending Two-Year Institutions
1989/90
1992/93
1995/96
Student Observations
Adjusted R-squared
267
0.4407
398
0.5888
499
0.6479
Coefficient / Standard Error
State Appropriations
Exp. Family Contribution
Cum.Grade Point Average
Strongly Regulated
Moderately Regulated
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Female
White
African American
Asian
Hispanic
Constant / Intercept
- 0.001 / 0.007
0.002 / 0.002
- 0.916 / 2.66
- 143 / 73
158 / 65
563 / 115
575 / 71
288 / 85
219 / 99
40 / 74
- 41 / 103
832 / 94
- 1.8 / 40
- 81 / 312
- 167 / 320
18 / 324
- 116 / 316
586 / 323
0.003 / 0.004
0.005 / 0.002
0.524 / 0.424
- 468 / 123
228 / 104
1570 / 252
1450 / 130
1410 / 137
1593 / 169
450 / 140
- 14 / 160
954 / 232
1.4 / 61
- 19 / 230
- 6.8 / 273
100 / 264
85 / 257
575 / 278
0.008 / 0.018
- 0.002 / 0.003
0.338 / 0.327
- 983 / 77
252 / 63
944 / 1686
964 / 970
710 / 97
63 / 110
- 363 / 96
- 570 / 99
223 / 194
100 / 45
20 / 145
- 14 / 157
- 182 / 182
- 85 / 159
1584 / 182
126
APPENDIX TABLE 6 (continued)
Student Level Analysis - In-State Tuition Regression
Students Attending Two-Year Institutions
1999/00
2003/04
Student Observations
Adjusted R-squared
693
0.5096
3156
0.5829
Coefficient / Standard Error
State Appropriations
Expected Family Contribution
Cumulative Grade Point Average
Strongly Regulated
Moderately Regulated
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Female
White
African American
Asian
Hispanic
Constant / Intercept
0.062 / 0.018
- 0.002 / .003
0.464 / 0.388
- 1033 / 112
- 140 / 96
380 / 255
1720 / 119
612 / 128
558 / 162
- 301 / 129
- 82 / 132
237 / 201
0.865 / 59
239 / 129
21 / 159
428 / 195
95 / 151
1034 / 199
0.044 / 0.009
0.003 / 0.001
0.416 / 0.139
-820 / 36
125 / 35
699 / 86
1298 / 41
585 / 43
883 / 52
225 / 47
-457 / 50
41 / 158
- 40 / 20
- 39 / 52
- 239 / 80
- 130 / 70
- 132 / 84
1346 / 79
127
Appendix Table 7
Descriptive Statistics – IPEDS Institutions
2003 – 2004
Four Year Institutions (N = 432)
Variable
Mean
Std. Dev.
Min
Max
In-State Tuition/Fee
Out-of-State Tuition/Fee
State App / Student
4443.17
11,211.44
5594.72
1494.94
3406.94
2322.65
2032
4254
417.39
13,868
29,064
15,949.48
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.081
0.194
0.725
0.273
0.396
0.447
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.058
0.206
0.433
0.171
0.123
0.234
0.405
0.496
0.377
0.328
0
0
0
0
0
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.074
0.141
0.157
0.111
0.313
0.106
0.053
0.045
0.262
0.349
0.365
0.315
0.464
0.309
0.225
0.197
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
128
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
2003 – 2004
Two Year Institutions (N = 900)
Variable
In-District Tuition
In-State Tuition
Out-of-State Tuition
State App. / Student
Mean
1924.71
2319.62
5287.20
4481.37
Std. Dev.
883.87
1293.37
2483.59
10691.73
Min
48
48
179
0.01
Max
7294
7391
17,630
24,078
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.148
0.176
0.677
0.355
0.381
0.468
0
0
0
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.044
0.099
0.150
0.106
0.309
0.112
0.039
0.141
0.206
0.299
0.357
0.307
0.462
0.316
0.193
0.340
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
129
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1999 – 2000
Four Year Institutions (N = 463)
Variable
Mean
Std. Dev.
