Higher Education: Tuition Increases and Colleges' Efforts to Contain
Costs (Letter Report, 09/30/1998, GAO/HEHS-98-227).

On a percentage basis, tuition has risen faster at both four-year and
two-year public colleges and universities than it has at four-year
private institutions--30 percent versus 17 percent in the past five
years. However, four-year private schools, which had much higher tuition
costs to begin with, had dollar increases that were greater than those
of four-year and two-year public schools. The relative size of a school,
whether public or private, appeared to have no relationship to the rate
of increase. At both four-year public and private schools, the size of
tuition increases was statistically linked to several other financial
variables, including schools' revenues from grants, contracts, and gifts
and changes in schools' costs to provide education. In the case of
public schools, changes in government appropriations for higher
education were also a factor. Tuition rises at community colleges have
closely paralleled increases at four-year public schools in recent
years. States, consortiums of schools, and individual colleges and
universities have all taken steps to cut schools' operating costs in
recent years--from eliminating academic departments to boosting class
sizes to privatizing such operations as bookstores and food services.
However, these measures may not necessarily lead to lower tuition for
students because schools' costs may be only one of several factors
considered when decisions on setting tuition are made.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-98-227
     TITLE:  Higher Education: Tuition Increases and Colleges' Efforts
	     to Contain Costs
      DATE:  09/30/1998
   SUBJECT:  Community colleges
	     Teacher salaries
	     Education or training costs
	     Student financial aid
	     Inflation
	     Higher education
	     College students
	     Colleges and universities
IDENTIFIER:  Higher Education Price Index

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************
GAO/HEHS-98-227

Cover
================================================================ COVER

Report to the Honorable
Charles E.  Schumer
House of Representatives

September 1998

HIGHER EDUCATION - TUITION
INCREASES AND COLLEGES' EFFORTS TO
CONTAIN COSTS

GAO/HEHS-98-227

Cost of College Update

(104910)

Abbreviations
=============================================================== ABBREV

  CPI - Consumer Price Index
  FIPSE - Fund for the Improvement of Postsecondary Education
  FTE - full-time-equivalent
  HEPI - Higher Education Price Index
  IPEDS - Integrated Postsecondary Education Data System
  MHEC - Midwestern Higher Education Commission
  NACUBO - National Association of College and University Business
     Officers
  TWU - Texas Woman's University

Letter
=============================================================== LETTER

B-279220

September 30, 1998

The Honorable Charles E.  Schumer
House of Representatives

Dear Mr.  Schumer:

The rising cost of tuition at the nation's colleges and universities
has been a topic of considerable attention.  Over the past two
decades, tuition increases have substantially outstripped rises in
median family income and the cost of living.  In our study on tuition
increases (including related fees) at 4-year public colleges and
universities from school year 1980-81 through 1994-95, we found that
the two major factors associated with these increases were the rise
in schools' expenditures and schools' need to increase tuition
revenue to make up for smaller increases in state appropriations.\1 A
subsequent study by the National Commission on the Cost of Higher
Education covering fiscal years 1987 through 1996 had similar
findings and made a number of recommendations on the actions that
could be taken such as strengthening institutional cost control to
keep college affordable for anyone wishing to obtain a postsecondary
education.\2

You asked us to provide information about various issues regarding
college tuition increases and what schools have done to keep down
their costs.  More specifically, you asked that we address the
following four questions:

  -- To what extent have tuition increases varied by type and size of
     school?

  -- To what extent is there a statistical relationship between
     increases in tuition and other factors such as cost increases at
     the schools?

  -- To what extent have tuition increases at 4-year schools affected
     tuition at community colleges?

  -- What are some examples of strategies that schools have employed
     to reduce their own costs?

Our information covers community colleges and 4-year public and
private colleges and universities.\3 The data we used are from the
Department of Education and cover the most recent 5-year period
available.\4 The tuition and related fee figures (which we refer to
simply as "tuition") we use reflect schools' full prices that have
not been adjusted for discounts or other forms of student financial
aid.  We conducted a statistical analysis using multiple regression
techniques to determine the factors associated with variations in
tuition increases.\5 In addition to analyzing the data, we
interviewed various school officials and experts knowledgeable about
college finance and policy issues, and we reviewed published
research.  Appendix I describes our scope and methodology in further
detail.

--------------------
\1 Higher Education:  Tuition Increasing Faster Than Household Income
and Public Colleges' Costs (GAO/HEHS-96-154, Aug.  15, 1996).

\2 National Commission on the Cost of Higher Education, Straight Talk
About College Costs & Prices (Phoenix, Ariz.:  Oryx Press, 1998).

\3 We excluded 2-year private schools from this analysis because they
constitute a very small segment of the total postsecondary education
community.

\4 For our analysis of variation by type and size of school, this
period was school years 1990-91 to 1995-96; for our analysis of the
relationship between tuition and other factors, which required
somewhat more detailed information than was available for 1995-96,
the period began 1 year earlier, in 1989-90, and ended with 1994-95.

\5 These methods did not identify the causes of tuition increases;
they identify characteristics of schools with larger or smaller
tuition increases.

   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

On a percentage basis, tuition has risen faster at both 4- and 2-year
public colleges and universities than at 4-year private schools--30
percent versus 17 percent in the past 5 years.\6 However, 4-year
private schools, which had much higher tuitions to begin with, had
dollar increases that were greater--$1,763 at 4-year private schools
versus $670 at 4-year public schools and $288 at community colleges.
The relative size of a school, whether public or private, appeared to
have no relationship to the rate of increase.  Large schools tend to
have higher tuitions than small schools, and therefore, although the
increase was greater in dollar terms at large schools, the percentage
increases were about the same.

At both 4-year public and private schools, our analysis showed that
the size of tuition increases was statistically related to a variety
of other financial variables such as schools' revenues from grants,
contracts, and gifts; for public schools, changes in government
appropriations for higher education; and changes in schools' costs of
providing education including noninstructional costs.  For example,
4-year public schools that experienced larger reductions in
government appropriations tended to have larger tuition increases.
Likewise, schools that experienced greater losses in grant revenues
or greater increases in operating costs also tended to contribute to
increase tuitions more than those schools that did not have these
experiences.

Tuition increases at community colleges have closely paralleled
increases at 4-year public schools in recent years.  However, there
is little evidence that tuition increases at community colleges have
been caused by increases at 4-year schools, according to officials
and researchers we contacted and the literature we reviewed.
Instead, the increases at both 4-year public schools and community
colleges are similar either because they are set legislatively or
administratively on a statewide basis or because the tuition-setting
entities consider the same kinds of factors such as the Consumer
Price Index (CPI) or state funding levels when deciding on tuition
increases for both kinds of schools.

States, consortiums of schools, and individual colleges and
universities have all taken steps to cut schools' operating costs in
recent years.  In our review of published reports and discussions
with education officials, a variety of cost-cutting strategies
surfaced, such as eliminating degree programs or academic
departments, increasing class sizes, streamlining administrative
operations, and privatizing operations like bookstores and food
service operations.  While these actions serve to reduce schools' own
costs, they may not necessarily result in lower tuition for students
because schools' costs may be only one of several factors considered
when tuition-setting decisions are made.

--------------------
\6 These percentages are based on adjusting all dollar amounts to
1995 dollars to account for inflation.  All dollar amounts shown in
this report are adjusted to the last year of the 5-year period
analyzed.

   BACKGROUND
------------------------------------------------------------ Letter :2

During the 5-year period from school years 1990-91 through 1995-96,
the average tuition for a full-time resident undergraduate student
rose 43.8 percent, compared with an increase of 15.4 percent in CPI
and 13.8 percent in median household income in current dollars
(dollars that are not adjusted for inflation).  (See fig.  1.) This
difference continued a pattern that has been apparent for a much
longer period.  From school year 1980-81 through 1994-95, for
example, the average tuition for a resident undergraduate to attend
school increased by about 234 percent.  During approximately the same
period, the cost of living rose 74 percent and median household
income rose 82 percent.

   Figure 1:  Cumulative
   Percentage Increases in
   Tuition, Median Household
   Income, and Consumer Prices,
   School Years 1990-91 Through
   1995-96

   (See figure in printed
   edition.)

Note:  This comparison was made using data that are in dollars not
adjusted for inflation.

Sources:  Tuition and fee data are from the Department of Education;
median household income data are from the U.S.  Census Bureau; and
consumer price data are from the Bureau of Labor Statistics.

The topic of college affordability has become an issue of growing
concern not only for students and their parents but also for both the
state and federal governments.  Congressional committees have held a
number of hearings on this issue and, in 1997, the Congress
established the National Commission on the Cost of Higher Education
to conduct a comprehensive review of college costs and prices.  In
its final report, issued in January 1998, the Commission stated that
its members were convinced that if colleges and universities did not
take steps to reduce their costs, federal and state policy makers
would intervene and take up the task for them.

   TUITION INCREASES VARIED BY
   TYPE AND SIZE OF SCHOOL
------------------------------------------------------------ Letter :3

During the most recent 5-year period (school years 1990-91 through
1995-96), the rise in tuition was highest in percentage terms at
public schools and highest in dollar terms at private schools.
Tuition at public 2-year and 4-year schools rose 30 and 31 percent,
respectively, compared with 17 percent at private 4-year schools.
Private 4-year school tuition, which is generally much higher, had
greater dollar increases--$1,763 compared with $670 and $288 for
public 4-year and 2-year schools, respectively (see table 1).