Min
Max
In-State Tuition
Out-of-State Tuition
State App. / Student
3483.28
9463.50
6581.12
1154.05
2682.98
2448.23
1611.33
2934.68
1543.58
8841.41
21,719.92
19,160.98
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.149
0.179
0.672
0.357
0.384
0.470
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.069
0.205
0.434
0.177
0.112
0.254
0.404
0.496
0.382
0.316
0
0
0
0
0
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.067
0.136
0.147
0.104
0.287
0.099
0.058
0.102
0.250
0.343
0.354
0.305
0.453
0.299
0.235
0.299
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
130
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1999 – 2000
Two Year Institutions (N = 941)
Variable
Mean
Std. Dev.
Min
Max
In-State Tuition
Out-of-State Tuition
State App. / Student
1768.96
4667.94
5297.11
1148.12
2542.20
9493.90
62.65
137.39
148.71
7175.13
19,740.38
27,315.80
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.143
0.173
0.683
0.351
0.379
0.465
0
0
0
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.043
0.087
0.145
0.099
0.330
0.125
0.034
0.137
0.202
0.282
0.352
0.299
0.470
0.331
0.181
0.338
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
131
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1995 – 1996
Four Year Institutions (N = 460)
Variable
In-State Tuition
Out-of-State Tuition
State App. / Student
Mean
3278.29
8759.35
5923.47
Std. Dev.
1147.41
2814.10
2236.98
Min
1596.06
2708
661.47
Max
8322.40
22,000.75
17,071.23
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.143
0.183
0.674
0.351
0.387
0.469
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.067
0.202
0.437
0.172
0.115
0.251
0.402
0.497
0.376
0.320
0
0
0
0
0
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.070
0.141
0.146
0.102
0.289
0.094
0.059
0.099
0.255
0.349
0.353
0.303
0.454
0.291
0.235
0.297
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
132
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1995 – 1996
Two Year Institutions (N = 797)
Variable
Mean
Std. Dev.
Min
Max
In District Tuition
In-State Tuition
Out-of-State Tuition
State App. / Student
1590.63
1810.70
4597.46
4437.39
802.54
1014.04
2274.73
2366.53
187.91
187.91
722.74
12.66
5173.65
5432.63
15,723.30
28,887.29
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.132
0.208
0.660
0.338
0.406
0.474
0
0
0
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.050
0.108
0.118
0.103
0.325
0.123
0.043
0.130
0.218
0.310
0.323
0.304
0.469
0.329
0.202
0.326
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
133
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1992 – 1993
Four Year Institutions (N = 468)
Variable
Mean
Std. Dev.
Min
Max
In-State Tuition
Out-of-State Tuition
State App. / Student
2934.75
7689.33
5798.13
1058.59
2564.56
2408.77
1151.70
2158.46
595.03
8051.47
19,281.22
18,514.47
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.150
0.179
0.671
0.357
0.384
0.470
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.068
0.205
0.432
0.175
0.111
0.253
0.404
0.496
0.381
0.315
0
0
0
0
0
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.068
0.145
0.141
0.103
0.288
0.096
0.058
0.101
0.253
0.353
0.348
0.304
0 .454
0.295
0.233
0.289
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
134
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1992 – 1993
Two Year Institutions (N = 864)
Variable
Mean
Std. Dev.
Min
Max
In-District Tuition
In-State Tuition
Out-of-State Tuition
State App. / Student
1437.70
1751.86
4431.30
3862.16
784.40
1088.34
2080.41
4921.34
104.46
104.46
143.64
52.10
6254.709
6386.593
15,205.86
13,744.10
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.153
0.189
0.659
0.360
0.391
0.474
0
0
0
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.050
0.105
0.156
0.097
0.289
0.123
0.037
0.143
0.218
0.307
0.363
0.296
0.454
0.328
0.189
0.342
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
135
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1989 – 1990
Four Year Institutions (N = 453)
Variable
Mean
Std. Dev.