                                Table 1

                Comparison of Average Undergraduate In-
                State Tuition and Fee Increases by Type
                  of School, School Years 1990-91 and
                                1995-96

                                                           Average
                                           Average         tuition
                                          tuition\a        increase
                                        --------------  --------------
                                         1990-   1995-  Amount  Percen
Type of school                              91      96      \a       t
--------------------------------------  ------  ------  ------  ------
4-year private                          $10,48  $12,24  $1,763     17%
                                             0       3
4-year public                            2,178   2,848     670      31
2-year public                              951   1,239     288      30
----------------------------------------------------------------------
\a Dollar amounts are in constant 1995-96 dollars.

A school's size did not appear to have a substantial effect on the
percentage increase in tuition.  To determine the degree to which
tuition increases were related to a school's size, we divided each
type of school into quartiles according to fall 1995-96 enrollment
levels.  We did this for each type of school--public 2-year and
4-year and private 4-year.  For each type of school, comparisons
across the four quartiles showed that the percentage increases in
tuition were about the same.

Although tuition rose at about the same percentage increase across
the different sizes of schools, size did have some bearing on the
dollar amount of the increases.  For all kinds of schools, tuition
tends to be somewhat higher at schools with higher enrollments.  Even
though the percentage increase was generally about the same, larger
schools saw higher dollar increases (see table 2).

                         Table 2

          Tuition Increases, by Type and Size of
          School, School Years 1990-91 and 1995-
                            96

                           Amount of tuition increase
                        --------------------------------
Enrollment quartile (1    4-year        4-year    2-year
being the largest)        public       private    public
----------------------  --------  ------------  --------
1                           $718        $1,991      $359
2                            675         1,691       345
3                            651         1,594       314
4                            522         1,293       252
--------------------------------------------------------
For example, private 4-year schools in the quartile with the largest
enrollments (quartile 1) had an average 1990-91 tuition of $11,327,
compared with $6,925 for schools in the quartile with the smallest
enrollments.  Although tuition rose by about the same percentage in
both quartiles, tuition increased by $1,991 for the largest schools
compared with $1,293 for the smallest schools.  Appendix II provides
more detailed information.

   TUITION INCREASES AND CERTAIN
   FINANCIAL FACTORS ARE RELATED
------------------------------------------------------------ Letter :4

The size of a school's tuition increase is associated with certain
aspects of the school's financial situation, our analysis found.  To
determine what factors might help account for variations in tuition
increases, we conducted a statistical analysis using multiple
regression techniques.  Multiple regression is a method for
determining the existence and strength of associations between one
variable--in this case, the amount of the tuition increase--and
various other factors.  These methods did not identify the causes of
tuition increases; they identify variables related to larger or
smaller tuition increases.  It is particularly difficult to identify
factors causing variation in tuition increases because schools'
decisions to increase tuition and to increase expenditures are likely
to be interdependent.  For example, one study pointed out that
schools with larger tuition increases may increase spending more
because they have more income and because they may need to increase
instructional expenditures more to maintain adequate enrollment.\7 In
addition, the study reported that it is likely that schools'
decisions to increase tuition and increase institutional financial
aid may be interdependent, and it also suggested that larger
increases in tuition could lead to larger increases in institutional
financial aid, both because a school had more income and because it
desired to maintain a diverse student body.

We examined a judgmentally selected set of 15 factors derived from
information the Department of Education obtains annually in surveys
of all postsecondary institutions eligible for federal education
programs.  This information includes a variety of financial trends
(such as increases in instructional costs or school endowments) as
well as other institutional characteristics (such as the ratio of
in-state tuition to average tuition receipts per student).  We
focused this analysis on tuition (and required fees) for in-state
full-year, full-time undergraduate students at 4-year public and
private schools over the period covering school years 1989-90 and
1994-95 (the latest years for which information was available).  We
adjusted tuition and all financial variables for inflation using the
Bureau of Labor Statistics' CPI.  For the list of all 15 variables
and details concerning our analyses and results, see appendix III.

In general, the size of tuition increases at both 4-year public and
private schools was most strongly related to the amount of change in
school revenue from sources other than tuition and schools' costs of
providing education, including noninstructional costs.  Changes in
the market value of schools' endowments were not related to tuition
changes.  Because the results differed to some degree between public
and private institutions, we discuss each individually.

--------------------
\7 M.S.  McPherson and M.O.  Schapiro, Keeping College Affordable:
Government and Educational Opportunity (Washington, D.C.:  Brookings
Institution, 1991).

      INCREASES AT 4-YEAR PUBLIC
      COLLEGES AND UNIVERSITIES
---------------------------------------------------------- Letter :4.1

For 4-year public schools, the 15 factors collectively accounted for
about 85 percent of the variation in tuition increases.  Seven
important factors, which together accounted for about half of the
variation, were as follows:\8

  -- Change in government appropriations.  On average, 4-year public
     schools saw declines in governmental appropriations (federal,
     state, and local) during the period--about $6,200 per
     full-time-equivalent (FTE) student in 1994-95, compared with
     about $7,100 in 1989-90.\9 Schools that saw the greatest
     declines in appropriations per FTE student typically had larger
     increases in tuition.  These appropriations included
     unrestricted funding as well as appropriations for restricted
     purposes, such as particular building projects.  These did not
     include, however, federal appropriations to fund grants to
     individual students.

  -- Change in instruction expenditures.  Schools that had the
     largest increases in instructional costs per FTE student tended
     to have larger tuition increases after adjusting for the other
     factors we examined.  Schools varied considerably in the extent
     to which their instruction costs changed.  Some schools had
     increases while others had decreases.  As a result, instruction
     costs per student changed little over the period.  They were
     about $5,300 in both 1989-90 and 1994-95.

  -- Change in revenue from grants, contracts, and other sources.
     These "other" revenues, which totaled about $3,700 per
     full-time-equivalent student in 1994-95, include such nontuition
     items as grants and contract revenue from government and private
     sources.  Schools that had larger increases in these revenues
     typically had smaller increases in tuition after taking into
     account the other factors analyzed.  The average amount of these
     revenues increased from about $3,400 to about $3,700 from
     1989-90 to 1994-95.

  -- Change in the amount by which revenues exceeded or fell short of
     expenditures.  This is the amount by which a school's revenue
     exceeded or fell short of expenditures for current operations,
     which excludes revenue and expenditures for such things as
     noninstitutional grants to students, independent operations,
     auxiliary enterprises, and hospitals.\10

On average, public schools saw little or no change in this factor.
On average, they had revenues exceeding expenses (a surplus) of about
$160 per student in both years.  Schools that had larger decreases in
the annual current fund surplus, or increases in their deficit,
tended to have larger tuition increases after taking into account the
other factors analyzed.

  -- Change in other student-related expenditures.  These
     expenditures were for the portion of "overhead" expenditures
     proportionally allocated as student-related expenses
     (instruction and student services) versus research and public
     service.  Such "overhead" expenses included expenditures for
     libraries, museums, galleries, academic computing support,
     academic administration, general administrative services, legal
     and fiscal operations, public relations, operation and
     maintenance of grounds and facilities, funds for debt
     retirement, and improvements or additions to facilities.  On
     average, institutions spent about $3,000 per FTE student in this
     category in 1989-90 and about $2,800 in 1994-95.  Schools that
     had the largest decreases in this category typically had smaller
     increases in tuition.

  -- Change in nonstudent-related expenditures.  These other
     expenditures include schools' spending on public service and
     overhead expenditures proportionally allocated to research and
     public service rather than student services and instruction.\11
     These expenditures averaged about $2,100 in 1989-90 and about
     $2,400 in 1994-95.  Schools with larger increases in this
     category were more apt to have larger tuition increases after
     controlling for other factors (that is, after taking into
     account the relationship between tuition increases and the other
     variables we studied).

  -- Change in research expenditures.  One of the missions of many
     4-year colleges and universities is research.  Schools report to
     the Department of Education amounts of their expenditures on
     research commissioned by an external agency or budgeted within a
     school.  Research expenditures increased from about $2,000 to
     about $2,100 per student from 1989-90 to 1994-95.  After
     controlling other factors, including changes in all current
     revenue, schools with larger increases in research expenditures
     were more apt to have larger increases in tuition.

--------------------
\8 See appendix III for a more detailed discussion of results for all
factors analyzed.  We identified variables as "important" here if the
decline in the portion of variation accounted for (R squared) by
omitting the variable from the model was greater than the decline
observed from omitting other variables, one at a time.  However, when
the independent variables are highly correlated, it is difficult to
distinguish their independent contribution to the portion of
variation accounted for (R squared).  Some of the independent
variables we examined were highly correlated.  For example, the
change in revenue from grants, contracts, and other sources had a
correlation of 0.76 with the change in research expenditures in the
sample of public 4-year schools.

\9 The amounts are per FTE student at schools included in our
regression analysis, weighted by each school's estimated FTE student
count.

\10 The Department of Education's Integrated Postsecondary Education
Data System (IPEDS) surveys, from which we obtained these data, allow
schools to account separately for current funds activity (educational
and general revenue expenditures and transfers) and for activity in
other funds such as endowment funds, funds for retirement of plant
debt, and funds for renewals and replacements of facilities.