Min
Max
In-State Tuition
Out-of-State Tuition
State App. / Student
2455.82
6461.59
6726.88
883.22
2110.15
2859.50
632.98
1787.58
846.62
5986.91
17,562.19
22,704.71
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.152
0.181
0.667
0.360
0.385
0.472
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.068
0.196
0.433
0.181
0.115
0.253
0.398
0.496
0.385
0.319
0
0
0
0
0
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.062
0.152
0.146
0.104
0.276
0.099
0.060
0.101
0.241
0.360
0.353
0.306
0.447
0.299
0.237
0.297
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
136
Appendix Table 7 (continued)
Descriptive Statistics – IPEDS Institutions
1989 – 1990
Two Year Institutions (N = 858)
Variable
In-District Tuition
In-State Tuition
Out-of-State Tuition
State App. / Student
Mean
1221.18
1545.97
3842.89
4456.01
Std. Dev.
721.22
1096.44
1881.91
2335.29
Min
73.26
73.26
146.52
76.72
Max
5778.84
5975.19
13,902.06
23,646.96
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.161
0.186
0.653
0.368
0.390
0.476
0
0
0
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.055
0.106
0.157
0.094
0.283
0.119
0.037
0.149
0.228
0.308
0.364
0.293
0.451
0.324
0.190
0.348
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
137
Appendix Table 8
Descriptive Statistics
2003 – 2004
Four Year In-State Students (N = 7669)
Variable
Mean
Std. Dev
Min
Max
Tuition
Net Tuition
EFC
Inst. Grant
Unmet Need
State App / Student
4738.64
3977.71
12,409.05
760.94
3602.85
6137.98
1542.32
2407.55
14,630.83
2104.18
4948.81
2412.24
1224
-16646
0
0
-35,082
1762.76
11,445
11,295
99,000
21,883
24,700
15,852.95
GPA
Female
White
African American
Asian
Hispanic
Other
293.54
0.553
0.772
0.084
0.041
0.060
0.033
65.02
0.498
0.421
0.287
0.201
0.238
0.120
0
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.051
0.158
0.213
0.075
0.277
0.107
0.072
0.048
0.220
0.365
0.409
0.263
0.448
0.309
0.258
0.214
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.067
0.206
0.727
0.250
0.404
0.445
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.089
0.280
0.493
0.074
0.063
0.285
0.449
0.500
0.262
0.243
0
0
0
0
0
1
1
1
1
1
138
Appendix Table 8 (continued)
Descriptive Statistics
2003 – 2004
Four Year Out-of-State Students (N = 1061)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / Student
12,777.01
10,545.88
14,997.73
2231.13
8003.92
6299.31
5024.84
6429.683
16,906.26
4266.88
9065.11
2378.00
3471
-12,386
0
0
- 22,500
1762.76
26,924
26,924
99,000
22,191
33,346
15,852.95
GPA
Female
White
African American
Asian
Hispanic
Other
297.79
0.475
0.725
0.131
0.072
0.039
0.033
61.88
0.500
0.447
0.338
0.259
0.193
0.103
64
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.082
0.102
0.146
0.128
0.282
0.084
0.121
0.054
0.275
0.303
0.353
0.334
0.450
0.278
0.326
0.226
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.022
0.126
0.853
0.145
0.332
0.355
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.089
0.386
0.425
0.050
0.049
0.285
0.487
0.495
0.218
0.217
0
0
0
0
0
1
1
1
1
1
139
Appendix Table 8 (continued)
Descriptive Statistics
2003 – 2004
Two Year In-State Students (N = 3156)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
1816.37
1495.72
10,022.28
320.65
2680.89
2991.71
861.65
1604.53
14,477.88
1375.82
3583.69
1320.00
378
- 18020
0
0
- 20842
164.00
5192
5079
99,000
20,842
18,163
8652.61
GPA
Female
White
African American
Asian
Hispanic
Other
283.17
0.513
0.679
0.125
0.042
0.114
0.040
73.53
0.500
0.467
0.331
0.202
0.318
0.131
0
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.021
0.176
0.183
0.106
0.260
0.107
0.005
0.141
0.143
0.381
0.387
0.308
0.439
0.310
0.067
0.348