\11 Schools' spending on public service is for noninstructional
services beneficial to groups external to the institution such as
seminars and projects provided to sectors of the local community and
schools' cooperative extension services funded by the Department of
Agriculture.

      INCREASES AT 4-YEAR PRIVATE
      COLLEGES AND UNIVERSITIES
---------------------------------------------------------- Letter :4.2

Our analysis for 4-year private schools included the same 15 factors
used in analyzing public schools.\12 Together, these factors
accounted for about 70 percent of the variation in tuition changes at
private 4-year schools.  Most of the factors that helped account for
the variation were the same as those for 4-year public schools.  For
private 4-year schools, the seven factors that accounted for most of
the variation were

  -- Change in revenue from grants, contracts, and other sources.
     Schools that had large decreases in these revenues typically had
     larger increases in tuition.  These revenues averaged about
     $7,000 per FTE student in 1994-95, up from about $6,300 in
     1989-90.

  -- Change in the amount by which revenues exceeded or fell short of
     expenditures.  Although private 4-year schools on average saw
     little, if any change in these amounts, schools that had a large
     increase tended to have larger tuition increases.  Conversely,
     schools that had smaller increases tended to have smaller
     tuition increases.  On average, schools with smaller increases
     had a $200 decrease per FTE student in 1989-90 and a $300
     decrease per FTE student in 1994-95.

  -- Change in other student-related expenditures.  These
     expenditures averaged $5,700 per student in 1994-95, up from
     about $5,600 in 1989-90.  Schools that had the greatest
     increases (or the smallest decreases) in this category typically
     had larger increases in tuition.

  -- Change in instruction expenditures.  Unlike public schools,
     where instruction costs changed little over the period, private
     4-year schools showed an increase in instructional costs during
     the 5-year period after adjusting for inflation.  On average,
     the private 4-year schools we studied spent $7,000 per student
     in 1994-95, up from $6,200 in 1989-90.  We found that schools
     with larger increases in instructional costs (and those that had
     smaller decreases) per FTE student tended to have larger tuition
     increases.

  -- Change in the ratio of undergraduate tuition to average tuition
     received per FTE student.  Some schools increased in-state
     undergraduate tuition charges more than other tuition charges,
     such as tuition for graduate or professional schools.  The
     schools that did so saw larger increases in undergraduate
     tuition.  On average, in-state undergraduate tuition was close
     to the average amount of tuition that schools reported receiving
     per FTE student in both 1989-90 and 1994-95.

  -- Change in research expenditures.  Schools with larger increases
     in research expenditures were more likely to have larger
     increases in tuition after controlling for other factors.  This
     does not mean that tuition increases with additional research,
     because much research is paid for by other revenue.\13 We were
     not able to identify how much of each school's "other" revenue
     was for research.  Without that information, we could not
     determine whether research activity (including associated costs
     and revenue) was related to changes in tuition.  Our analysis
     indicated only that research costs were related to tuition
     increases after controlling for other factors, including revenue
     and other costs.  The average amount of research expenditures
     per student increased from about $1,300 to about $1,500 from
     1989-90 to 1994-95.

  -- Change in other nonstudent-related expenditures.  On average,
     these other expenditures increased from about $2,900 to about
     $4,100 per student over the 5-year period.  Schools that saw
     larger increases in these expenditures saw larger tuition
     increases.

--------------------
\12 See appendix III for a more detailed discussion of results for
the 15 factors analyzed.

\13 Without controlling for other factors, changes in research were
not correlated with changes in tuition to a statistically significant
extent.

   EFFECT OF 4-YEAR SCHOOL TUITION
   INCREASES ON COMMUNITY COLLEGE
   TUITION PRICES
------------------------------------------------------------ Letter :5

Although community college tuition increases have generally
paralleled those of 4-year public schools, in our review of the
literature and discussions with officials in the higher education
community we found little evidence of any direct causative
relationship between the two, at least on any widespread basis.  That
is, there is little evidence that indicates that community colleges,
in general, raised tuition just because 4-year schools did.

In some states, however, there is some commonality in the
tuition-setting policies and practices of both 4-year and 2-year
public schools, and this can occur in a variety of ways.  The
following examples illustrate:

  -- In Alaska, the Alaska Board of Regents sets tuition for the
     state's 4-year and 2-year public schools and it applies the same
     policy to both types of schools.  The Board's policy is to
     adjust schools' tuition annually on the basis of the average of
     the prior 3 years' Higher Education Price Index (HEPI).\14 The
     Board can also suspend this policy and adjust tuition by other
     amounts.

  -- In Nevada, the Nevada Board of Regents sets the tuition rates
     for both 4-year and 2-year public schools on the basis of
     recommendations by a committee composed of elected student
     government representatives, college presidents, and university
     system administration staff.  In its deliberations, the
     committee is to consider a common set of factors for both kinds
     of schools.  These factors include charges at peer institutions,
     changes in HEPI, and the cost of education.

  -- In Washington State, the legislature sets tuition for both
     4-year and 2-year public schools, and it increased tuition for
     both kinds of schools by 4 percent for each of the 2 years in
     the 1997-99 biennium budget.

Thus, the tuition increases for public 2-year schools in states such
as these are linked in one way or another to the tuition increases of
4-year schools.  However, even in these states it cannot be concluded
that 2-year school tuition increases were caused by tuition increases
at 4-year schools.

--------------------
\14 HEPI is an index specifically designed to measure changes in the
prices of goods and services commonly purchased by higher education
institutions.  Items that HEPI measures include faculty and
administrators' salaries, fringe benefits, communication and data
processing services, supplies and materials, library acquisitions,
and utilities.

   EXAMPLES OF SCHOOL EFFORTS TO
   REDUCE COSTS AND IMPROVE
   EFFICIENCY AND EFFECTIVENESS
------------------------------------------------------------ Letter :6

Many schools are taking a wide variety of actions to reduce or
contain their own costs or improve the efficiency and effectiveness
of their operations.  In our 1996 study on the cost of college, we
reported on examples of approaches taken by states and schools to
ease the college cost burden to students.\15 For example, we found
some initiatives that focus directly on charges students incur, such
as limiting the amount of tuition increases, and other initiatives
such as prepaid tuition plans that provide alternative ways of
helping students or their parents pay for college.

To identify additional examples of schools' cost containment
initiatives, we conducted a search of publications in the field, and
we contacted researchers and representatives of schools and major
higher education associations such as the American Council on
Education.  In some cases, the actions we identified were being
driven by state-level initiatives.  In other cases, schools had
initiated efforts either on their own or in conjunction with other
schools.

Schools have initiated cost-cutting initiatives for a variety of
reasons, such as to compenstate for cuts in state funding, to
reallocate resources from one department or area of activity to
another, or simply just to improve the economy and efficiency of
their operations.  While these kinds of activities help schools
reduce their own costs, they do not necessarily result in lower costs
to students because, in some states, schools' costs are only one of
several factors considered when tuition-setting decisions are made.

--------------------
\15 GAO/HEHS-96-154, Aug.  15, 1996.

      RESPONSES TO STATE-LEVEL
      INITIATIVES
---------------------------------------------------------- Letter :6.1

In some states, such as Ohio, Oregon, Virginia, and Wisconsin, the
governor, legislature, or board of regents initiated efforts to
contain costs or improve efficiency and quality in state-funded
colleges and universities.  For example, in 1994, Virginia's Governor
and General Assembly asked the state's public colleges and
universities to submit restructuring plans to improve quality and
reduce costs.  In subsequent years, the schools responded by taking
or planning a wide variety of actions such as reducing the number of
credits needed to graduate from most bachelor's degree programs to
120 credits, closing or merging a number of degree programs,
increasing faculty productivity by teaching more students with fewer
instructors, and privatizing various services such as food service
and bookstores.  In 1996, these actions had a value of $63.3 million
in cost savings, and once all plans are fully implemented, the
schools estimate financial benefits to be about $110 million
annually.  An official of the Virginia Council of Higher Education
said this amounts to about 5 to 6 percent of the state's total
educational and general appropriations for its postsecondary
education system, which is almost $2 billion.

      ACTIONS INITIATED BY SCHOOLS
---------------------------------------------------------- Letter :6.2

Some schools have hired consulting firms to help them identify cost
savings opportunities or otherwise become efficient.  For example, in
our discussions with representatives of five of the largest
accounting firms, we found that four of them have departments or
persons who specialize in helping schools develop strategies to
contain costs or improve efficiency.  These firms help schools
develop strategic plans and institutional goals and evaluate
opportunities for enhancing efficiencies.  They also assist schools
in redesigning and reengineering their business processes and
information systems.  According to one firm, it has worked with over
two dozen colleges and universities, medical schools, and state
education agencies.

Since 1994, the National Association of College and University
Business Officers (NACUBO), together with Barnes and Noble
Bookstores, Inc., have sponsored a program that recognizes schools
that have improved quality and reduced costs.  The Higher Education
Award program is fashioned after the Malcolm Baldrige National
Quality Management program for businesses.  It grants awards of
$2,500 to $10,000 to schools submitting the best proposals for
improving quality or reducing costs.  For example, Loyola College in
Maryland was one of six winners of the 1997 Higher Education Award.
Its physical plant department improved operational efficiency and
reduced expenses by forming self-directed work teams to assist with
management functions.  The end result was a decrease in overtime and
an increase in the housekeeping staff's duties by 50 percent and more
than $625,000 in annual savings.

A recent study by RAND, a private nonprofit research institution,
found that some schools have joined together to undertake
cost-savings initiatives.  For example, 17 independent nonprofit
colleges and universities in Virginia--the Council of Independent
Colleges in Virginia--joined together to purchase a contract with a
long-distance network carrier.  As a group, the schools expect to
save $4.65 million over the 3-year period of the contract.\16

--------------------
\16 Tessa Kaganoff, Collaboration, Technology, and Outsourcing
Initiatives in Higher Education:  A Literature Review (Chicago:  The
Foundation for Independent Higher Education, 1998).

      DEPARTMENT OF EDUCATION
      INITIATIVES
---------------------------------------------------------- Letter :6.3

According to a Department of Education official, the Department does
not have any programs that deal specifically with cost containment or
efficiency improvement activities at postsecondary institutions.
However, it operates a grant program, the Fund for the Improvement of
Postsecondary Education (FIPSE), that offers schools funding for
projects that will help them improve and become more efficient.\17
The University of Delaware, for example, is using a FIPSE grant to
establish a longitudinal database on instructional costs and
productivity to compare these factors among many institutions.  Such
information will eventually help schools identify departments,
programs, or other areas where cost-savings opportunities are
available.

For fiscal year 1998, the FIPSE grant competition specifically
addresses cost control at postsecondary institutions.  For the
purposes of the competition, FIPSE defines cost control as reducing
the total amount that is actually spent by a school to bring a given
number of students up to the level of postsecondary knowledge and
skill that was achieved by comparable groups in previous years or as
lowering the annual rate of increase of such costs significantly
below its previous level.  The Department awarded FIPSE grants to
nine institutions that proposed projects to demonstrate innovative
ways to reduce costs.  Brief descriptions of these projects are
contained in appendix V.

For more specific examples of schools' cost containment and
efficiency improvement efforts, see appendix IV.

--------------------
\17 Created by the Higher Education Amendments of 1972, FIPSE
provides grants intended to support innovative educational reform
projects that can serve as national models for the improvement of
postsecondary education.  From fiscal year 1994 to 1998, $83.9
million in FIPSE grant funds were used for over 1,100 new and
continuing projects.

   AGENCY COMMENTS
------------------------------------------------------------ Letter :7

The Department of Education provided comments on a draft of this
report.  The comments are included as appendix V.  The Department
said that the report demonstrates that public school tuition and fees
have increased primarily in response to reductions in state funding
for higher education and that the report provides helpful examples of
how states and schools are reducing costs.  The Department also
provided technical suggestions that we incorporated as appropriate.

---------------------------------------------------------- Letter :7.1

Copies of this report are being sent to the Chairman and Ranking
Minority Member of the Committee on Education and the Workforce of
the House of Representatives; Chairman and Ranking Minority Member of
the Committee on Labor and Human Resources of the U.S.  Senate; the
Secretary of Education; other appropriate congressional committees
and Members; and others who are interested.

If you have any questions about this report, please call me or Joseph
J.  Eglin, Jr., Assistant Director, at (202) 512-7014.  Major
contributors to this report include Charles M.  Novak, Benjamin P.
Pfeiffer, Charles H.  Shervey, and Dianne L.  Whitman-Miner.

Sincerely yours,

Carlotta C.  Joyner
Director, Education and
 Employment Issues

SCOPE AND METHODOLOGY
=========================================================== Appendix I

For this study, we used annual tuition and school financial and
enrollment data collected by the Department of Education from 4-year
private and public colleges and universities and community colleges.
We did not include 2-year private schools in this study because they
constitute a very small segment of the total number of schools
providing postsecondary education.  We also excluded proprietary
(for-profit trade) schools because, unlike the schools included in
our study, they operate on a for-profit basis.  For our various
analyses, we used annual tuition and related fees for full-time
in-state undergraduate students.  We did not include data on other
costs that students may incur while attending college such as for
room and board, books and supplies, and commuting.

To determine the extent to which tuition increases from school years
1990-91 through 1995-96 varied by type of school, we used average
undergraduate in-state tuition and fee data published by the
Department of Education in the 1997 Digest of Education Statistics.
To determine the extent to which the tuition increases varied by
school size, we obtained and analyzed annual school-by-school tuition
and enrollment data collected by the Department of Education through
its Integrated Postsecondary Education Data System (IPEDS) for school
years 1990-91 through 1995-96.

To determine the extent to which there is a statistical relationship
between increases in tuition and other factors, such as school
revenues and expenditures, we conducted linear regression analyses
using IPEDS data for school years 1989-90 and 1994-95.  We developed
two regression models:  one for public 4-year schools and one for
private 4-year schools.  We limited the scope of these analyses to
4-year public and private schools in the 50 states and the District
of Columbia where more than half of the students enrolled in the fall
term were undergraduates and where more than half of all students
enrolled in the fall term were full-time students.  To be included,
schools had to report tuition revenue, instruction expenses, fall
enrollment, and the typical tuition and required fee charge for
in-state full-time full-year undergraduates.  We excluded schools for
which their tuition charge was less than 50 percent or more than 150
percent of reported tuition revenue per estimated
full-time-equivalent (FTE) student.  Our analysis included a total of
383 public 4-year schools and 761 private 4-year schools.

Our dependent variable in each model was the dollar change in
in-state, full-time, full-year undergraduate tuition and required
fees that each school reported to the Department in its IPEDS survey.
We converted 1989-90 tuition figures to constant 1994-95 dollars
using the Department of Labor's Consumer Price Index (CPI) for all
urban consumers for that school year.  This index indicated an
18.4-percent increase in prices over the period compared with a
19.5-percent increase in the Higher Education Price Index (HEPI),
which tracks changes in the costs of items purchased by postsecondary
schools.

Most of our independent variables were the dollar amounts of change
in selected revenue and expense items per FTE student.  Here again,
we adjusted 1989-90 dollar figures for inflation using the CPI.  We
estimated the number of FTE students at each school by multiplying
the number of part-time fall-term students by one-third and adding
the number of full-time fall-term students.  We recognize that the
calculation of FTE students is a contentious subject and may differ
considerably by type of school.  However, we believe the calculation
method we selected is appropriate for the purposes of our analyses.

Our independent variables included two categories of revenue from
Part A of the IPEDS finance survey:  governmental appropriations and
revenue from grants, contracts, and other sources.  We excluded the
amount of (1) student scholarships, fellowships, and grants other
than those from institutional funding sources and (2) revenue from
hospital, independent, and auxiliary operations.

We placed expenditure items reported in Part B of the IPEDS finance
survey into the following categories:

  -- instruction expenditures;

  -- student services expenditures;

  -- other student-related expenditures (a pro rata share of other
     expenditures excluding research, public service, hospital,
     independent and auxiliary operations and amounts of student
     scholarships, fellowships, and grants other than those from
     institutional funding sources);

  -- research expenditures;

  -- scholarships, fellowships, and grants to students from
     institutional funding sources (versus federal, state, local, or
     private sources); and

  -- other nonstudent-related expenditures including public service,
     pro-rata portion of academic support, institutional support,
     operation and maintenance of the physical plant, and transfers
     to other funds for purposes such as debt service and additions
     to the physical facilities.  (We excluded hospital, independent
     and auxiliary operations and amounts of student scholarships,
     fellowships, and grants.)

Our analysis also included independent variables for the
inflation-adjusted dollar change per student in (1) the market value
of each school's endowment and (2) the amounts of tuition used for
purposes not reported in Part A of the IPEDS survey.  Our analysis
excluded measures of quality of education and other services offered
by schools.

As agreed with your office, the scope of this regression analysis
does not identify the causes of changes in tuition.  It is
particularly difficult to identify factors causing variation in
tuition increases because decisions to increase tuition and to
increase expenditures are likely to be interdependent.  In addition,
it is likely that decisions to increase tuition and increase
institutional (compared with federal) financial aid may be
interdependent.

The Department uses separate IPEDS survey forms to collect fall
enrollment data, finance data, and tuition data.  In many cases,
schools reported some but not all data for separate campuses or
affiliates.  As a result, in many cases schools used one
identification number for finance data and used several
identification numbers to report enrollment data.  In some cases,
schools reported data separately in one of the study years and in
aggregate in other years.  Where possible, we aggregated an
institution's data, calculating estimated FTE weighted average
tuition rates for the whole institution where tuition and enrollment
data were available for subunits.  Where this was not the case, we
used tuition for the school entity for which financial data were
reported.  In many cases, data for some of the variables were missing
for schools in either the 1989-90 or 1994-95 survey.  We used imputed
data from the Department where these were available.  In many cases,
data were missing and schools were excluded from our analysis.

To determine the extent that tuition increases at 4-year schools
affect the costs of attending community colleges, we conducted a
literature search and contacted representatives of major higher
education associations, such as the American Association of Community
Colleges, and individuals who have conducted research in the subject
area.