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.221
0.170
0.608
0.415
0.376
0.488
0
0
0
1
1
1
140
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1999 – 2000
Four Year In-State Students (N = 5863)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
4001.33
3435.862
12,379.84
565.47
3496.25
7110.89
1755.67
2283.93
13,056.70
1602.81
3888.81
2553.23
48.36
-14,428.29
0
0
0
1789.543
18,254.36
18,254.36
109,074.40
19,256.77
29,236.87
18,449.29
GPA
Female
White
African American
Asian
Hispanic
Other
288.24
0.545
0.750
0.092
0.061
0.062
0.034
63.67
0.498
0.433
0.289
0.239
0.242
0.176
0
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
0.039
0.113
0.192
0.089
0.270
0.102
0.039
0.193
0.317
0.394
0.285
0.444
0.302
0.194
0
0
0
0
0
0
0
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.190
0.198
0.612
0.392
0.399
0.487
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.148
0.304
0.408
0.092
0.048
0.356
0.460
0.491
0.290
0.213
0
0
0
0
0
1
1
1
1
1
141
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1999 – 2000
Four Year Out-of-State Students (N = 1039)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
10095.41
8005.90
14,509.57
2089.51
6068.22
6899.52
4719.36
5911.59
13,508.54
4190.32
6642.26
2357.53
311.05
-13,155.49
0
0
0
1789.54
25,928.49
21,138.48
84,507.75
26,868.25
54,956.53
18,449.29
GPA
Female
White
African American
Asian
Hispanic
Other
291.24
0.520
0.752
0.108
0.067
0.035
0.037
61.52
0.500
0.432
0.311
0.249
0.184
0.179
62
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.083
0.127
0.191
0.104
0.266
0.097
0.061
0.276
0.333
0.393
0.305
0.442
0.296
0.240
0
0
0
0
0
0
0
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.133
0.117
0.750
0.340
0.321
0.433
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.150
0.432
0.344
0.048
0.026
0.357
0.496
0.475
0.215
0.158
0
0
0
0
0
1
1
1
1
1
142
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1999 – 2000
Two Year In-State Students (N = 693)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
1689.83
1541.30
10,044.03
148.53
2388.89
3969.11
1090.62
1190.88
10,611.92
426.26
2958.26
1852.59
20.88
-1428.87
0
0
0
384.76
8988.69
8988.69
68,484.63
3028.11
13,952.36
12,039.71
GPA
Female
White
African American
Asian
Hispanic
Other
270.09
0.520
0.701
0.098
0.038
0.105
0.058
76.81
0.500
0.458
0.297
0.191
0.307
0.108
0
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
0.017
0.166
0.175
0.063
0.242
0.109
0.031
0.128
0.372
0.380
0.243
0.428
0.311
0.174
0
0
0
0
0
0
0
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.212
0.215
0.573
0.409
0.411
0.495
0
0
0
1
1
1
143
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1995 – 1996
Four Year In-State Students (N = 5533)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
3583.60
3139.38
10,584.41
444.22
3389.82
6851.25
1281.77
1772.23
12,324.46
1337.56
4028.41
2490.19
327.64
-13,014.22
0
0
-16,683.35
1551.85
12,720.30
12,720.30
114,434.50
16,081.06
22,886.91
13,426.33
GPA
Female
White
African American
Asian
Hispanic
Other
278.38
0.542
0.781
0.080
0.074
0.054
0.011
63.49
0.498
0.414
0.271
0.262
0.226
0.075
4
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky
Far West
0.029
0.123
0.214
0.077
0.266
0.084
0.052
0.154
0.168
0.329
0.410
0.267
0.442
0.277
0.223
0.361
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.244
0.147
0.610
0.429
0.354
0.488
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.122
0.306
0.410
0.123
0.039
0.327
0.461
0.492
0.328
0.194
0
0
0
0
0
1
1
1
1
1
144
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1995 – 1996
Four Year Out-of-State Students (N = 1161)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