To identify examples of schools' efforts to reduce their own costs,
we conducted a literature search and contacted representatives of
example schools and major higher education associations, such as the
American Council on Education, and individuals who have conducted
research in higher education finance issues.  We also contacted
officials from the Department of Education and representatives of
accounting and consulting firms that advise schools on cost
containment strategies.

We conducted our work in accordance with generally accepted
government auditing standards between October 1997 and July 1998.  We
did not verify the accuracy of data in the Department's IPEDS data
sets.  The Department uses several procedures for ensuring that the
data are accurate, including a wide range of edit checks to ensure
that numbers reported in different places in the survey forms are
consistent and that numbers are within a range of tolerance compared
with the schools' report for the prior years.  Nonetheless, the
Department does not conduct tests to ensure that schools' reported
tuition receipts are consistent with the numbers of students enrolled
and the tuition charges the schools report.  In several cases, we
found that schools' reported in-state tuition charges were less than
50 percent or more than 150 percent of receipts per estimated FTE
student.  Concerned that the data reported by these schools may not
be consistent, we excluded them from our analysis.

VARIATIONS IN TUITION INCREASES BY
KIND AND SIZE OF SCHOOL
========================================================== Appendix II

                               Table II.1

                 Variations in Tuition Increases at 4-
                Year Public Schools, School Years 1990-
                     91 and 1995-96, by School Size

                                           Average
                                          tuition of
                                          schools in       Tuition
                                           quartile        increase
                                        --------------  --------------
Enrollment          Fall 1995            1990-   1995-          Percen
quartile            enrollment              91      96  Amount       t
------------------  ------------------  ------  ------  ------  ------
1                   14,157 to 51,445    $2,227  $2,945    $718     32%
2                   7,095 to 14,108      2,079   2,755     675      33
3                   3,259 to 7,061       1,963   2,615     651      33
4                   287 to 3,256         2,278   2,800     522      23
----------------------------------------------------------------------

                               Table II.2

                 Variations in Tuition Increases at 4-
                Year Private Schools, School Years 1990-
                     91 and 1995-96, by School Size

                                           Average
                                          tuition of
                                          schools in       Tuition
                                           quartile        increase
                                        --------------  --------------
Enrollment          Fall 1995            1990-   1995-          Percen
quartile            enrollment              91      96  Amount       t
------------------  ------------------  ------  ------  ------  ------
1                   2,500 to 35,835     $11,32  $13,31  $1,991     18%
                                             7       8
2                   1,352 to 2,492       9,707  11,398   1,691      17
3                   666 to 1,343         8,893  10,487   1,594      18
4                   3 to 664             6,925   8,218   1,293      19
----------------------------------------------------------------------

                               Table II.3

                 Variations in Tuition Increases at 2-
                Year Public Schools, School Years 1990-
                     91 and 1995-96, by School Size

                                           Average
                                          tuition of
                                          schools in       Tuition
                                           quartile        increase
                                        --------------  --------------
Enrollment          Fall 1995            1990-   1995-          Percen
quartile            enrollment              91      96  Amount       t
------------------  ------------------  ------  ------  ------  ------
1                   6,583 to 47,060     $1,300  $1,659    $359     28%
2                   3,268 to 6,562       1,401   1,745     345      25
3                   1,891 to 3,261       1,240   1,554     314      25
4                   124 to 1,890         1,197   1,449     252      21
----------------------------------------------------------------------

REGRESSION ANALYSES
========================================================= Appendix III

To better distinguish the characteristics of schools that experienced
larger increases in tuition from 1989-90 to 1994-95, we undertook a
series of regression analyses.  As shown in table III.1, we examined
15 variables in relation to changes in tuition.

                              Table III.1

                   Variables Examined in Relation to
                           Changes in Tuition

Variable                            Description
----------------------------------  ----------------------------------
Revenue
----------------------------------------------------------------------
Change in government                These included federal, state, and
appropriations                      local appropriations to schools,
                                    excluding appropriations for
                                    grants to students such as Pell
                                    grants. This variable is expressed
                                    as the change in the 1994-95
                                    constant dollar amount of these
                                    appropriations from 1989-90 to
                                    1994-95 (the amount for 1994-95
                                    less the amount for 1989-90).

Change in revenue from grants,      This is educational and general
contracts, and other sources        revenue excluding appropriations
                                    and tuition revenue. It includes
                                    grants, contracts, and gifts. It
                                    excludes funding received for
                                    noninstitutional student grants,
                                    scholarships, and fellowships and
                                    revenue for hospitals, auxiliary,
                                    and independent operations. This
                                    is the change in the 1994-95
                                    constant dollar amount of these
                                    revenues from 1989-90 to 1994-95.

Expenditures
----------------------------------------------------------------------
Change in instruction expenditures  This includes expenditures for
                                    both credit and noncredit
                                    instructional activities expressed
                                    as the change in the 1994-95
                                    constant dollar amount of these
                                    expenditures from 1989-90 to 1994-
                                    95. The Department's survey form
                                    for collecting this information
                                    instructs schools to include
                                    research expenditures not budgeted
                                    by the school.

Change in student services          This includes career guidance,
expenditures                        counseling, financial aid
                                    administration, and student health
                                    services. This is expressed as
                                    change in the 1994-95 constant
                                    dollar amount of these
                                    expenditures from 1989-90 to 1994-
                                    95.

Change in institutional             This is the change in the 1994-95
scholarships, fellowships, and      constant dollar amount of these
grants                              expenditures from 1989-90 to 1994-
                                    95.

Change in other student-related     This includes a prorated portion
expenditures                        of academic and institutional
                                    support, physical plant
                                    maintenance, and transfers to
                                    other funds. The variable is the
                                    change in the 1994-95 constant
                                    dollar amount of these
                                    expenditures from 1989-90 to 1994-
                                    95.

Change in research expenditures     The variable is the change in the
                                    1994-95 constant dollar amount of
                                    these expenditures from 1989-90 to
                                    1994-95. These figures include
                                    only research budgeted by the
                                    school. Other research
                                    expenditures are included with
                                    instruction expenditures.

Change in other nonstudent-         This includes public service and a
related expenditures                prorated portion of academic and
                                    institutional support, physical
                                    plant maintenance, and transfers
                                    to other funds. The variable is
                                    expressed as the change in the
                                    1994-95 constant dollar amount of
                                    these expenditures from 1989-90 to
                                    1994-95.

Other items
----------------------------------------------------------------------
Change in the ratio of in-state     This is expressed as the
undergraduate tuition to average    percentage point change in in-
tuition received per FTE student\a  state undergraduate tuition as a
                                    percentage of average tuition
                                    revenue received per estimated FTE
                                    student.

Change in the amount of tuition     This is the change in the 1994-95
used for noncurrent fund purposes   constant dollar amount of these
                                    expenditures or uses from 1989-90
                                    to 1994-95. These are uses not
                                    noted among current funds
                                    expenditures. These may include
                                    uses in renewal, replacement, or
                                    addition to the school's physical
                                    plant or retirement of debt.

Ratio of in-state undergraduate     This is expressed as in-state
tuition to average tuition          full-time, full-year undergraduate
received per FTE student\a          tuition as a percentage of the
                                    average amount of tuition and
                                    required fee revenue received per
                                    estimated FTE student in 1989-90.

Change in market value of the       This is the change in the 1994-95
endowment fund                      constant dollar amount of the fund
                                    balance from 1989-90 to 1994-95.

Level of in-state undergraduate     This is the in-state full-time,
tuition in 1989-90                  full-year undergraduate tuition
                                    rate in constant 1994-95 dollars.

Change in the amount by which       This variable is the change from
revenues exceeded or fell short of  1989-90 to 1994-95 in the 1994-95
expenditures (surplus or deficit    constant dollar amount by which
in current account)                 general and educational revenue
                                    exceeded or fell short of general
                                    and educational expenditures. This
                                    excluded funds for
                                    noninstitutional scholarships,
                                    fellowships, and grants to
                                    students, independent operations,
                                    auxiliary enterprises, and
                                    hospitals.

Change in surplus (deficit) from    This variable is the change over
independent operations, auxiliary   the 1989-90 to 1994-95 period in
enterprises, and hospitals          the 1994-95 constant dollar amount
                                    by which revenue from independent
                                    operations, auxiliary enterprises,
                                    and hospitals exceeded or fell
                                    short of expenditures for the
                                    same.
----------------------------------------------------------------------
\a Private school tuition is generally the same for in-state and
out-of-state students.  This variable may be affected by the mix of
in-state and out-of-state students; the mix of undergraduate,
graduate, and first professional students; and the mix of part-time
and full-time students.  While we recognize these mixes may differ
considerably among schools, we did not attempt to analyze them.

As shown in table III.2, we computed several statistics for 4-year
public and private schools indicating relationships between variables
and changes in tuition.  We indicate the portion of variation
accounted for with and without adjustments for other variables.

                              Table III.2

                Statistics Concerning Selected Variables
                 and Variation in Tuition Changes, From
                           1989-90 to 1994-95

Statistic (letter designates
column in tables III.3 and III.4)   Description
----------------------------------  ----------------------------------
A. Portion of variation accounted   The figures in this column
for after controlling for the       indicate the extent to which the
other factors                       portion of variation accounted for
                                    (multiple R squared, the multiple
                                    coefficient of determination)
                                    declines with the deletion of each
                                    variable one at a time. In table
                                    III.3, for example, all the
                                    variables together accounted for
                                    86 percent of the variation in
                                    tuition increases. When the
                                    analysis is repeated with all the
                                    variables listed except the first
                                    (changes in governmental
                                    appropriations), the portion
                                    accounted for declines by 64
                                    percentage points to 25 percent.

B. Portion of variation accounted   \These figures indicate the
for without including factors       cumulative portion of variation in
listed lower                        tuition accounted for with the
                                    inclusion of each successive
                                    variable moving down the list. For
                                    example, the first two variables-
                                    -change in governmental
                                    appropriations and change in
                                    instruction expenditures--
                                    accounted for 8.52 percent of
                                    variation in tuition increases.
                                    Adding a third variable to the
                                    analysis (change in
                                    nonappropriation revenue)
                                    increased the portion of variation
                                    explained to 10.97 percent.

C. Portion of variation accounted   These figures show an increase in
for by the addition of each         the portion of variation explained
variable to the model including     with the addition of each variable
factors listed in rows above        as you move down the table. Since
                                    adding the third variable, change
                                    in revenues from grants,
                                    contracts, and other sources, to
                                    the preceding two variables
                                    increased the portion of variation
                                    accounted for from 8.52 to 10.97
                                    percent, the change associated
                                    with the addition of this variable
                                    was 2.4 percent.

D. Portion of variation accounted   These figures show the portion of
for by each variable separately     variation in tuition accounted for
(without controlling for other      by each variable without
factors)                            controlling for any other
                                    variables. This is the correlation
                                    coefficient (column E) squared.

E. Correlation coefficient with     These are the Pearson correlation
change in tuition                   coefficients of each variable with
                                    the dependent variable--change in
                                    tuition from 1989-90 to 1994-95.
                                    If increases in a variable were
                                    consistently associated with
                                    increases in tuition, the
                                    correlation would approach 1.0. If
                                    decreases in the variable were
                                    consistently associated with
                                    increases in tuition, the
                                    correlation would approach -1.0.
                                    If there were no statistical
                                    relationship between the two
                                    variables (without controlling for
                                    other factors), the correlation
                                    coefficient would approach 0.

F. Regression coefficient           These are the regression
                                    coefficients, which indicate the
                                    model's estimate of how many
                                    dollars more tuition would
                                    increase with each dollar (or
                                    other unit of measure) change in
                                    each independent variable after
                                    controlling for all the other
                                    variables. For example, if
                                    instruction expenditures increased
                                    by one dollar more, the model
                                    predicts that tuition increases by
                                    about 76 cents more. Similarly, if
                                    governmental appropriations
                                    declined by one more dollar,
                                    tuition would increase by about 75
                                    cents more (after adjusting for
                                    other variables).

G. Probability of the T statistic   These figures are the probability
                                    of the T statistic. They indicate,
                                    for each variable in the model,
                                    the probability that the
                                    statistical relationship between
                                    the variable and the variation in
                                    average tuition increases not
                                    accounted for by all the other
                                    variables was due to random
                                    factors.
----------------------------------------------------------------------

   RESULTS FOR 4-YEAR PUBLIC
   COLLEGES AND UNIVERSITIES
------------------------------------------------------- Appendix III:1

Together the 15 variables we examined accounted for about 85 percent
of the variation in tuition increases among 4-year public colleges
and universities.  (Multiple R squared was 85.69 percent and adjusted
multiple R squared was 85.11 percent.) The first seven variables
shown in table III.3 were particularly important:\18

1.  change in government appropriations (federal, state, and local);

2.  change in instruction expenditures;

3.  change in revenue from grants, contracts, and other sources;

4.  change in the amount by which revenues exceed or fell short of
expenditures;

5.  change in other student-related expenditures;

6.  change in other nonstudent-related expenditures; and

7.  change in research expenditures.

Schools that experienced larger increases in these expenditure items,
larger decreases in these revenue items, or larger increases in
surpluses or decreases in deficits were more likely than other
schools to have larger increases in tuition.  Deleting any one of
these variables from the statistical analysis resulted in a decline
in the portion of variation accounted for of more than 30 percentage
points.  (In other words, after adjusting for all the other factors
in the model, each of these variables accounted for more than 30
percent of the variation in tuition increases.)

In contrast to other financial variables, such as the change in the
value of the school's endowments, changes in amounts of institutional
grants, scholarships, and fellowships, and amounts of tuition used
for noncurrent fund purposes accounted for little, if any, of the
variation in tuition increases.

                                       Table III.3

                            Factors Associated With Changes in
                          Tuition at Public 4-Year Schools, From
                                    1989-90 to 1995-96

                                                                           Regression
                                                                        equation with all
                                                                             listed
                  Portion of variation accounted for                       variables\a
              ------------------------------------------                -----------------
                A\b         B           C          D           E           F         G
              --------  ----------  ----------  --------  ------------  --------  -------
                                        by the
                                      addition   by each
                                       of each  variable
                                      variable  separate
                 after                  to the        ly
              controll     without      model,  (without
               ing for   including   including  controll   Correlation  Regressi
                   the     factors     factors   ing for   coefficient        on
                 other      listed   listed in     other   with change  coeffici  Probabi
Variable       factors     lower\c  rows above  factors)    in tuition       ent     lity
------------  --------  ----------  ----------  --------  ------------  --------  -------
Change in       64.01%       2.07%       2.07%     2.07%        -0.144     -0.75        <
 government                                                                        0.0001
 appropriati
 ons \d,e
Change in        52.60        8.52        6.45      1.57         0.125      0.76        <
 instruction                                                                       0.0001
 expenditure
 s\e
Change in        52.50       10.97        2.44      0.15        -0.039     -0.79        <
 revenue                                                                           0.0001
 from
 grants,
 contracts,
 and other
 sources\f
Change in        46.52       12.53        1.56      0.11        -0.034      0.77        <
 the amount                                                                        0.0001
 by which
 revenues
 exceeded or
 fell short
 of
 expenditure
 s
Change in        41.17       18.90        7.93      0.49         0.070      0.76        <
 other                                                                             0.0001
 student-
 related
 expenditure
 s
Change in        31.79       32.67       13.77      0.43         0.065      0.76        <
 other                                                                             0.0001
 nonstudent-
 related
 expenditure
 s\f
Change in        30.53       46.70       14.03      0.24         0.049      0.80        <
 research                                                                          0.0001
 expenditure
 s\f
Change in        26.59       66.06       19.36      4.02         0.200     23.34        <
 the ratio                                                                         0.0001
 of in-
 state
 undergradua
 te tuition
 to average
 tuition
 received
 per FTE
 student
Change in        12.06       75.44        9.38      3.15         0.178      0.78        <
 student                                                                           0.0001
 services
 expenditure
 s
Change in         5.84       80.26        4.82      1.77        -0.133      0.67        <
 amount of                                                                         0.0001
 tuition
 used for
 noncurrent
 fund
 purposes
Ratio of in-      4.73       85.61        5.35      1.52        -0.123      6.66        <
 state                                                                             0.0001
 undergradua
 te tuition
 to average
 tuition
 received
 per FTE
 student
Change in         0.08       85.69        0.08      0.83         0.091     -0.03   0.1484
 surplus or
 deficit
 from
 independent
 operations,
 auxiliary
 enterprises
 , and
 hospitals
Level of in-      0.00       85.69        0.00      2.28         0.151      0.00   0.7526
 state
 undergradua
 te tuition
 in 1989-90
Change in         0.00       85.69        0.00      6.16         0.248      0.00   0.9267
 institution
 al
 scholarship
 s,
 fellowships
 , and
 grants
Change in         0.00       85.69        0.00      0.10         0.031      0.00   0.9984
 market
 value of
 the
 endowment
 fund
-----------------------------------------------------------------------------------------
Note:  The dependent variable in each regression analysis was the
change in full-time, in-state, undergraduate tuition from school year
1989-90 to 1994-95.  This analysis does not identify the causes of
changes in tuition.  It is particularly difficult to identify factors
causing variation in tuition increases because decisions to increase
tuition, increase expenditures, and increase institutional financial
aid are likely to be interdependent.  Further analysis using
techniques that could shed light on the nature of these
interdependencies was beyond the scope of our work.

\a The constant in the regression equation is -515.

\b See table III.2 for an explanation of the statistics in each
column.

\c Analysis of changes in R squared by adding variables one at a time
in reverse order (starting with change in institutional scholarships,
fellowships, and grants and working up one variable at a time)
further demonstrated the importance of changes in government
appropriations.  Only two other variables added more than 5
percentage points to R squared:  adding institutional scholarships,
fellowships, and grants to the analysis increased R squared by 6
percentage points, and adding change in government appropriations
increased R squared by 64 percentage points.

\d This and each of the other financial variables were expressed as a
dollar change per estimated fall FTE student in constant 1994-95
dollars, except as noted.

\e The utility of assessing the importance of a variable by deleting
it from the model (as shown in column A) or adding it to a model
(columns B and C) is limited where independent variables are
correlated with one another.  Change in government appropriations and
change in instruction expenditures were correlated (r = 0.5658).
Removing both these variables from the model decreased R squared by
66 percentage points to 20 percent.