10042.34
8644.19
14,203.44
1398.15
5882.15
6719.33
3989.99
4988.13
15,024.16
3287.36
6723.89
2518.01
1911.66
-20,537.99
0
0
-15,057.17
2574.19
31,480.34
31,480.34
114,434.50
31,061.14
39,589.53
13,426.33
GPA
Female
White
African American
Asian
Hispanic
Other
280.95
0.490
0.771
0.078
0.093
0.034
0.025
63.41
0.500
0.421
0.268
0.291
0.180
0.132
25
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.071
0.093
0.175
0.102
0.322
0.132
0.062
0.042
0.257
0.291
0.381
0.303
0.468
0.339
0.240
0.200
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.146
0.146
0.707
0.353
0.353
0.455
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.212
0.310
0.363
0.067
0.048
0.409
0.462
0.481
0.249
0.214
0
0
0
0
0
1
1
1
1
1
1
145
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1995 – 1996
Two Year In-State Students (N = 489)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
1677.18
1457.29
7966.49
219.88
2509.36
3591.99
796.33
945.39
8326.47
669.53
3289.73
1492.91
80.71
-3376.42
0
0
-4741.20
692.14
4599.06
3922.09
69,858.06
4818.30
14,456.09
7838.499
GPA
Female
White
African American
Asian
Hispanic
Other
264.90
0.472
0.750
0.115
0.033
0.078
0.025
68.89
0.500
0.434
0.319
0.180
0.268
0.138
43
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountains
Far West
0.023
0.148
0.119
0.096
0.338
0.139
0.015
0.123
0.150
0.355
0.324
0.295
0.473
0.346
0.123
0.329
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.170
0.199
0.632
0.376
0.399
0.483
0
0
0
1
1
1
146
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1992 – 1993
Four Year In-State Students (N = 5433)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
3126.53
2767.51
11,986.69
359.02
5394.53
6536.62
1361.503
1731.696
14,129.09
1180.055
2678.402
2543.673
261.16
- 9474.77
0
0
0
1574.616
14,630.01
14,630.01
117,520.6
18,254.87
16,647.45
15,356.11
GPA
Female
White
African America
Asian
Hispanic
Other
276.21
0.512
0.810
0.069
0.043
0.050
0.028
61.22
0 .500
0.392
0.254
0.203
0.218
0.146
7
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky Mountain
Far West
0.032
0.110
0.173
0.123
0.298
0.108
0.041
0.116
0.175
0.312
0.378
0.329
0.457
0.310
0.198
0.320
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Highly Regulated
0.151
0.169
0.678
0.358
0.375
0.467
0
0
0
1
1
1
More Selective
Most Selective
Middle Selective
Less Selective
Non Selective
0.149
0.316
0.398
0.081
0.055
0.357
0.465
0.490
0.273
0.228
0
0
0
0
0
1
1
1
1
1
147
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1992 – 1993
Four Year Out-of-State Students (N = 1233)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
7339.24
6214.39
14,192.56
1124.85
8091.14
6192.68
4218.65
4964.99
16,925.3
2975.71
4386.17
2414.69
407.40
-13,979.73
0
0
0
1591.75
31,338.83
30,685.94
117,520.6
20,761.98
34,420.48
15,356.11
GPA
Female
White
African American
Asian
Hispanic
Other
283.05
0.514
0.762
0.095
0.067
0.031
0.045
60.53
0.500
0.426
0.293
0.250
0.174
0.205
100
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
MidEast
Great Lakes
Plains
Southeast
Southwest
Rocky
Far West
0.060
0.075
0.149
0.137
0.387
0.092
0.055
0.044
0.238
0.264
0.356
0.344
0.487
0.289
0.229
0.205
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.096
0.112
0.792
0.295
0.315
0.406
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.143
0.384
0.355
0.064
0.055
0.351
0.487
0.479
0.244
0.227
0
0
0
0
0
1
1
1
1
1
148
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1992 – 1993
Two Year In-State Students (N = 398)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
1419.67
1286.59
10,573.95
133.08
4739.52
3744.52
931.17
1088.33
14,022.53