\f Change in revenue from grants, contracts, and other sources was
correlated with change in research expenditures (r = 0.7594) and
change in research expenditures was correlated with change in other
nonstudent-related expenditures (r = 0.2962).  The correlation
between revenue from grants, contracts, and other sources and change
in other nonstudent-related expenditures was similar (r = 0.3791).
Removing these three variables from the model decreased R squared by
54 percentage points to 32 percent.

--------------------
\18 We identified variables as "important" here if the decline in the
portion of variation accounted for (R squared) by omitting the
variable from the model was greater than the decline observed from
omitting other variables, one at a time.  This technique is affected
by strong correlations between the variables.  An omitted variable
may appear to be less important if it is strongly correlated with
another independent variable.

   RESULTS FOR 4-YEAR PRIVATE
   COLLEGES AND UNIVERSITIES
------------------------------------------------------- Appendix III:2

Our analysis accounted for less of the variation in tuition increases
at private 4-year colleges and universities than it did at 4-year
public institutions.  The 15 variables we examined accounted for
about 70 percent of the variation in tuition increases from 1989-90
to 1994-95.  The R squared was 0.7199 and the adjusted R squared was
0.7142.  Most of the seven most highly related variables accounting
for variation among 4-year public institutions also figured
prominently in our 4-year private college analysis.  In addition,
changes in the ratio of in-state undergraduate tuition to average
tuition received per FTE student were helpful in accounting for
variation in tuition changes.  These highly associated variables
were\19

1.  change in revenue from grants, contracts, and other sources;

2.  change in the amount by which revenues exceeded or fell short of
expenditures;

3.  change in other student-related expenditures;

4.  changes in the ratio of undergraduate tuition to average tuition
received per FTE student;

5.  change in instruction expenditures;

6.  change in research expenditures; and

7.  change in other nonstudent-related expenditures.

Other variables helped account for less than 10 percent of additional
variation in changes in tuition after controlling for all the other
variables.

                                   Table III.4

                        Factors Associated With Changes in
                     Tuition at Private 4-Year Schools, From
                                1989-90 to 1995-96

                                                                    Regression
                                                                  equation with
                         Portion of variation accounted             all listed
                                      for                          variables\a
                        --------------------------------          --------------
                         A\b       B        C       D       E       F       G
                        ------  --------  ------  ------  ------  ------  ------
                                          by the
                                          additi
                                           on of      by
                                            each    each
                                          variab  variab
                                           le to      le
                                             the  separa
                                          model,    tely  Correl
                         after            includ  (witho   ation
                        contro               ing      ut  coeffi
                         lling   without  factor  contro   cient
                           for  includin       s   lling    with
                           the         g  listed     for  change  Regres
                         other   factors      in   other      in    sion
                        factor    listed    rows  factor  tuitio  coeffi  Probab
Variable                     s   lower\c   above      s)       n   cient   ility
----------------------  ------  --------  ------  ------  ------  ------  ------
Change in revenue from  27.81%     0.27%   0.27%   0.27%   0.052   -0.62       <
 grants, contracts,                                                       0.0001
 and other sources
 \d,e
Change in the amount     27.02      1.07    0.80    0.61       -    0.61       <
 by which revenues                                         0.078          0.0001
 exceeded or fell
 short of expenditures
Change in other          26.12      1.84    0.77    1.84   0.136    0.61       <
 student-related                                                          0.0001
 expenditures
Change in instruction    16.91      4.14    2.30    3.03   0.174    0.55       <
 expenditures                                                             0.0001
Change in the ratio of   13.51     12.76    8.62    2.22       -   42.83       <
 undergraduate tuition                                     0.234          0.0001
 to average tuition
 received per FTE
 student
Change in research       12.02     15.89    3.13    0.28   0.053    0.65       <
 expenditures\e                                                           0.0001
Change in nonstudent-     9.79     52.92   37.03   13.33   0.365    0.59       <
 related expenditures                                                     0.0001
 \e
Change in government      7.22     57.55    4.63    0.00       -   -0.78       <
 appropriations                                            0.005          0.0001
Change in student         6.96     64.95    7.40    1.50   0.123    0.65       <
 services expenditures                                                    0.0001
Ratio of undergraduate    2.28     67.81    2.86    2.22       -   15.28       <
 tuition to average                                        0.149          0.0001
 tuition received per
 FTE student
Level of undergraduate    1.01     69.75    1.94   27.90   0.528    0.04       <
 tuition in 1989-90\f                                                     0.0001
Change in                 0.63     70.39    0.64   25.48   0.505    0.17       <
 institutional                                                            0.0001
 scholarships,
 fellowships, and
 grants\f
Change in amount of       0.33     70.72    0.33    0.10       -    0.15  0.0038
 tuition used for                                          0.031
 noncurrent fund
 purposes
Change in surplus or      0.04     70.76    0.04    0.24   0.049   -0.02  0.2897
 deficit from
 independent
 operations, auxilary
 enterprises, and
 hospitals
Change in market value    0.00     70.77   0.004    0.29   0.054    0.00  0.7561
 of the endowment fund
--------------------------------------------------------------------------------
Note:  The dependent variable was the change in full-time, full-year
in-state, undergraduate tuition from school year 1989-90 to 1994-95.
This analysis does not identify the causes of changes in tuition.  It
is particularly difficult to identify factors causing variation in
tuition increases because decisions to increase tuition, increase
expenditures, and increase institutional financial aid are likely to
be interdependent.  Further analysis using techniques that could shed
light on the nature of these interdependencies was beyond the scope
of our work.

\a The constant in the regression equation is -1,230.

\b See table III.2 for an explanation of the statistics in each
column.

\c Analysis of changes in R squared by adding variables one at a time
in reverse order (starting with change in the market value of the
endowment fund and working up one variable at a time) indicated
greater importance for changes in institutional scholarships,
fellowships, and grants.  This variable added 25 percentage points to
R squared in an analysis including the three variables listed below
it in the table.  Only two other variables increased R squared by
more than 5 percentage points.  These were the level of undergraduate
tuition in 1989-90 and the change in revenue from grants, contracts,
and other sources (adding 9 and 28 percentage points to R squared,
respectively).

\d This and each of the other variables were expressed as changes in
constant 1994-95 dollars per estimated FTE student between school
years 1989-90 and 1994-95, the amount for 1994-95 less the amount for
1989-90, except as noted.

\e The utility of assessing the importance of a variable by deleting
it from the model (as shown in column A) or adding it to a model
(columns B and C) is limited where independent variables are
correlated with one another.  Change in revenue from grants,
contracts, and other sources was correlated with change in research
expenditures (r = 0.5455) and change in research expenditures was
correlated with change in other nonstudent-related expenditures (r =
0.3858).  The correlation between change in revenue from grants,
contracts, and other sources and change in other nonstudent-related
expenditures was in the same range (r = 0.4146).  Removing these
three variables from the model decreased R squared by 36 percentage
points to 43 percent.

\f Changes in other nonstudent-related expenditures were correlated
with changes in amounts of institutional scholarships, fellowships,
and grants (r = 0.6242).  The level of undergraduate tuition in
1989-90 was also correlated with changes in other nonstudent-related
expenditures (r = 0.4126).  Change in institutional scholarships,
fellowships, and grants and the level of undergraduate tuition in
1989-90 had correlation in the same range (r = 0.5453).  Removing
these three variables from the model decreased R squared by 49
percentage points to 23 percent.

--------------------
\19 We identified variables as "important" here if the decline in the
portion of variation accounted for (R squared) by omitting the
variable from the model was greater than the decline observed from
omitting other variables, one at a time.  This technique is affected
by strong correlations between the variables.  An omitted variable
may appear to be less important if it is strongly correlated with
another independent variable.

EXAMPLES OF SCHOOL COST
CONTAINMENT EFFORTS
========================================================== Appendix IV

   STATEWIDE INITIATIVES
-------------------------------------------------------- Appendix IV:1

      OREGON PUBLIC POSTSECONDARY
      SCHOOLS
------------------------------------------------------ Appendix IV:1.1

In 1989, Oregon voters passed a ballot measure that reduced property
taxes and shifted hundreds of millions of state tax dollars from
state agencies, including higher education, to local school districts
and prisons.  To make up for the lost state tax revenues, the state's
public postsecondary colleges and universities reduced the number of
programs, closed a college, cut services, and laid off hundreds of
nontenured faculty and staff.  The schools were also required to
develop scenarios for administrative cuts of up to 30 percent.

The University of Oregon, for example, implemented various measures
to cut 15 percent, or $7.7 million, in the first round of cuts from
its administrative budget.  Subsequent savings efforts included the
implementation of a new on-line financial information system.  These
actions have eliminated nearly all the paperwork and manual
processing that was necessary before.  Previously, it cost about $25
to manually issue a check, and now the cost is about $6.  At slightly
over 100,000 transactions per year, the estimated annual cost savings
are between $150,000 and $190,000.  In addition, the university
expanded its procurement card program wherein departments are allowed
to make purchases with a credit card instead of purchase orders.
Through the combined expansion of procurement cards and reduced cost
or elimination of checks for purchases made on the cards, annual
savings are estimated to be about $18,000.  This new system
reportedly not only saves money and time but is also more convenient;
products are delivered faster, and the schools' vendor base is larger
because some businesses that do not accept purchase orders accept
cards.

      OHIO PUBLIC POSTSECONDARY
      SCHOOLS
------------------------------------------------------ Appendix IV:1.2

In 1992, the Ohio Board of Regents initiated an effort to contain
costs at public postsecondary schools.  From fiscal year 1993 through
1997, the average annual savings was $55 million, totaling $270
million during the 5-year period.  According to the Board of Regents'
1997 annual progress report on cost containment,

  -- nearly all campuses reported significant savings through the
     adoption of cost- and time-saving changes in the purchasing
     procedures, including the adoption of electronic ordering
     systems and participation in cooperative purchasing agreements
     with other institutions;

  -- several campuses are using technology to eliminate or reduce the
     paperwork for processing financial aid;

  -- nearly half of the campuses have reported reducing or containing
     the cost of providing health care benefits to their employees;

  -- many campuses have achieved significant savings through
     administrative restructuring and downsizing and by outsourcing
     many services previously done in-house; and

  -- many creative projects have been undertaken to significantly
     reduce the cost of providing energy for the lighting, heating,
     and cooling of campus facilities.

      UNIVERSITY OF WISCONSIN
      SYSTEM
------------------------------------------------------ Appendix IV:1.3

The University of Wisconsin System and its 13 4-year schools and 13
2-year schools have taken a wide variety of efficiency-related
measures.  In addition to eliminating or consolidating a number of
administrative functions and academic programs, the system and its
schools have taken many other cost reduction or efficiency
improvement actions.  These include

  -- the implementation of injury prevention programs, safety
     awareness training, and better claims management, resulting in
     savings of $1.1 million;

  -- the establishment of preferred rate agreements with hotels and
     hotel chains, generating approximately $180,000 in annual
     savings; and

  -- the installation of more efficient and effective lighting at the
     Madison campus, bringing an annual savings of $475,000.

   MULTIPLE SCHOOL COLLABORATIVE
   EFFORTS
-------------------------------------------------------- Appendix IV:2

      FIVE COLLEGES, INC.
------------------------------------------------------ Appendix IV:2.1

Five Colleges, Inc., a consortium of five Massachusetts schools
(Amherst College, Hampshire College, Mount Holyoke College, Smith
College, and the University of Massachusetts at Amherst), was formed
to share resources, services, and programs.  The schools pooled
faculty to create two consortium departments (astronomy and dance),
to confer minor degrees and certificates, and to make joint faculty
appointments.  The consortium also offers open club membership, open
theater productions, open library access, meal exchange, and
cross-registration, all at no additional cost to students.  The
schools also have common contracts for key services and a joint
purchasing agreement that saves more than $1 million a year.  They
also share a risk-management expert, saving $300,000 a year, and a
recycling coordinator, saving another $35,000 a year.

      MIDWESTERN HIGHER EDUCATION
      COMMISSION (MHEC)
------------------------------------------------------ Appendix IV:2.2

MHEC, a regional consortium established in 1991, is made up of public
and private nonprofit colleges and universities in Illinois, Indiana,
Kansas, Michigan, Minnesota, Missouri, Nebraska, Ohio, and Wisconsin.
The consortium has multiple programs under way to help MHEC member
schools, such as

  -- Academic scheduling and management software, which was created
     in response to the needs of commission schools and offered to
     them at discounts saving the schools $800,000 in the lower
     purchase price.

  -- Virtual Private Network, which provides low-cost voice, video,
     and data transmission services at some of the lowest
     telecommunication rates in the country.  This effort was
     especially helpful to smaller institutions that cannot leverage
     lower telecommunication rates on their own.

  -- Interactive Video Program, which provides over 100 members with
     discounted prices on interactive video equipment at a total
     savings of more than $7 million.

   INDIVIDUAL SCHOOL EFFORTS
-------------------------------------------------------- Appendix IV:3

      EMPORIA STATE UNIVERSITY
------------------------------------------------------ Appendix IV:3.1

Emporia State University in Emporia, Kansas, has achieved cost
savings and other efficiencies from actions taken, such as the
following.

  -- Implemented an energy conservation program by installing
     building automation equipment.  The retrofit networked about 70
     percent of the campus with new controls and will produce an
     anticipated savings of about $571,500 per year.

  -- Purchased furniture that was constructed by Kansas Correctional
     Industries for the school's lounges, residence halls, most of
     the school's residence hall lounges, married student apartments,
     and several departments.  Estimated saving was over $112,000
     compared with furniture from the private sector.  The school
     also purchased paint and custodial supplies from the
     Correctional Industries for about half the price of buying them
     in the public sector.

  -- Negotiated a new lower-cost contract for deregulated natural
     gas, saving about $150,000 a year (depending upon winter weather
     conditions).

  -- Purchased 11,000 gallons of No.  2 fuel oil from federal surplus
     property for 10 cents a gallon versus the going price of 80
     cents, saving about 88 percent, or $7,700.  The school also
     purchased six new rooftop air-conditioning units from federal
     surplus property for $6,000 compared with the retail cost of
     $36,000.

  -- Negotiated with the telephone company for new phone service to
     the residence halls at a savings of $650,000, plus an additional
     annual $110,000 saved from 1985 to 1991 by eliminating the
     positions of switchboard operators used with the old telephone
     system.

  -- Saved $190,000 a year by training and qualifying employees as
     commercially licensed pest control applicators, thus eliminating
     the need to hire outside exterminators.

      SAN DIEGO STATE UNIVERSITY
------------------------------------------------------ Appendix IV:3.2

San Diego State University in San Diego, California, has achieved
savings through various measures.  For example,

  -- the school's Department of Electrical and Computer Engineering
     obtained donated software packages valued at over $12 million,

  -- the school negotiated its local telephone contract at an
     estimated annual savings of $33,000 compared with its previous
     services and

  -- the school hired an auditing firm to audit the phone bills
     received from vendors, resulting in $51,600 in savings from
     erroneous charges and $34,300 in refunds.

      TENNESSEE TECHNOLOGICAL
      UNIVERSITY
------------------------------------------------------ Appendix IV:3.3

Tennessee Technological University in Cookeville, Tennessee, issued a
nonexclusive office supply contract with a local vendor, which saved
about $320,000 in 6 months.  The vendor will also make an annual
contribution to the university's scholarship fund, based on a
percentage of sales during the course of the contract.  In addition,
by procuring carpets through promotional sales or from a carpet
mill's special goods section, the school saved as much as 75 percent
of the retail cost.

      TEXAS WOMAN'S UNIVERSITY
------------------------------------------------------ Appendix IV:3.4

Texas Woman's University (TWU), with campuses in Dallas, Denton, and
Houston, Texas, implemented a reorganization in 1989, reducing the
number of deans from 11 to 8 and the number of academic departments
from 28 to 21.  These changes also achieved modest savings in
administrative costs and promoted cooperation, coordination, and
interdisciplinary collaboration.  In the ensuing 5 years, TWU also
phased out or eliminated approximately one-third of its degree
programs.

      THOMAS EDISON STATE COLLEGE
------------------------------------------------------ Appendix IV:3.5

Thomas Edison State College in Trenton, New Jersey, reviews its
collegewide operations periodically to determine opportunities for
finding better and less costly ways to conduct its operations.
Examples of cost-savings efforts it has undertaken include

  -- filling only staff positions determined to be the most critical
     in terms of efficiency or revenue enhancement, which saved the
     school nearly $725,000;

  -- using part-time temporary hourly staff (rather than full-time
     employees) to supplement full-time staff during periods of peak
     need and operation, saving the college about $240,000;

  -- restructuring and consolidating several departments, which
     enabled the school to achieve approximately $128,000 in
     personnel savings and additional savings of $108,000 by closing
     its Newark office facility; and

  -- restructuring its reprographic and data processing contracts,
     saving the school $135,000, and investing in automated postal
     zip coding equipment, saving it $18,000 in postal costs.

      WASHINGTON STATE UNIVERSITY
------------------------------------------------------ Appendix IV:3.6

According to documentation provided by a school budget office
official, Washington State University in Pullman, Washington, through
a wide variety of cost-savings measures, reduced its costs per
student from $11,273 in school year 1986-87 to $10,998 in 1995-96 and
its spending per degree granted from $48,294 in 1986-87 to $46,066 in
1995-96 (all figures adjusted for inflation).  Some of the measures
the school has taken to achieve these savings include

  -- reallocating funding from areas of declining enrollment to areas
     of expanding enrollment;

  -- increasing faculty productivity (over the past 4 years, FTE
     students have increased 12 percent while FTE faculty increased
     only 5.99 percent);

  -- operating a highly efficient composting program that diverts
     cafeteria food scraps, heating plant coal ash, and agricultural
     school animal waste and bedding materials to beneficial use at
     an annual savings of $200,000 in avoided disposal fees; and

  -- acquiring, at no cost, used motor vehicles, including tractors,
     backhoes, farm trucks, and pickups as well as electronic repair
     parts and computer components from the U.S.  Department of
     Agriculture and the Boeing Company as federal excess property.

(See figure in printed edition.)Appendix V
COMMENTS FROM THE DEPARTMENT OF
EDUCATION
========================================================== Appendix IV

(See figure in printed edition.)

(See figure in printed edition.)

*** End of document. ***