559.16
2634.49
6984.52
78.35
-3879.49
141.02
0
0
758.15
5464.71
5464.71
117,520.60
5223.14
12,988.64
137,369.90
GPA
Female
White
African American
Asian
Hispanic
Other
265.50
0.490
0.808
0.047
0.058
0.067
0.019
73.14
0.502
0.394
0.213
0.235
0.250
0.108
7
0
0
0
0
0
0
400
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky
Far West
0.020
0.134
0.163
0.078
0.276
0.082
0.025
0.222
0.140
0.341
0.370
0.269
0.448
0.275
0.155
0.416
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.248
0.145
0.607
0.433
0.353
0.489
0
0
0
1
1
1
149
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1989 – 1990
Four Year In-State Students (N = 3352)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
1631.16
1483.53
13,669.98
147.6312
1737.48
7160.34
1274.29
1448.20
16,368.16
707.34
4209.32
3015.57
397.08
-15,028.81
1025.66
0
-17099.18
1729.20
36,630.63
36,630.63
144,431.6
17,099.18
41,397
19,908.83
GPA
Female
White
African American
Asian
Hispanic
Other
26.80
0.518
0.845
0.062
0.046
0.042
0.005
6.84
0.500
0.362
0.242
0.210
0.200
0.073
0
0
0
0
0
0
0
40
1
1
1
1
1
1
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky
Far West
0.148
0.221
0.111
0.151
0.113
0.105
0.132
0.356
0.415
0.314
0.358
0.317
0.306
0.339
0
0
0
0
0
0
0
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.218
0.193
0.590
0.413
0.395
0.492
0
0
0
1
1
1
Most Selective
More Selective
Middle Selective
Less Selective
Non Selective
0.102
0.268
0.501
0.110
0.019
0.302
0.443
0.500
0.313
0.138
0
0
0
0
0
1
1
1
1
1
150
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1989 – 1990
Four Year Out-of-State Students (N = 395)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
4291.11
4079.00
19,165.27
212.12
2483.70
7114.31
2717.05
2868.89
22,295.17
980.55
6289.78
2490.57
397.076
-5515.11
1025.66
0
-14,835.4
2745.64
16,996.61
16,996.61
136,431.50
13,157.72
31,997.59
14,946.56
GPA
Female
White
African American
Asian
Hispanic
Other
26.99
0.460
0.809
0.095
0.073
0.019
0.005
6.73
0.499
0.394
0.294
0.261
0.136
0.068
0
0
0
0
0
0
0
40
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky
Far West
0.091
0.153
0.205
0.085
0.242
0.089
0.120
0.043
0.288
0.361
0.404
0.279
0.429
0.285
0.325
0.203
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.192
0.103
0.705
0.394
0.304
0.456
0
0
0
1
1
1
Most selective
More selective
Middle selective
Less selective
Non Selective
0.120
0.361
0.392
0.110
0.017
0.326
0.481
0.489
0.314
0.129
0
0
0
0
0
1
1
1
1
1
151
Appendix Table 8 (continued)
Descriptive Statistics – NPSAS Students
1989 – 1990
Two Year In-State Students (N = 269)
Variable
Mean
Std. Dev.
Min
Max
Tuition
Net Tuition
EFC
Institutional Grant
Unmet Need
State App. / student
692.75
626.84
9092.49
65.90
1426.71
4113.41
413.12
491.45
10,435.99
267.367
3836.84
2974.57
109.89
-1002.21
1025.66
0
-5263.09
453.32
2530.44
2530.44
84,511.26
1667.43
35,650.39
23,634.90
GPA
Female
White
African American
Asian
Hispanic
Other
25.88
0.571
0.816
0.058
0.057
0.065
0.004
7.50
0.496
0.388
0.234
0.232
0.248
0.064
0
0
0
0
0
0
0
40
1
1
1
1
1
1
New England
Mid East
Great Lakes
Plains
Southeast
Southwest
Rocky
Far West
0.043
0.194
0.136
0.075
0.281
0.054
0.063
0.163
0.203
0.397
0.344
0.263
0.450
0.227
0.244
0.370
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
Weakly Regulated
Moderately Regulated
Strongly Regulated
0.159
0.220
0.622
0.366
0.415
0.486
0
0
0
1
1
1
152
REFERENCES
Archibald, Robert B. & Feldman, David H. (2008) Explaining Increases in Higher
Education Costs, The Journal of Higher Education, Vol. 79, No. 3
Baum, Sandy, and Payea, Kathleen (2003) Trends in Student Aid, College Board
Publications. New York.
Baumol, William J. and Sue Anne Batey Blackman (1989) “How to Think about Rising
College Costs” Planning for Higher Education. 23, Summer, 1-7.
Bowen, Howard R. (1980) The Cost of Higher Education: How Much Do Colleges and
Universities Spend Per Student and How Much Should They Spend? San Francisco:
Jossey-Bass. pp. 10-24
Cheslock, J. & Gianneschi (2008) Replacing State Appropriations with Alternative
Revenue Sources: The Case of Voluntary Support. The Journal of Higher Education 79
(2). March/April 2008.
Dill, D.D. (1997) Higher education markets and public policy, Higher Education Policy,
10, 167 – 185.
Dimaggio, Paul J. and Walter W. Powell (1983). The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizations.
Doyle, William R., Delaney, Jennifer A., and Blake Alan Naughton (2003) Public
Institutional Aid and State Policy: Compensation or Compliance? Association for the
Study of Higher Education
Ehrenberg, Ronald G. (2002). Tuition Rising: Why College Costs So Much. Harvard
University Press.
Ehrenberg, Ronald G. (2002). Reaching for the Brass Ring: The U.S. News & World
Report Rankings and Competition. The Review of Higher Education. Winter 2002 Vol.
26, No. 2. pp 145-162.
Garvin, David A. (1980). The Economics of University Behavior, Academic Press. New
York, NY. Pp 17-39.
Gerald, Danette and Haycock, Kati (2007). Engines of Inequality: Diminishing Equity in
the Nation’s Premier Public Universities. The National Education Trust. p. 7.
153
Getz, M., & Siegfried, J. J. (1991). Cost and productivity in American colleges and
universities. In C. Clotfelter, R. Ehrenberg, M. Getz, & J. J. Siegfried (Eds.), Economic
challenges in higher education (261–392). Chicago: University of Chicago Press.
Greene, K.V. (1994) The public choice of non-resident college tuition levels, Public
Choice, 78, 167 – 185.
Griswold, Carolyn P. and Ginger M. Marine. (1996) “Political Influences on State Policy:
Higher Tuition, Higher Aid, and the Real World” Review of Higher Education. 19(4).
Summer. Pp. 361-389.
Hearn J. C., & Griswold, C. P. (1994) State-level centralization and policy innovation in
U.S. postsecondary education. Educational Evaluation and Policy Analysis, 16(2), 161 –
190.
Hearn, Griswold, & Marine (1996). Region, Resources, and Reason: A Contextual
Analysis of State Tuition and Student Aid Policies, Research in Higher Education, Vol.
37, No. 3 1996. pp 241-278.
Hearn, James C. (1998) “The Growing Loan Orientation in Federal Financial Aid Policy”
in ASHE Reader on Finance in Higher Education, 2nd Edition. J. Yaeger, G. Nelson, E.
Potter, J. Weidman, and T. Zullo, eds. Boston: Pearson Custom Publishing.
Hearn, James C. & David Longanecker (1985) “Enrollment Affects of Alternative
Postsecondary Pricing Policies” Journal of Higher Education. 56(5). pp. 485-508.
Heller, Donald. E. (1997) “Student Price Response in Higher Education: An Update to
Leslie and Brinkman” Journal of Higher Education 68(6), Nov./Dec. pp. 624-659.
Heller, Donald E. (2002). “The Policy Shift in State Financial Aid Programs” in Higher
Education: Handbook of Theory and Research. Vol. 17. John C. Smart, ed. New
York:Agathon Press. Pp 221-261.
Hossler, D., Lund, J., Ramin, J., Westfall,, S. & Irish, S. (1997) State funding for higher
education; The Sysyphean task, Journal of Higher Education, 68, 160 – 190.
Hossler, D. & Anderson, D.K.(2004) The Enrollment Management Process. In
Challenging and supporting the first-year student: A Handbook for improving the first
year of college., edited by M.L. Upcraft, J.N. Gardner and B.O. Barefoot. San Francisco:
Jossey-Bass.
154
Johnstone, D. Bruce. (1993). “The ‘High Tuition –High Aid’ Model of Public Higher
Education Finance: The Case Against” paper from a presentation at the 1993 Annual
Meeting of the National Association of Systems Heads.
Johnstone, D. Bruce. (2004) “The Economics and Politics of Cost Sharing in Higher
Education: Comparative Perspectives” Economics of Education Review. 23, pp.403-410.
Leslie, L.L., & Brinkman, P.T. (1987). Student price response in higher education: The
student demand studies. Journal of Higher Education, 58(2);, 181 – 204.
Leslie, L.L. & Ramey G. (1986) State appropriation and enrollment: does enrollment
growth still pay? Journal of Higher Education, 57, 1 – 19.
Levy, Stanley R. (1995). “Sources of Current and Future Funding. New Directions for
Student Services. No. 70. Summer. pp. 39-50.
MacPherson, Michael S., and Morton O. Shapiro (1998) The Student Aid Game: Meeting
Need and Rewarding Talent in American Higher Education. Princeton, NJ. Princeton
University Press.
McKeon, Howard P. (2003) “Controlling the Price of College” Chronicle of Higher
Education. July 11. p. B20.
McKeown, Mary P. (1982) “State Policies on Tuition and Fees for Public Higher
Education” Journal of Education Finance, 8. (Summer 1982): 1-19.
Rhoades, G. (1998) Managed Professionals: Unionized Faculty and Restructuring
Academic Labor. Albany: State University of New York Press.
Selingo, Jeffrey (2003) “The Disappearing State in Public Higher Education” Chronicle
of Higher Education. February 28.
Slaughter, Sheila, and Larry Leslie (1997) Academic Capitalism: Politics, Policies, and
the Entrepreneurial University. Baltimore: Johns Hopkins University Press.
Slaughter, Sheila, and Gary Rhoades (2004) Academic Capitalism and the New Economy.
Baltimore: Johns Hopkins University Press.
Tolbert, Pamela S. (1985) “Institutional Environments and Resource Dependence:
Sources of Administrative Structure in Institutions of Higher Education.” Administrative
Science Quarterly, 30.
155
Trow, Martin A. (1984) “The Analysis of Status,” in Burton R. Clark (Ed.), Perspectives
on Higher Education: Eight Disciplinary and Comparative Views. Los Angeles: The
University of California Press.
US College Board, Trends in Higher Education Series, 2007.
Weertz, David J., and Ronca, Justin M (2006) Examining Differences in State Support
for Higher Education: A Comparative Study of State Appropriations for Research I
Universities, The Journal of Higher Education, 77(6), 935 – 967.
Wei, C.C., Li, X., and Berkner, L. (2004). A Decade of Undergraduate Student Aid:
1989-90 to 1999-2000 (NCES 2004-158). U.S. Department of Education, National Center
for Education Statistics. Washington, D.C: U.S. Government Printing Office.
Wellman, Jane V., Desrochers, Donna M. and Colleen M. Lenihan (2008). The Growing
Imbalance: Recent trends in U.S. postsecondary education finance. Delta Project on
Postsecondary Education Costs, Productivity and Accountability. Washington, D.C. pp
14-17.
Wilkinson, Rupert (2004) Aiding Students, Buying Students: Financial Aid in America.
Vanderbilt University Press.
Winston, Gordon C. (1999). “Subsidies, Hierarchies, and Peers” Journal of Economic
Perspectives. 13(1). Pp. 13-36.
Wolanin, Thomas R. (1998) “Pell Grants: A 25-Year History” in Memory, Reason,
Imagination: A Quarter Century of Pell Grants. L. Gladieux, ed. Washington, D.C.:
College Board. pp. 13-31.
Zemsky, Robert (1994). Faculty Discretionary Time: Departments and the Academic
Ratchet. Journal of Higher Education 65, 1 – 22.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement