Higher Education: Students Have Increased Borrowing and Working to Help
Pay Higher Tuitions (Letter Report, 02/18/98, GAO/HEHS-98-63).

To cope with rising tuition costs, more students are borrowing money and
more are working. At four-year public schools, graduating seniors who
had borrowed at some point during their college careers rose to 60
percent, up from 42 percent in 1992-93. During the same period, the
percentage of graduating seniors (recipients of bachelor's degrees) who
owed $20,000 or more rose from nine percent to 19 percent. At the
graduate and professional levels, the picture is mixed, but in general
the percentage of borrowers and the level of debt increased. Higher
borrowing levels were especially pronounced at professional schools,
where average debt among borrowers completing their programs soared from
$45,000 in 1992-93 to nearly $60,000 in 1995-96. More than two-thirds of
full-time undergraduate students held jobs during 1995-96 and worked an
average of 23 hours  a week.

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

 REPORTNUM:  HEHS-98-63
     TITLE:  Higher Education: Students Have Increased Borrowing and 
             Working to Help Pay Higher Tuitions
      DATE:  02/18/98
   SUBJECT:  Colleges/universities
             Higher education
             College students
             Education or training costs
             Student loans
             Part-time employment
IDENTIFIER:  National Postsecondary Student Aid Study
             William D. Ford Federal Direct Loan Program
             Federal Family Education Loan Program
             Dept. of Education Parent Loans for Undergraduate Students 
             Program
             
******************************************************************
** 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.                         **
******************************************************************


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


Report to the Honorable
Carol Moseley-Braun, U.S.  Senate

February 1998

HIGHER EDUCATION - STUDENTS HAVE
INCREASED BORROWING AND WORKING TO
HELP PAY HIGHER TUITIONS

GAO/HEHS-98-63

College Student Borrowing and Working

(104879)


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

  FDLP - William D.  Ford Federal Direct Loan Program
  FFELP - Federal Family Education Loan Program
  NPSAS - National Postsecondary Student Aid Study
  PLUS - Parent Loans for Undergraduate Students

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


B-278947

February 18, 1998

The Honorable Carol Moseley-Braun
United States Senate

Dear Senator Moseley-Braun: 

In August 1996 we reported that tuition increases at 4-year public
colleges had outstripped increases in household income,\1 and parents
and others were questioning how students and their families would
find the money to pay for school.  We reported, for example, that
tuition increased 234 percent from school year 1980-81 through
1994-95, in contrast to median household income, which rose 82
percent during this period.  Recent sharp increases in the annual
volume of new federal student loans suggest that the cumulative
amounts of student debt may be rising as well.  In response to this
concern, you asked us to answer the following questions: 

  -- What changes have occurred in recent years in the percentage of
     undergraduate and graduate/professional students who borrow and
     in the cumulative amount of their borrowing?\2

  -- What changes have occurred in the percentage of undergraduate
     and graduate/professional students who work and the number of
     hours they work? 

  -- How do undergraduate borrowing and work patterns differ by type
     of school, year in school, dependency status, family income, and
     race/ethnicity? 

  -- What information is available concerning the amounts of
     education debt parents incur? 

Our review was based in large part on our analysis of data collected
by the Department of Education as part of the National Postsecondary
Student Aid Study (NPSAS).\3 We focused our analysis primarily on two
academic or student years:  1992-93 and 1995-96.\4 We focused on
1995-96 because it was the most recent year for which data were
available.  We chose 1992-93 for a comparison because it (1) predated
the most recent round of increases in maximum limits on the amounts
of loans available through the government's major student loan
programs, and (2) like 1995-96, was a year in which the Department
identified students who had graduated from their academic programs. 
The use of these survey data allowed us to estimate averages for the
student population as a whole, as well as narrow the comparison to
groups that had completed their educational programs.  Appendix I
describes our methodology in further detail.  Appendix II provides
additional data, including confidence intervals, for the estimates
presented. 


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

\2 We identify "professional" students as those seeking professional
degrees at, for example, law schools or medical schools, excluding
students who have already received their first professional degree. 

\3 NPSAS is a Department of Education survey done periodically of a
nationally representative sample of all students in postsecondary
education institutions, and it includes questions about how students
and their families pay for this education. 

\4 The Department's NPSAS for 1992-93 covered students enrolled at
any time during school terms that began between May 1, 1992, and
April 30, 1993.  Similarly, the 1995-96 NPSAS survey covered students
enrolled during school terms that began between May 1, 1995, and
April 30, 1996. 


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

Over the past several years, students have turned increasingly to
borrowing to cope with rising education costs.  At the undergraduate
level, the percentage of postsecondary students who had borrowed by
the time they completed their programs (received a bachelor's degree,
associate degree, or award or certificate) increased from 41 percent
in 1992-93 to 52 percent in 1995-96, and the average amount of debt
per student increased from about $7,800 to about $9,700 in constant
1995-96 dollars.\5 For graduating seniors (recipients of bachelor's
degrees) and who had borrowed, the average rose from about $10,100 to
about $13,300.  And the portion of these graduates with $20,000 or
more of student debt grew from 9 percent to 19 percent during the
period. 

Students attending 4-year public institutions showed the largest
increase in the number of borrowers.  Sixty percent of seniors
graduating from these schools in 1995-96 borrowed at some point in
their program, up from 42 percent in 1992-93 and about even with the
percentage of borrowers at private 4-year schools.  Students at
2-year public institutions borrowed least often and in lesser
amounts.  At the graduate and professional levels, the picture is
somewhat mixed, but in general the percentage of borrowers and the
level of debt increased.  Higher borrowing levels were especially
pronounced at professional schools, where average debt among
borrowers completing their programs climbed from about $45,000 in
1992-93 to nearly $60,000 in 1995-96. 

More full-time undergraduates worked while attending school in
1995-96 than in 1992-93.  More than two-thirds of full-time
undergraduate students held jobs during 1995-96, working an average
of 23 hours a week while enrolled.  At graduate and professional
schools, the percentage of full-time students who worked changed
little over the same period.  About two-thirds of master's and
doctoral students worked, usually in part-time jobs directly related
to their field of study.  At professional schools, less than half
worked while enrolled. 

Some variations in borrowing and work patterns can also be seen on
the basis of the cost of attendance, dependency status,\6 family
income, and gender.  However, most characteristics are not very
strong predictors of how much undergraduates were likely to borrow or
work. 

Little information is available about amounts of debt parents
accumulate in order to pay for their children's postsecondary
education.  The information we reviewed--which is also used by others
within the education community--was obtained from NPSAS, the
University of Michigan, and the Federal Reserve Board.  In general,
household debt for education (primary, secondary, or postsecondary)
remains a small share of household debt. 


--------------------
\5 These are estimates of cumulative borrowing (not just federal
student loans) by undergraduate students who completed programs
(including associate degree, award, and certificate programs) during
1995-96.  Unless otherwise noted, all dollar figures shown in this
report are expressed in constant 1995-96 dollars. 

\6 The Higher Education Act of 1965, as amended, identifies students
as independent if any of the following apply:  The student (1) is 24
years of age or older, (2) is married, (3) is a graduate or
professional student, (4) has legal dependents other than a spouse,
(5) is an orphan or ward of the court, (6) is a veteran, or (7) is
determined and documented by his or her financial aid administrator
to be independent because of other unusual circumstances.  If none of
these apply, the student is classified as dependent. 


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

The 1985 through 1995 period saw an increase in both the number of
college students and the proportion of the college-aged population in
colleges, universities, training schools, and other postsecondary
institutions.  In 1995, more than 34 percent of all 18- to
24-year-old U.S.  residents were attending postsecondary schools,
compared with slightly less than 28 percent in 1985.  Many who attend
also plan to stay longer:  Two-thirds of college freshmen now intend
to go beyond a baccalaureate degree, compared with about half in
1980.  In part, this interest in postsecondary education likely
reflects students' recognition that such education is associated with
higher incomes later in life.  Bureau of the Census statistics
indicate that, on average, households headed by persons with
bachelor's degrees have average incomes nearly 70 percent higher than
households headed by persons with no more than high school diplomas. 
Households with a member that has a professional degree have incomes
that average about three times those of households in which members'
highest certificate is a high school diploma. 

As the number of students has increased, so has the size of the
government's student loan programs.  By the end of fiscal year 1996,
the estimated outstanding amount of loans provided by the Department
of Education's two largest loan programs,\7 the principal sources of
loans for postsecondary education, had reached $112 billion, up from
$91 billion a year earlier and from $65 billion in 1990, in constant
1995-96 dollars.  The Higher Education Amendments of 1992 increased
the maximum amount that students could borrow.  For example, the
limit for graduate and professional students rose from $74,750 to
$138,500 (in current dollars, including both graduate and
undergraduate loans). 

Borrowing and working both play significant roles in how students pay
for their education.  Figure 1 shows how an "average" full-time
student met the cost of his or her education at various types of
postsecondary institutions in school year 1995-96.\8 Together,
borrowing and working constituted more than half of the amount of
funds students needed to pay their cost of attendance at all types of
schools, except private 4-year schools.\9

   Figure 1:  Role of Borrowing
   and Working in Paying Cost of
   Attendance for an "Average"
   Full-Time Undergraduate for a
   School Year, by Type of School,
   1995-96

   (See figure in printed
   edition.)

Note:  The height of each stacked bar represents the estimated
average budgeted cost of attending each type of school full time for
a full academic year.  Costs vary by school and by students'
dependency and residence status (whether they live with their
parents, for example).  Although many students at proprietary schools
(for-profit, vocational schools) and public and private
less-than-4-year schools were enrolled in programs less than a school
year in length, for comparison, we show the estimated cost as if all
students were attending a proprietary school for a full school year
(8 to 9 months).  The estimates are based on data from NPSAS. 

While this "average" view is instructive as a way to see the general
role of student borrowing and working patterns, it does not show the
wide range of methods students use to finance college.  Some students
do not borrow or work at all, while others earn more than enough to
cover the cost of attending college.  To provide a more complete
picture, our report focuses on those students who borrow and those
who work, showing the annual and cumulative amounts of their
borrowing and the number of hours worked per week while they were
enrolled. 


--------------------
\7 These are the William D.  Ford Federal Direct Loan Program (FDLP)
and the Federal Family Education Loan Program (FFELP).  Under FDLP,
students obtain loans from schools, which use federal funds to make
them.  Under FFELP, students obtain loans from private lenders, such
as banks, with the Department of Education ensuring that lenders will
receive payment if students default. 

\8 We estimated average costs for all undergraduate students, whether
or not they borrowed or worked while enrolled. 

\9 "Cost of attendance" includes tuition, required fees, room, board,
transportation, books, and supplies. 


   MORE STUDENTS ARE BORROWING IN
   GENERALLY LARGER AMOUNTS
------------------------------------------------------------ Letter :3

The proportion of students who borrowed to finance the cost of
postsecondary education increased between school years 1992-93 and
1995-96, and the amounts they borrowed increased, after taking
inflation into account.  In general, this was true for both
undergraduates and graduate and professional students. 


      BORROWING TRENDS FOR
      UNDERGRADUATES
---------------------------------------------------------- Letter :3.1

An increasing percentage of undergraduates in all types of programs
turned to borrowing to finance part of their education.  To provide
as complete a picture as possible of how students used borrowing
during their entire period of enrollment, we focused our analysis on
undergraduates who had completed their 2-year, 4-year, or other
programs in 1992-93 and 1995-96.\10 In 1992-93, 41 percent of
undergraduates who completed their programs had borrowed in 1 or more
years.  By 1995-96, this number had risen to 52 percent.  The
percentage varied, however, by type of degree or certificate, with
the greatest increase in the group receiving bachelor's degrees (see
table 1). 



                          Table 1
          
          A Greater Portion of Undergraduates Are
          Borrowing in All Types of Programs, With
            Students Seeking Bachelor's Degrees
          Borrowing More, School Years 1992-93 and
                          1995-96

                      Percentage of
                        recipients        Average total
                    borrowing in 1 or       principal
                        more years          borrowed\a
                    ------------------  ------------------
Degree/award
received             1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Certificate or            39        53    $5,171    $5,597
 award
Associate degree          34        42     5,580     5,059
Bachelor's degree         46        60    10,080    13,269
----------------------------------------------------------
\a Dollars are adjusted for inflation to 1995-96 values. 

The average amount borrowed by undergraduates completing their
program (excluding those who had not borrowed) rose from about $7,800
to about $9,700 over the 1992-93 to 1995-96 period, after adjusting
for inflation.  The amounts borrowed by those receiving bachelor's
degrees in 1995-96 were the highest.  Among bachelor's degree
recipients, the portion of students who had borrowed $20,000 or more
for the 1992-93 through 1995-96 time period rose from about 9 percent
to about 19 percent of graduating seniors who had borrowed; see table
2.  (See tables II.2, II.3, and II.4 for supporting data, including
confidence intervals (degree of precision) for the estimates.)



                                Table 2
                
                Many More Graduating Seniors Accumulated
                 $20,000 or More of Debt, School Years
                          1992-93 and 1995-96

                                                          Estimated
                                                        percentage of
                                                          graduating
                                                         seniors who
                                                         had borrowed
                                                          $20,000 or
                                                             more
                                                        --------------
                                                         1992-   1995-
Type of school                                              93      96
------------------------------------------------------  ------  ------
Public 4-year                                                6      16
Private 4-year                                              15      25
======================================================================
Total\a                                                      9      19
----------------------------------------------------------------------
\a Totals include graduating seniors at proprietary institutions. 

The most substantial increases in the number of graduating students
who borrowed occurred at public schools.  At 4-year public schools,
the percentage of graduating seniors who borrowed in 1 or more years
rose from 42 percent in 1992-93 to 60 percent in 1995-96 (see table
3.) This increase eliminated the earlier difference between public
and private 4-year schools in the percentage of students borrowing in
1 or more years--public school students "caught up" to private school
students in terms of the percentage of the group that borrowed. 
Students at private schools, however, still borrowed larger amounts
during both school years.  Students graduating from public schools
offering less than 4-year degrees also borrowed in substantially
higher numbers, although the average amount borrowed changed little
after taking inflation into account. 



                          Table 3
          
              Greatest Percentage Increase in
           Borrowers Has Come at Public Schools,
          While Dollar Amounts Are Still Greatest
          at 4-Year Private Schools, School Years
                    1992-93 and 1995-96

                      Percentage of
                    program completers    Average total
                    borrowing in 1 or       principal
                        more years          borrowed\a
                    ------------------  ------------------
Type of school       1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Public 4-year             42        60    $8,536   $11,554
Private 4-year            50        58    12,016    15,559
Public less-than-         28        38     4,959     4,009
 4-year
Private less-             53        53     5,848     8,129
 than-4-year
Proprietary               63        73     6,318     6,718
----------------------------------------------------------
\a Dollars are adjusted for inflation to 1995-96 values. 


--------------------
\10 Graduates of proprietary schools generally receive certificates
or awards; graduates of public less-than-4-year schools generally
receive associate degrees, and graduates of 4-year schools generally
receive bachelor's degrees. 


      BORROWING TRENDS FOR
      STUDENTS IN GRADUATE AND
      PROFESSIONAL SCHOOLS
---------------------------------------------------------- Letter :3.2

In the aggregate, borrowing by graduate and professional students
also increased.  In 1992-93, about 55 percent of graduate and
professional students who completed their degrees had borrowed in 1
or more years, and those who had borrowed had a cumulative debt (for
graduate, professional, and undergraduate education) averaging
$16,990.\11 By 1995-96, about 62 percent of this group borrowed, and
their cumulative debt averaged $24,340. 

Students in professional programs were the most likely to borrow and
had the highest levels of debt.  For 1995-96, students completing
professional programs had an average debt of $59,909, and the
percentage of students who borrowed more than $50,000 had increased
from 34 percent to 60 percent.\12 (See table 4.)



                          Table 4
          
               Most Students in Graduate and
             Professional Programs Borrow, With
          Increased Debt Concentrated Mainly Among
          Students Receiving Professional Degrees,
              School Years 1992-93 and 1995-96

                                            Percentage of
            Percentage of                   borrowers with
              recipients    Average total    principal of
            borrowing in 1    principal       $50,000 or
            or more years     borrowed\a         more
            --------------  --------------  --------------
Type of      1992-   1995-   1992-   1995-   1992-   1995-
program         93      96      93      96      93      96
----------  ------  ------  ------  ------  ------  ------
Master's        55      63  $12,87  $19,24       2       7
                                 0       5
Doctoral        56      59  22,973  18,045       8       9
Profession      79      73  45,100  59,909      34      60
 al
----------------------------------------------------------
\a Dollars are adjusted for inflation to 1995-96 values. 


--------------------
\11 Except where otherwise indicated we analyzed (1) debt for all
students who had completed their programs and (2) work patterns for
all full-time, full-year students, whether or not they completed
their program during NPSAS' survey year. 

\12 Cumulative debt for graduate and professional students includes
undergraduate loans.  Because of heavy borrowing by some students in
professional programs, median values would be more representative of
the typical amount borrowed; however, data were not available for us
to develop these estimates. 


   POSTSECONDARY STUDENTS
   TYPICALLY WORKED WHILE ENROLLED
------------------------------------------------------------ Letter :4

Changes in students' employment have been less pronounced than
changes in borrowing.  Compared with 1992-93, the percentage of
full-time undergraduate students who worked while attending school
rose slightly, while the percentage of graduate and professional
students who worked generally declined.  Among those who worked, the
average number of hours remained relatively steady. 


      WORK TRENDS FOR
      UNDERGRADUATES
---------------------------------------------------------- Letter :4.1

Most full-time undergraduate students worked during the school year
in both 1992-93 and 1995-96.  The percentage of full-time students
who worked rose in all three program categories--certificate or
award, associate degree, and bachelor's degree.\13 Overall, during
1995-96 more than two-thirds of full-time undergraduates worked while
enrolled.  On average, undergraduates worked 23 hours per week;
however, this varied considerably by program, with students in
associate and certificate or award programs working the most.  The
average number of hours worked per week did not change appreciably
from 1992-93, although it rose somewhat among students completing
associate degree programs.  (See table 5.)



                          Table 5
          
           A Greater Proportion of Undergraduates
           in All Types of Programs Worked, With
           Hours Worked Remaining Steady, School
                 Years 1992-93 and 1995-96

                      Percentage of
                    full-time students  Average number of
                     who worked while    hours worked per
                         enrolled              week
                    ------------------  ------------------
Program completed    1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Certificate or            56        71        28        26
 award
Associate degree          72        76        25        27
Bachelor's degree         61        72        20        21
----------------------------------------------------------
At 4-year and proprietary schools, the percentage of full-time,
full-year undergraduates who worked during the 1995-96 school year
was substantially higher than the percentage who worked in 1992-93. 
(See table 6.) Average hours worked per week did not change
significantly.  (See tables II.6, II.7, and II.8.)



                          Table 6
          
           A Greater Proportion of Undergraduates
           Attending All Types of Schools Worked,
          With Hours Worked Little Changed, School
                 Years 1992-93 and 1995-96

                      Percentage of
                        full-time
                    undergraduates who  Average number of
                       worked while      hours worked per
                         enrolled             week\
                    ------------------  ------------------
Type of school       1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Public 4-year             62        71        21        22
Private 4-year            60        73        18        19
Public less-than-         71        77        25        27
 4-year
Private less-             63        67        24        26
 than-4-year
Proprietary               55        70        30        30
----------------------------------------------------------

--------------------
\13 Although student employment levels rose from 1992-93 to 1995-96,
NPSAS data we reviewed from 1989-90 showed employment levels closer
to the 1995-96 rates, suggesting that the percentage of students who
work may be affected by factors other than the price of education,
such as job availability.  For example, during 1992-93, the national
unemployment rate was about 30 percent higher than in 1995-96. 


      WORK TRENDS FOR GRADUATE AND
      PROFESSIONAL STUDENTS
---------------------------------------------------------- Letter :4.2

Students in master's and doctoral programs in school year 1995-96
were more likely to work, and to work more hours per week, than were
students in professional programs.  Working students in professional
programs averaged about 20 hours of work a week, while those in
master's and doctoral programs averaged about 25 to 30 hours per
week.  Many of these students held jobs in their field of study, such
as teaching or research assistance.  About 80 percent of full-time
doctoral students who worked while enrolled said they held positions
directly related to their studies, compared with about 63 percent of
students in master's programs and about two-thirds of students in
professional programs.  However, even though more students in
master's and professional programs worked in off-campus jobs than did
doctoral students, most of them still regarded their jobs as closely
related to their field of study.  (See table 7.)



                          Table 7
          
           Graduate Students More Likely to Work
             and Work More Hours per Week Than
             Students in Professional Programs,
              School Years 1992-93 and 1995-96

                      Percentage of
                    full-time students  Average number of
                    working during the   hours worked per
                       school year             week
                    ------------------  ------------------
                     1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Master's                  72        69        24        26
Doctoral                  64        68        25        29
Professional              48        43        18        20
----------------------------------------------------------

   WORK AND BORROWING PATTERNS
   VARIED CONSIDERABLY FOR SOME
   GROUPS
------------------------------------------------------------ Letter :5

To gain a better understanding of student work and borrowing patterns
during school year 1995-96, we analyzed amounts borrowed and hours
worked by several factors, including type of school, cost of
attendance, year in school, dependency status, gender, family income,
race/ethnicity, cost of attendance, and expected family contribution. 
We focused this analysis on undergraduate students because the data
for graduate and professional students did not produce statistically
meaningful results when divided into many of the categories and
subcategories we analyzed.  (See app.  III for further details on our
analyses.)


      VARIATIONS IN BORROWING
---------------------------------------------------------- Letter :5.1

To help identify the relationships between the various factors
selected for analysis, we conducted a series of regression analyses. 
Regression analysis is a statistical technique that can analyze many
factors at the same time and estimate their relationship to a given
outcome.  In this case, our analyses were directed at determining
what factors, if any, help predict the amount of money that students
borrowed. 

Our results indicated that none of the factors we examined are strong
predictors of the amount of student borrowing.  Not surprisingly, the
most influential factor that emerged from our analyses was the cost
of the school attended.  However, this factor accounted for only
about 11 percent of the difference in the amounts of borrowing that
occurred, after controlling for other factors.  Several factors that
helped account for smaller portions of the variation in amounts
borrowed were the student's class level (freshmen, sophomore, junior,
or senior), the amount of grant aid received, and whether the student
was independent.  Other factors included in the analysis (type of
school, race/ethnicity, adjusted gross family income percentile,
expected family contribution, and hours worked per week while
enrolled) accounted for little, if any, variation.  Together, all of
the factors we examined accounted for about 31 percent of the
variation in the amounts borrowed. 

The relationship we saw in the regression analysis between the cost
of the school attended and the amounts borrowed is also apparent in
comparing graduating seniors' average cost of attendance for 1995-96
and the average cumulative amount of their borrowing.  As shown in
figure 2, seniors whose annual school costs were $20,000 or more
borrowed an average cumulative amount of about $18,000.  In contrast,
the comparable amount borrowed was about $11,000 for those whose
annual school costs were between $5,000 and $9,999. 

   Figure 2:  Variation in the
   Average Cumulative Amount That
   1995-96 Graduating Seniors
   Borrowed, by Cost of Attendance


   (See figure in printed
   edition.)

Appendix IV has additional data on variations in the average
cumulative amounts borrowed by undergraduates and variations in the
proportion who borrowed in 1 or more years as undergraduates.  For
example, appendix IV shows variations in these factors by student
year in school, by race/ethnicity, and by parental income. 


      VARIATIONS IN WORK PATTERNS
---------------------------------------------------------- Letter :5.2

As with borrowing, we conducted a regression analysis to determine
which factors, if any, would be strong predictors of how much
students will work.  We used the same list of factors as we did for
borrowing, but in this case, different factors emerged as important. 

Dependency status and type of school accounted for little of the
variation in hours worked (3.1 percent and 2.5 percent,
respectively).  Of students who worked, those who worked more hours
tended to be the independent students.  On average full-time,
full-year independent students who were employed while enrolled
worked about 28 hours per week, compared with an average of about 21
hours for their dependent counterparts.  (See table II.10 for further
details.)

Other factors included in our regression analysis each accounted for
less than 1 percent of the variation in hours worked.  (For
additional data on variation in work patterns, see apps.  III and
IV.)


   LITTLE INFORMATION AVAILABLE ON
   PARENTAL DEBT FOR CHILDREN'S
   EDUCATION
------------------------------------------------------------ Letter :6

In contrast with the substantial amount of information about
students' own borrowing experiences, little information is available
about the amounts that parents borrow to pay for their children's
postsecondary education.  In general, studies that provide data on
parents' education debt were dated or limited in scope, and they
often failed to differentiate between postsecondary education debt
and other types of education debt.  We found three studies that come
closest to describing the debt parents incur for their children's
postsecondary education.  Of these, the Department of Education's
work contained the most useful information. 


      INFORMATION FROM THE
      DEPARTMENT OF EDUCATION
---------------------------------------------------------- Letter :6.1

The best available data are in the Department of Education's NPSAS,
which we used as the basis for information on student borrowing and
work patterns.  As part of this survey, which is conducted
periodically, the Department collected some information through
telephone interviews with samples of parents. 

However, changes in NPSAS questions included in its 1995-96 survey
did not provide similar data that allowed for comparisons with
earlier survey results.  The most recent NPSAS (1995-96) included
parents' responses related only to certain groups of undergraduates,
such as dependent students who did not receive financial aid or those
whose schools' files did not include parents' adjusted gross income. 
Since such a sample of parents would not be representative of parents
of all undergraduates, estimates based on responses from that year's
survey are not included here. 

The 1992-93 NPSAS provided a more wide-ranging sampling of parents
selected to represent a group of graduating seniors.  Parents of
between 8 and 11 percent of seniors under 24 years of age who
graduated in 1992-93 reported borrowing to help finance their child's
education during 1992-93.  The average amount parents borrowed for
these seniors for 1992-93 was between $10,734 and $14,553.  Sources
of borrowing included home equity loans, home equity lines of credit,
signature loans, state- or school-sponsored parent loans, loans
against life insurance policies and retirement funds, commercial
loans, and federal PLUS loans.\14

Parents have borrowed a rapidly increasing amount of loan funds
through the Department of Education's PLUS program.  Parents of about
5 percent of dependent undergraduate students participated in this
program during 1995-96, about the same portion as in 1992-93.  Among
dependent students who graduated as seniors in 1995-96, about 10
percent had parents who had used the program during 1 or more years
of their child's postsecondary schooling.  The average cumulative
amount they borrowed was about $9,748.  NPSAS results indicate that
the average was about $9,022 for parents of students at public 4-year
schools and $10,673 for those at private 4-year schools. 

Amounts of PLUS borrowing have also risen in recent years, reflecting
the influence of higher loan limits.  According to the Department,
the average amount of these loans increased by about 55 percent (from
$3,588 to $5,556 in constant 1995-96 dollars) over the 1992-93 to
1995-96 period.  In the Higher Education Amendment of 1992, limits on
the amount of PLUS loans were lifted.  Currently, eligible parents
may borrow, regardless of financial need, up to their student's cost
of attendance, less the amount of other financial aid received. 


--------------------
\14 Federal PLUS loans are also referred to as federal Parent Loans
for Undergraduate Students. 


      OTHER SURVEY DATA
---------------------------------------------------------- Letter :6.2

Other survey data suggest that education has been an important use of
funds obtained from home equity loans.  Excluding first mortgages,
U.S.  home equity debt totaled about $255 billion in 1993, $110
billion of which was in home equity lines of credit and $145 billion
in traditional home equity loans.\15 \16 According to a school year
1993-94 survey by the University of Michigan, among borrowers using
home equity lines of credit, about 21 percent indicated that some or
all of these loan funds were used for education, up from 18 percent
in 1988.\17 Among borrowers using traditional equity loans, about 7
percent indicated that some or all of the funds were used for
education, up from 5 percent in 1988.\18 The survey did not indicate
what portion of these funds went for children's postsecondary
education and how much may have been used for other educational uses,
such as private elementary or secondary schools. 

The Federal Reserve Board's surveys of U.S.  households indicate that
education debt was about 1.9 percent of U.S.  household debt in 1989
and about 2.5 percent in 1992 and 1995.  However, the surveys are not
designed to capture parents' debt for their children's postsecondary
education.  The survey does not make a distinction between debt for
postsecondary education and debt for elementary and secondary
education, nor does it distinguish between debt owed by parents for a
child's education and debt owed by parents for their own education. 


--------------------
\15 These are Federal Reserve Board estimates. 

\16 Unlike other dollar amounts in this report, these numbers from
the Federal Reserve Board have not been adjusted for inflation. 

\17 The confidence intervals for these estimates are 15 to 27 percent
and 11 to 25 percent, respectively. 

\18 The confidence intervals for these figures are 2 to 12 percent
and 1 to 9 percent, respectively. 


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

The Department of Education reviewed a draft of this report and had
no formal comments, although it provided several technical
suggestions that we incorporated as appropriate. 


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

We are sending copies of this report to the Secretary of Education,
appropriate congressional committees, and other interested parties. 

Please call me at (202) 512-7014 if you or your staff have any
questions regarding this report.  Major contributors included Joseph
J.  Eglin, Jr., Assistant Director; Charles M.  Novak; Benjamin P. 
Pfeiffer; and Dianne L.  Whitman-Miner. 

Sincerely yours,

Carlotta C.  Joyner
Director, Education
 and Employment Issues


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

To analyze working and borrowing patterns among postsecondary
students, we reviewed literature and data from the Department of
Education and other sources, such as various professional
associations.  The data we analyzed included the Department of
Education's periodic National Postsecondary Student Aid Study
(NPSAS), the Federal Reserve Board's Survey of Consumer Finances,
Claritas Inc.'s (a private research firm) survey on use of credit
cards, and the University of Michigan's National Survey of Home
Equity Loans.  In connection with this effort, we also interviewed
Department of Education officials and staff of professional
associations and the Federal Reserve Board. 

The Department's NPSAS addresses how students and their families pay
for postsecondary education and involves nationally representative
samples of all students in postsecondary institutions.  In 1995-96,
for example, the Department selected a sample of over 950
institutions and about 50,000 students.  The researchers gathered
data about students from schools' institutional records and the
Department's records (including financial aid applications and the
National Student Loan Data System).  They also gathered information
by telephoning a subsample of about 27,000 undergraduates and about
4,000 graduate and professional students. 


   DATA USED FOR ASSESSING
   UNDERGRADUATES' BORROWING AND
   WORK PATTERNS
--------------------------------------------------------- Appendix I:1

We focused our analysis on the average amounts of borrowing and
average cumulative debt reported in the NPSAS for school years
1992-93 and 1995-96.\19

Unless otherwise indicated, the term "debt" in this report refers to
the cumulative total of the principal amounts borrowed by students
for undergraduate education (borrowing for all the costs of
attendance, including room and board).  The data on the amount of
students' cumulative debt were self-reported and, according to the
Department's NPSAS project officer, the extent to which it includes
credit card debt is unknown.  The portion of college students with
credit cards rose from about one-half to about two-thirds from 1990
to 1996, according to a study by Claritas Inc.  The estimated
aggregate average balance grew from about $900 in 1990 to about
$2,250 in the third quarter of 1997.  (These amounts have not been
adjusted for inflation, and Claritas Inc.  did not provide confidence
intervals for these numbers.)

Average annual amounts of borrowing came from NPSAS analysis of
school records for over 50,000 undergraduate students and Department
of Education records for students with federal student loans.  Data
on cumulative debt came from telephone surveys of about 27,000
respondents to NPSAS telephone surveys.  About 1,500 of these were
graduating seniors.  The 1989-90 NPSAS survey did not identify
students who completed their degree program in that year, so we
limited our analysis of those data to a comparison of 1992-93 and
1995-96 survey results.\20 We did use the 1989-90 survey as a point
of comparison for the overall portion of undergraduates who worked
while enrolled. 

Similarly, we focused our analysis of undergraduate students' work
patterns on students included in NPSAS' 1992-93 and 1995-96 surveys
who enrolled as undergraduates for their first term during the May 1
through April 30 time period, and attended full-time for a full year
(9 months).\21 \22

To assess the number of hours worked by undergraduate students while
enrolled, we used NPSAS for 1992-93 and 1995-96.  These data came
from a computer-aided telephone interview. 

To assess parents' borrowing for their child's postsecondary
education, we used parent responses to NPSAS' 1992-93 survey, Federal
Reserve Board data from its Survey of Consumer Finances for 1995, and
the University of Michigan's National Survey of Home Equity Loans. 


--------------------
\19 The median amount borrowed (the amount at which half of the
borrowers had larger amounts and half had lower amounts) may differ
from the average amount borrowed by group.  Because they are less
susceptible to being skewed by students who borrow at the highest
levels, median amounts borrowed is preferable to averages.  However,
it was not feasible for us to develop estimates of median values
because the Department has not published median amounts borrowed and
the data analysis system it provides does not provide functions for
the computation of medians. 

\20 Estimates for 1992-93 were not strictly comparable with those for
1995-96, because the researchers assumed that the amount a student
borrowed from parents, relatives, or friends was $0 if the sampled
student was not interviewed.  For 1995-96, we weighted the estimates
to include only sampled students who were interviewed.  Comparable
weights were not available for 1992-93. 

\21 Although NPSAS' 1995-96 survey broadened the definition of
full-year students to include those who were enrolled for 8 months,
we retained the 9-month definition in order to provide data that are
consistent with data from prior surveys. 

\22 To evaluate employment trends, we used a somewhat different group
than we had used to evaluate borrowing.  Instead of focusing on
students who graduated in 1992-93 or 1995-96 (so that we could
analyze their borrowing history over several years), we based this
analysis on all students who attended school full time during 1992-93
or 1995-96.  To maintain our focus on full-time students, we excluded
those students who attended only part time or who did not attend for
the full school year.  Our estimates are for students employed at
some point during the year. 


   DATA USED FOR ASSESSING
   GRADUATE AND PROFESSIONAL
   BORROWING AND WORK PATTERNS
--------------------------------------------------------- Appendix I:2

Our analysis of graduate and professional students included those in
NPSAS who were enrolled in a postbaccalaureate program that began
between May 1 and April 30 in the 1992-93 or 1995-96 NPSAS years.\23
Data on hours worked while enrolled came from NPSAS telephone
interviews with about 4,000 students.  We limited our analysis of
hours worked and earnings to those who were enrolled full time for a
full year (9 months).  Data on cumulative borrowing came from NPSAS
telephone interviews of about 2,800 graduate students and about 1,200
professional students. 

Because we were unable to identify students who were in the last year
of their graduate or professional degree program in school year
1989-90 or who completed their degree during that year, we limited
our analysis of graduate and professional degree students' cumulative
debt to 1992-93 and 1995-96. 


--------------------
\23 Students identified as "professional" were first professional
students in programs leading toward degrees in chiropractic therapy,
dentistry, law, medicine, optometry, osteopathy, pharmacy, podiatry,
theology, and veterinary medicine. 


   STATISTICAL TESTS
--------------------------------------------------------- Appendix I:3

Analysts use various statistical techniques to help evaluate the
relative strength of relationships that can be found in sets of data. 
To calculate confidence intervals for survey results, we used
standard errors provided by the Department and a 95-percent
confidence interval.\24 Similarly, we tested for the statistical
significance of differences between groups using t-tests and a p =
0.05 criterion. 

To further assess statistical relationships between variables
discussed in this report, we performed two linear regression analyses
with the following dependent variables:  (1) the amount
undergraduates borrowed for 1995-96 and (2) the average hours
full-time, full-year undergraduates worked per week while enrolled
during 1995-96. 


--------------------
\24 The Federal Reserve Board did not publish confidence intervals
for its estimates, but indicated that they are "subject to a fairly
wide margin of error."


   CONSTANT DOLLAR CALCULATIONS
--------------------------------------------------------- Appendix I:4

To indicate the extent to which borrowing and debt have changed at a
rate faster or slower than changes in consumer prices, we analyzed
levels of cumulative borrowing in constant 1995-96 dollars.  To
calculate constant 1995-96 dollars we used the Bureau of Labor
Statistics' Consumer Price Index for all urban consumers.\25

We conducted our work from April to December 1997 in accordance with
generally accepted government auditing standards.  Because the
Department uses several methods to check and review NPSAS data and
these data are widely relied upon in the education community, we did
not validate the reliability of the data derived from the sources
indicated. 


--------------------
\25 While we are aware that the Consumer Price Index has limitations,
we chose to use it to adjust for inflation because we concluded that
it was the best available index for this purpose.  See, for example,
Consumer Price Index:  More Frequent Updating of Market Basket
Expenditure Weights Is Needed (GAO/GGD/OCE-98-2, Oct.  9, 1997). 


ADDITIONAL DATA CONCERNING
POSTSECONDARY EDUCATION DEBT AND
WORK DURING PERIODS OF ENROLLMENT
========================================================== Appendix II

The tables in this appendix contain additional details regarding the
information presented in the letter portion of this report.  The
tables present category-by-category estimates for various aspects of
student debt and work, along with confidence intervals for each.  The
estimated averages shown are based on analysis of the results from a
sample of students.  The confidence intervals are the ranges in which
the averages are likely to fall for the entire population of
postsecondary students within the category indicated. 

The table notes indicate whether differences in the estimated
averages for various sample groups are statistically significant.  We
identified differences as statistically significant when our
statistical tests showed less than a 5-percent chance that the
differences between groups occurred purely by chance. 



                         Table II.1
          
          Role of Borrowing and Working in Paying
             Costs for the "Average" Full-Time
            Student, by Type of School, 1995-96

                                  Estimated average
                            ------------------------------
                                                Confidence
Type of school                      Amount        interval
--------------------------  --------------  --------------
Average cost of attendance\a
----------------------------------------------------------
Public 4-year                      $10,883       $10,559 -
                                                   $11,108
Private 4-year                    19,946\b        19,175 -
                                                    20,716
Public less-than-4-year            7,275\b    6,901 -7,650
Private less-than-4-year          12,912\b        12,150 -
                                                    13,675
Proprietary                       14,112\b        13,383 -
                                                    14,481

Average amount borrowed for 1995-96\c,d
----------------------------------------------------------
Public 4-year                        1,944    1,818 -2,070
Private 4-year                     2,660\b    2,477 -2,843
Public less-than-4-year              639\b        487 -791
Private less-than-4-year             2,329    1,362 -3,296
Proprietary                        3,128\b    2,688 -3,568

Average amount earned while enrolled\c
----------------------------------------------------------
Public 4-year                        3,282    3,010 -3,553
Private 4-year                       3,145    2,719 -3,572
Public less-than-4-year            5,313\b    4,087 -6,539
Private less-than-4-year             4,806    3,205 -6,408
Proprietary                        6,006\b    3,847 -8,165

Average total amount of grants and scholarships received\c
----------------------------------------------------------
Public 4-year                        1,492    1,371 -1,612
Private 4-year                     4,823\b    4,414 -5,232
Public less-than-4-year              803\b        672 -933
Private less-than-4-year             1,972    1,453 -2,491
Proprietary                          1,572    1,361 -1,784
----------------------------------------------------------
Note:  These data are for full-time, full-year undergraduates only. 
We classified postsecondary institutions as follows:  (1) public
4-year institutions that offer a baccalaureate degree or more
advanced program; (2) private, not-for-profit 4-year institutions
that offer a baccalaureate degree or more advanced program; (3)
public less-than-4-year institutions that only offer programs below
the baccalaureate level; (4) private less-than-4-year institutions
that only offer programs below the baccalaureate level; and (5)
proprietary (for-profit) institutions. 

\a The cost of attendance includes tuition, fees, room, board, books,
supplies, transportation, and other necessary living costs identified
in the school's budget. 

\b The estimated value differs to a statistically significant extent
from the estimated value for public 4-year schools. 

\c These averages were calculated including "0" for undergraduates
who did not borrow, did not work while enrolled, or did not receive
grants. 

\d The amount of loans indicated excludes loans from parents,
friends, and relatives. 



                         Table II.2
          
             Undergraduate Degree or Credential
            Recipients Who Borrowed in 1 or More
          Years, School Years 1992-93 and 1995-96

                                      Estimated average
             Estimated percentage     cumulative amount
              of recipients who       borrowed (constant
                   borrowed            1995-96 dollars)
            ----------------------  ----------------------
Degree or
other
credential              Confidence              Confidence
received    Percentage    interval      Amount    interval
----------  ----------  ----------  ----------  ----------
1992-93
----------------------------------------------------------
Certificat        39.5      35.1 -      $5,171    $4,758 -
 e or                         43.9                  $5,584
 award
Associate       34.5\a      30.6 -      $5,580     4,876 -
 degree                       38.4                   6,284
Bachelor's      46.3\a      44.6 -    10,080\a     9,762 -
 degree\b                     47.9                  10,398
Average           41.1      39.3 -     7,834\a     7,489 -
                              42.9                   8,178

1995-96
----------------------------------------------------------
Certificat      52.7\c      41.5 -       5,597     4,420 -
 e of                         63.8                   6,774
 award
Associate         42.0      34.4 -       5,059     4,243 -
 degree                       49.6                   5,875
Bachelor's      60.1\c      56.4 -  13,269\a,c    12,364 -
 degree\b                     63.9                  14,173
Average         52.4\c      48.7 -  9,701\a,\c     9,013 -
                              56.0                  12,587
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the estimated value for certificate or award recipients. 

\b For bachelor's degree recipients we used figures for students
identified as graduating seniors. 

\c This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table II.3
          
             Percentage of Graduates of 4-Year
           Schools Who Borrowed $20,000 or More,
              School Years 1992-93 and 1995-96

                    Percentage of graduates borrowing in 1
                     or more years a total of $20,000 or
                                     more
                    --------------------------------------
                                                Confidence
Type of school              Percentage            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Public 4-year                      5.6            4.7 -6.5
Private 4-year                  15.2\a          13.0 -17.3
Total                            9.1\a           8.1 -10.2

1995-96
----------------------------------------------------------
Public 4-year                   15.6\b          11.5 -19.7
Private 4-year                25.0\a,b          19.5 -30.4
==========================================================
Total                           19.0\b          15.7 -22.2
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the estimated value for those attending public 4-year schools. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table II.4
          
            Undergraduate Program Completers Who
           Borrowed and Average Amount Borrowed,
              School Years 1992-93 and 1995-96

                                        Average total
            Percentage of program     principal borrowed
                completers who        (constant 1995-96
                   borrowed                dollars)
            ----------------------  ----------------------
Type of                 Confidence              Confidence
school         Percent    interval    Amount      interval
----------  ----------  ----------  --------  ------------
1992-93
----------------------------------------------------------
Public 4-         42.4      40.5 -    $8,536      $8,205 -
 year                         44.2                  $8,868
Private 4-      50.0\a      47.1 -  12,016\a      11,350 -
 year                         52.8                  12,683
Public          27.7\a      24.7 -   4,959\a       4,267 -
 less-                        30.8                   5,651
 than-4-
 year
Private           52.7      42.1 -   5,848\a       3,643 -
 less-                        63.2                   8,053
 than-4-
 year
Proprietar      62.6\a      56.5 -   6,318\a       5,665 -
 y                            68.8                   6,972

1995-96
----------------------------------------------------------
Public 4-       60.2\b      55.8 -  11,554\b      10,521 -
 year                         64.6                  12,587
Private 4-      57.6\b      52.6 -  15,559\a      14,450 -
 year                         62.5        ,b        16,669
Public        38.4\a,b      30.8 -   4,009\a       3,176 -
 less-                        46.1                   4,842
 than-4-
 year
Private           53.3      34.3 -   8,129\a       6,869 -
 less-                        72.2                   9,338
 than-4-
 year
Proprietar      71.6\a      63.4 -   6,718\a       5,781 -
 y                            81.7                   7,655
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for students at public 4-year
schools. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                                        Table II.5
                         
                          Students in Graduate and Professional
                         Programs Who Borrowed, Amount Borrowed,
                           and Percentage With $50,000 or More
                          Debt, School Years 1992-93 and 1995-96

                                         Estimated average amount
                                         of cumulative borrowing    Estimated percentage
                                            for undergraduate,     of recipient borrowers
                  Estimated percentage         graduate, or         borrowing a total of
                   of recipients who      professional education    $50,000 or more for
                 borrowed in 1 or more      (constant 1995-96          postsecondary
                         years                   dollars)                education
                 ----------------------  ------------------------  ----------------------
                             Confidence                Confidence              Confidence
Degree received   Percent\a    interval    Amount\b      interval     Percent    interval
---------------  ----------  ----------  ----------  ------------  ----------  ----------
1992-93
-----------------------------------------------------------------------------------------
Master's               54.6      51.5 -     $12,870     $11,806 -         1.9    0.7 -3.0
                                   57.7                   $13,934
Doctoral               55.5      47.6 -    22,973\c      18,516 -       8.3\c   5.5 -11.2
                                   63.4                    27,431
Professional         79.0\c      73.7 -    45,100\c      40,191 -      34.1\c      28.9 -
                                   84.2                    50,009                    39.4

1995-96
-----------------------------------------------------------------------------------------
Master's               62.6      54.2 -    19,245\d      14,685 -       6.9\d   2.6 -11.2
                                   71.0                    23,806
Doctoral               59.2      38.5 -      18,045      11,605 -         9.2     0 -19.3
                                   79.8                    24,485
Professional         73.3\c      63.8 -  59,909\c,d      52,072 -    60.2\c,d      49.7 -
                                   82.8                    67,746                    70.6
-----------------------------------------------------------------------------------------
\a Numbers are estimated averages for recipients who borrowed either
for undergraduate education, for graduate or professional education,
or both, whether or not any loans were still owed when they completed
their graduate or professional programs. 

\b Numbers are estimates of the average cumulative amounts of loans
for undergraduate, graduate, and professional education. 

\c This estimate differs to a statistically significant extent from
the estimate for master's degree recipients. 

\d This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table II.6
          
              Portion of Full-Time, Full-Year
           Undergraduates Who Worked and Average
            Hours Worked per Week, School Years
                    1992-93 and 1995-96

                        Estimated
                      percentage of
                     full-time, full-
                           year
                    undergraduates who  Estimated average
                       worked while      number of hours
                         enrolled        worked per week
                    ------------------  ------------------
                              Confiden            Confiden
                                    ce                  ce
Program              Percent  interval     Hours  interval
------------------  --------  --------  --------  --------
1992-93
----------------------------------------------------------
Certificate or          56.0    51.7 -      27.9    26.6 -
 award                            60.4                29.2
Associate degree      72.1\a    68.8 -    24.8\a    23.9 -
                                  75.3                25.8
Bachelor's degree     61.5\a    60.2 -    20.3\a    19.9 -
                                  62.8                20.7

1995-96
----------------------------------------------------------
Certificate or        71.3\b    66.9 -      26.0    24.0 -
 award                            75.8                28.0
Associate degree      75.6\a    70.9 -    26.5\b    25.3 -
                                  80.3                27.8
Bachelor's degree     71.7\b    70.0 -    20.8\a    20.2 -
                                  73.4                21.4
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for students in certificate or
award programs. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table II.7
          
             Percentage of Full-Time, Full-Year
          Undergraduates Who Worked While Enrolled
             and Average Hours Worked per Week,
              School Years 1992-93 and 1995-96

             Estimated percentage
            of undergraduates who     Estimated average
                 worked while       number of hours worked
                  enrolled\a               per week
            ----------------------  ----------------------
Type of                 Confidence              Confidence
school         Percent    interval       Hours    interval
----------  ----------  ----------  ----------  ----------
1992-93
----------------------------------------------------------
Public 4-         62.1      60.6 -        21.1      20.7 -
 year                         63.6                    21.6
Private 4-        60.1      58.0 -      18.5\b      17.5 -
 year                         62.1                    19.4
Public          71.3\b      67.9 -      25.3\b      24.3 -
 less-                        74.7                    26.3
 than-4-
 year
Private           62.9      51.2 -      24.0\b      21.2 -
 less-                        74.6                    26.8
 than-4-
 year
Proprietar      55.5\b      50.3 -      30.0\b      28.6 -
 y                            60.6                    31.4

1995-96
----------------------------------------------------------
Public 4-       70.8\c      68.6 -        21.6      20.9 -
 year                         73.0                    22.2
Private 4-      72.8\c      70.5 -      18.8\b      17.8 -
 year                         75.2                    19.8
Public            76.6      71.9 -      26.8\b      22.4 -
 less-                        81.4                    23.7
 than-4-
 year
Private           66.6      58.7 -      26.4\b      22.9 -
 less-                        74.5                    29.9
 than-4-
 year
Proprietar      70.2\c      65.3 -      29.9\b      27.9 -
 y                            75.1                    32.0
----------------------------------------------------------
\a Numbers include students who worked for only a portion of the time
they were enrolled during the year.  On average, full-time, full-year
students who worked while enrolled during 1995-96 worked about 8
months. 

\b This estimated value differs to a statistically significant extent
from the corresponding estimated value for public 4-year
undergraduates. 

\c This estimated value differs to a statistically significant extent
from that for 1992-93. 



                         Table II.8
          
            Percentage of Full-Time Graduate and
            Professional Students Working While
           Enrolled and Average Hours Worked per
           Week, School Years 1992-93 and 1995-96

                                      Estimated average
             Estimated percentage   hours worked per week
            ----------------------  ----------------------
Degree                  Confidence              Confidence
program        Percent    interval       Hours    interval
----------  ----------  ----------  ----------  ----------
1992-93
----------------------------------------------------------
Master's          71.7      67.9 -        24.5      23.3 -
                              75.5                    25.7
Doctoral        63.4\a      58.0 -        24.9      22.9 -
                              68.9                    27.0
Profession      47.6\a      41.8 -      18.4\a      16.7 -
 al                           53.4                    20.1

1995-96\b
----------------------------------------------------------
Master's          69.0      61.5 -        25.9      23.3 -
                              76.6                    28.4
Doctoral          67.5      54.5 -        29.2      24.2 -
                              80.6                    34.1
Profession      43.1\a      37.8 -      20.1\a      17.3 -
 al                           48.3                    22.9
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for master's degree
recipients. 

\b None of the estimated values for 1995-96 differed to a
statistically significant extent from the estimated values for
1992-93. 



                         Table II.9
          
           Variation in Average Cumulative Amount
          1995-96 Graduating Seniors Borrowed, by
                     Cost of Attendance

                    Average cumulative amount borrowed for
                     those who borrowed during 1 or more
                                    years
                    --------------------------------------
1995-96 cost of                                 Confidence
attendance\a                    Amount            interval
------------------  ------------------  ------------------
$5,000 -$9,999                 $10,579     $9,274 -$11,885
$10,000 -$14,999              12,821\b      11,024 -14,617
$15,000 -$19,999              13,714\b      12,138 -15,289
$20,000 -$24,999              17,673\b      14,928 -20,417
$25,000 or more               17,663\b      14,490 -20,836
----------------------------------------------------------
\a This is the institution's budgeted amount for the costs of
attending the institution full-time for a school year, including
tuition, fees, books, supplies, room, board, and transportation. 
Each institution identified separate amounts for students on the
basis of dependency status and residence status (whether they live at
home, for example). 

\b This estimate differs to a statistically significant extent from
the estimate for schools with a cost of attendance between $5,000 and
$9,999. 



                        Table II.10
          
          Variation in the Average Number of Hours
          Worked per Week by Full-Time, Full-Year
           Undergraduates, by Dependency Status,
                    School Year 1995-96

                     Average hours worked per week while
                                  enrolled\a
                    --------------------------------------
                                                Confidence
Dependency status                Hours            interval
------------------  ------------------  ------------------
Dependent                         21.2          20.5 -21.8
Independent                       27.6          26.5 -28.7
----------------------------------------------------------
Note:  The Higher Education Act of 1965, as amended, identifies
students as independent if any of the following apply:  the student
(1) is 24 years of age or older, (2) is married, (3) is a graduate or
professional student, (4) has legal dependents other than a spouse,
(5) is an orphan or ward of the court, (6) is a veteran, or (7) is
determined and documented by his or her financial aid administrator
to be independent because of other unusual circumstances.  If none of
these apply, the student is classified as dependent. 

\a These numbers are 1995-96 averages for full-time, full-year
undergraduates who worked at some time during their enrollment.  The
estimate for dependent students differs from the estimate for
independent students to a statistically significant extent. 


ANALYSES OF UNDERGRADUATE WORK AND
BORROWING PATTERNS
========================================================= Appendix III

Although our work focused primarily on the extent to which borrowing
and working by undergraduates varied by each of several factors (type
of school and year in school, for example), we sought more
information about the extent to which these factors were predictive. 
To do this, we performed a series of regression analyses.\26 Each
analysis indicates what portion of variance in the working or
borrowing variable examined was accounted for by each factor after
taking the other factors into account. 

In tables III.1 and III.2, the portion of the variance accounted for
is the change in the portion of variance accounted for (R\2 expressed
as a percentage) after adding each variable to the model after
including (controlling for) all the other variables listed. 

"Total accounted for" is the percentage of variance accounted for
including all variables listed.  This is the coefficient of
determination, a statistic that indicates how well a statistical
model fits the data.  If there is no linear relationship between
dependent and independent variables, R\2 equals 0; if there is a
perfect statistical relationship, R\2 equals 1 (100 percent). 

The regression coefficients (B) shown in each table indicate the
extent to which a change in each independent variable is associated
with a change in the dependent variable.  For example, in table
III.1, the regression coefficient for graduating seniors is
$1,632.75.  This indicates that after taking into account the
relationships between all the variables listed, graduating seniors
borrowed an average of about $110.67 less than freshmen in their
first year of postsecondary education (the reference category). 

The standardized regression coefficients (beta) shown in each table
are statistics that are standardized to allow comparison when the
independent variables are measured in different units.  They help
analysts compare the extent to which variables help predict variation
in the dependent variable, such as the amount borrowed.  The unit of
measure for beta weights is a standard deviation in the dependent
variable.  (Standard deviations are measures of the extent to which,
for example, the amounts students borrowed typically differed from
the average amount borrowed.) A beta weight is an estimate of the
number of standard deviations more a student is expected to borrow
for a one standard deviation increase in an independent variable (see
table III.1). 

The significance test (probability based on the t-statistic) in each
table indicates, for the addition of each variable in the model, the
probability that the statistical relationship between each
independent variable and the variation in the dependent variable not
accounted for by other variables is due to random factors. 

In the analysis of the statistical relationship between each
dependent variable and each categorical variable, such as year in
school, we identified a reference category.  The tables provide
regression statistics that indicate the extent to which nonreference
groups compare statistically with the reference group.  In both
tables, reference groups are white non-Hispanic, dependent, men,
first-time beginning freshmen, and attending public 4-year schools. 



                                   Table III.1
                     
                       Regression Results for the Amount of
                      Annual Borrowing by Undergraduates for
                               School Year 1995-96

                            Portion of                Standardized   Probability
                              variance    Regression    regression  based on the
                             accounted   coefficient   coefficient             t
Variable                         for\a           (B)        (beta)   statistic\b
------------------------  ------------  ------------  ------------  ------------
Type of school (portion of variance accounted for = 0.18%)
--------------------------------------------------------------------------------
Public 2-year                              $ -110.67         -0.02        0.2078
Public less-than-2-year                       346.57          0.01        0.6299
Private 4-year                               -186.32         -0.04        0.0196
Private 2-year                                185.81          0.01        0.3390
Private less-than-4-                         -760.94         -0.01        0.6039
 year
Proprietary 4-year                            362.37          0.02        0.1471
Proprietary 2-year                             35.43          0.00        0.7839
Proprietary less-than-                         82.37          0.01        0.5144
 2-year

Class level (portion of variance accounted for = 6.46%)
--------------------------------------------------------------------------------
Freshmen--not beginning                       307.33          0.05        0.0004
 postsecondary education
 for the first time
Sophomore                                     682.81          0.12      < 0.0001
Junior                                      1,562.51          0.26      < 0.0001
Senior, 4th year                            1,697.56          0.21      < 0.0001
Senior, 5th + year                          1,878.90          0.07      < 0.0001
Graduating senior                           1,632.75          0.22      < 0.0001
Other                                         631.38          0.04        0.0005

Racial/ethnic group (portion of variance accounted for = 0.24%)
--------------------------------------------------------------------------------
Black, non-Hispanic                          -196.02         -0.03        0.0100
Hispanic                                     -301.85         -0.04        0.0013
Asian/Pacific Islander                       -244.82         -0.02        0.0457
American Indian/Alaskan                      -481.29         -0.02        0.1016
 Native
Other/unknown                                -132.93          0.00        0.6749

Other portion of variances accounted for
--------------------------------------------------------------------------------
Gender--women                    0.02%        -61.57         -0.01        0.2442
Cost of attendance              10.14%          0.19          0.51      < 0.0001
Adjusted gross family            0.03%         -2.28         -0.03        0.1109
 income percentile
Expected family                  0.53%         -0.04         -0.10      < 0.0001
 contribution
Grants received for              2.21%         -0.14         -0.22      < 0.0001
 1995-96
Hours worked per week            0.07%         -4.21         -0.03        0.0189
 while enrolled
Independent student              1.64%        778.00          0.17      < 0.0001
================================================================================
Total accounted for           30.85%\c           N/A           N/A           N/A
================================================================================
Total unaccounted for           69.15%           N/A           N/A           N/A
--------------------------------------------------------------------------------
Note:  N/A = not applicable. 

\a Change in R\2 from the addition of each variable or group of
variables after including (controlling for) the other variables
listed. 

\b These numbers reflect an adjustment for design effect. 

\c Adjusted R\2 was 0.3068, with 11,171 degrees of freedom. 



                                   Table III.2
                     
                       Regression Results for the Number of
                       Hours Worked per Week by Full-Time,
                       Full-Year Undergraduates Who Worked
                                  While Enrolled

                            Portion of                Standardized
                              variance    Regression    regression   Probability
                             accounted   coefficient   coefficient  based on the
Variable                         for\a           (B)        (beta)   t statistic
------------------------  ------------  ------------  ------------  ------------
Type of school (portion of variance accounted for = 2.55%)
--------------------------------------------------------------------------------
Public 2-year                                  $3.52          0.13      < 0.0001
Public less-than-2-year                         2.98          0.03        0.0833
Private 4-year                                  1.09          0.04        0.0997
Private 2-year                                  2.70          0.02        0.1404
Private less-than-4-                           13.79          0.04        0.0312
 year
Proprietary 4-year                             14.55          0.12      < 0.0001
Proprietary 2-year                              6.46          0.07      < 0.0001
Proprietary less-than-                          4.37          0.05        0.0074
 2-year

Class level (portion of variance accounted for = 0.65%)
--------------------------------------------------------------------------------
Freshmen -not beginning                         1.97          0.06        0.0017
 postsecondary education
 for the first time
Sophomore                                       2.06          0.07        0.0002
Junior                                          2.21          0.07        0.0004
Senior, 4th year                                1.32          0.03        0.1029
Senior, 5th + year                              0.74          0.00        0.8098
Graduating senior                               0.57          0.01        0.4683
Other                                           3.03          0.04        0.0196

Racial/ethnic group (portion of variance accounted for = 0.51%)
--------------------------------------------------------------------------------
Black, non-Hispanic                             1.51          0.04        0.0214
Hispanic                                        1.46          0.04        0.0353
Asian/Pacific Islander                         -2.19         -0.04        0.0083
American Indian/Alaskan                        -0.78         -0.01        0.7374
 Native
Other/unknown                                  -2.40         -0.02        0.3502

Other portion of variances accounted for
--------------------------------------------------------------------------------
Gender--women                    0.84%         -2.13         -0.09      < 0.0001
Cost of attendance               0.71%          0.00         -0.14      < 0.0001
Adjusted gross family            0.33%          0.04          0.09        0.0004
 income percentile
Expected family                  0.08%          0.00         -0.04        0.0858
 contribution
Grants received for              0.35%          0.00         -0.08        0.0003
 1995-96
Borrowing for 1995-96            0.01%          0.00         -0.01        0.6187
Independent student              3.08%          6.12          0.23      < 0.0001
================================================================================
Total accounted for           17.07%\c           N/A           N/A           N/A
================================================================================
Total unaccounted for           82.93%           N/A           N/A           N/A
--------------------------------------------------------------------------------
Note:  N/A = not applicable. 

\a Change in R\2 from the addition of each variable or group of
variables after including (controlling for) the other variables
listed. 

\b These numbers reflect an adjustment for design effect. 

\c Adjusted R\2 was 0.1681, with 8,682 degrees of freedom. 


--------------------
\26 The Department of Education's National Postsecondary Student Aid
Study, from which we obtained our data, is based on a survey of
students.  As survey respondents are divided into various categories
(such as the number of students in professional programs by family
income), the number of students sampled who have these
characteristics may be too small to allow reliable estimates.  For
students in graduate and professional programs, the confidence
intervals for such estimates became so great that we judged them not
meaningful enough to analyze in the manner we analyzed undergraduate
students.  To analyze amounts borrowed for the 1995-96 school year,
we limited our analysis to those undergraduates who had borrowed.  To
analyze the average number of hours worked per week, we limited our
analysis to those full-time, full-year undergraduates who worked
while enrolled during some portion of the school year. 


ADDITIONAL INFORMATION ON
VARIATION IN AVERAGE CUMULATIVE
BORROWING AND WORK PATTERNS AMONG
UNDERGRADUATES
========================================================== Appendix IV

We were also asked to analyze borrowing and work patterns in relation
to four other factors:  students' class level, students' dependency
status, parental income, and students' race/ethnicity.  This appendix
contains our findings with respect to these factors and concludes
with tables that provide additional information to supplement that
shown in tables IV.1 through IV.5 and figure IV.1. 


   VARIATION IN BORROWING PATTERNS
-------------------------------------------------------- Appendix IV:1


      CLASS LEVEL
------------------------------------------------------ Appendix IV:1.1

Not surprisingly, students who had attended school for several years
were more likely to have borrowed--and in greater amounts--than their
counterparts who had not been in school as long.  A greater portion
of students borrowed at all undergraduate levels, and the amounts
they borrowed increased to a statistically significant extent for
everyone but freshmen (see table IV.1). 



                         Table IV.1
          
          Percentage of Borrowers Was Up Among All
          Undergraduate Years, With Dollar Amounts
           Up in Each Year as Well, School Years
                    1992-93 and 1995-96

                Percentage of
                undergraduates
            borrowing in 1 or more      Average total
                    years            principal borrowed\a
            ----------------------  ----------------------
Class
level          1992-93     1995-96     1992-93     1995-96
----------  ----------  ----------  ----------  ----------
Freshman            38          41      $4,167      $4,296
Sophomore           40          49       5,428       6,318
Junior              49          61       7,710       9,487
Senior              49          61       9,187      12,063
----------------------------------------------------------
\a Cumulative amount of borrowing for undergraduate education
adjusted for inflation to 1995-96 values. 


      DEPENDENCY STATUS AND
      PARENTAL INCOME
------------------------------------------------------ Appendix IV:1.2

Students who were classified by the Department of Education's
financial aid needs analysis process as dependent on their parents\27
were less apt to borrow than those who were classified as
independent, but when they borrowed, they tended to borrow larger
amounts.  Among seniors graduating in 1995-96, 51 percent of those
who were dependent on their parents borrowed in 1 or more years,
compared with 71 percent of independent students.  On average,
dependent students borrowed $13,754, compared with $12,842 for
independent students. 

Comparing 1995-96 graduating dependent seniors with their
counterparts in 1992-93, borrowing was up across all income levels. 
As figure IV.1 shows, borrowing tended to be most common among
dependent students from families whose annual income is less than
$45,000.  However, the portion of dependent students who borrowed
increased at all family income levels, and at the highest level
($100,000 and above), it nearly doubled from 16.3 percent to 32.6
percent.  The increase in amounts borrowed was relatively uniform
among all income groups except the lowest and highest. 

   Figure IV.1:  Undergraduate
   Borrowing Among Dependent
   Students, by Parental Income,
   School Years 1992-93 and
   1995-96

   (See figure in printed
   edition.)


--------------------
\27 In general, the Higher Education Act identifies postsecondary
students as dependent unless they are (1) 24 years of age or older,
(2) married, (3) a graduate or professional student, (4) someone with
legal dependents other than a spouse, (5) an orphan or ward of the
court, or (6) a veteran.  If any of these conditions apply, the
Department considers the student independent of his or her parents or
guardians whether or not the parents/guardians provide support or
contribute to the student's postsecondary education. 


      RACE/ETHNICITY
------------------------------------------------------ Appendix IV:1.3

Analysis of cumulative borrowing by race and ethnicity showed that
all four groups analyzed (white, not Hispanic; black, not Hispanic;
Hispanic; and Asian/Pacific Islander) showed similar increases in the
portion of students borrowing.  Average cumulative amounts borrowed
ranged from about $11,910 for Hispanics to about $16,531 for students
with Asian and Pacific Islander backgrounds.  Greater portions of
black and Hispanic groups borrowed than whites.  (See table IV.2.)



                         Table IV.2
          
            Percentage of Borrowers Was Up Among
          Most Racial/Ethnic Groups, School Years
                    1992-93 and 1995-96\

                Percentage of
            graduating seniors who
            borrowed in 1 or more       Average total
                    years             principal borrowed
            ----------------------  ----------------------
Racial/
ethnic
group\a        1992-93     1995-96     1992-93     1995-96
----------  ----------  ----------  ----------  ----------
White,            44.6        58.6     $10,314     $13,076
 non-
 Hispanic
Black,            61.0        75.7       9,489      14,246
 non-
 Hispanic
Hispanic          56.5        72.8       8,146      11,910
Asian/            38.1        53.3       9,613      16,531
 Pacific
 Islander
American          64.3          \b       8,543          \b
 Indian/
 Alaskan
 Native
----------------------------------------------------------
\a These racial and ethnic categories are mutually exclusive.  For
example, in this table, "white" refers to persons who identified
themselves as white and not Hispanic. 

\b Information is only available for 1992-93.  Information was
collected in 1995-96 that included American Indians/Alaskan Natives,
but the data were insufficient for analysis. 


   VARIATION IN WORK PATTERNS
-------------------------------------------------------- Appendix IV:2


      CLASS LEVEL
------------------------------------------------------ Appendix IV:2.1

Increases in the percentage of students working were reflected across
all undergraduate years (see table IV.3).  As in 1992-93, juniors and
seniors enrolled full time in 1995-96 were more apt to work while
enrolled, but on average worked slightly fewer hours than freshmen or
sophomores. 



                         Table IV.3
          
          Undergraduates at All Levels More Likely
          to Work, With Number of Hours Remaining
           Steady, School Years 1992-93 and 1995-
                             96

                      Percentage of
                        full-time
                      undergraduates    Average number of
                    working during the   hours worked per
                       school year             week
                    ------------------  ------------------
Class level          1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Freshman                  60        69        23        23
Sophomore                 63        73        22        23
Junior                    64        75        21        22
Senior                    69        78        22        21
----------------------------------------------------------

      DEPENDENCY STATUS AND
      PARENTAL INCOME
------------------------------------------------------ Appendix IV:2.2

A higher percentage of both dependent and independent students worked
in 1995-96 compared with 1992-93.  For 1995-96, the percentage was
higher for dependent students, but among those students who worked,
independent students worked about 6 hours more per week.  Among
dependent undergraduates, students in all income groups were more apt
to work while enrolled in 1995-96 than in 1992-93 (see table IV.4). 
Those whose parents were in middle-income groups were more likely to
work while enrolled.  The average number of hours worked changed
little and varied little by income group. 



                         Table IV.4
          
          A Greater Portion of Dependent Students
           at All Parental Income Levels Worked,
               With Number of Hours Remaining
          Relatively Steady, School Years 1992-93
                        and 1995-96

                      Percentage of
                        full-time
                      undergraduates    Average number of
                    working during the   hours worked per
                       school year             week
                    ------------------  ------------------
Parental income      1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
Less than $15,000         61        70        21        20
$15,000-$29,999           68        78        20        22
$30,000-$44,999           70        80        21        22
$45,000-$59,999           71        77        22        20
$60,000-$74,999           61        75        20        21
$75,000-$99,999           57        74        19        21
$100,000 or more          54        65        18        21
----------------------------------------------------------

      RACE/ETHNICITY
------------------------------------------------------ Appendix IV:2.3

The percentage of students who worked in 1995-96 was higher than the
percentage for 1992-93 across racial and ethnic groups as well (see
table IV.5).  White, black, and Hispanic students had the highest
percentages of students who worked, and black and Hispanic students
had the highest average hours worked per week. 



                         Table IV.5
          
            A Greater Portion of Undergraduates
            Across All Racial and Ethnic Groups
           Worked, With Hours Remaining About the
            Same, School Years 1992-93 and 1995-
                            96\a

                      Percentage of
                        full-time
                      undergraduates    Average number of
                    working during the   hours worked per
                       school year             week
                    ------------------  ------------------
Racial/ethnic
group\a              1992-93   1995-96   1992-93   1995-96
------------------  --------  --------  --------  --------
White, non-               65        74        22        23
 Hispanic
Black, non-               58        74        24        24
 Hispanic
Hispanic                  60        70        23        25
Asian/Pacific             48        64        18        19
 Islander
American Indian/          61        62        25        23
 Alaskan
 Native
----------------------------------------------------------
\a These racial and ethnic categories are mutually exclusive.  For
example, in this table, "white" refers to persons who identified
themselves as white and not Hispanic. 


   DATA SUPPORTING FIGURE AND
   TABLES
-------------------------------------------------------- Appendix IV:3

The following tables provide data supporting the preceding figure and
tables, along with additional information, including confidence
intervals for each estimate. 



                         Table IV.6
          
           Data Supporting Table IV.1, Percentage
          of Undergraduates Who Borrowed During 1
          or More Years, School Years 1992-23 and
                          1995-96

                           Percentage who borrowed
                    --------------------------------------
                                                Confidence
Class level                    Percent            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Freshman                          37.7          35.8 -39.5
Sophomore                         39.5          37.4 -41.6
Junior                          48.7\a          46.7 -50.7
Senior                          49.3\a          47.7 -51.0

1995-96
----------------------------------------------------------
Freshman                        40.9\b          38.6 -43.1
Sophomore                     48.7\a,b          45.0 -52.3
Junior                        60.7\a,b          57.4 -64.0
Senior                        60.8\a,b          58.0 -63.7
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for freshmen. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table IV.7
          
            Data Supporting Table IV.1, Average
               Cumulative Amount Borrowed by
          Undergraduates Who Borrowed During 1 or
            More Years, School Years 1992-93 and
                          1995-96

                       Average total principal borrowed
                          (constant 1995-96 dollars)
                    --------------------------------------
                                                Confidence
Class level                     Amount            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Freshman                        $4,167      $3,993 -$4,340
Sophomore                      5,428\b        5,129 -5,727
Junior                         7,710\a        7,368 -8,051
Senior                         9,187\a        8,896 -9,479

1995-96
----------------------------------------------------------
Freshman                         4,296        4,033 -4,560
Sophomore                    6,318\a,b        5,850 -6,785
Junior                       9,487\a,b       8,931 -10,044
Senior                      12,063\a,b      11,454 -12,673
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for freshmen. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table IV.8
          
          Data Supporting Figure IV.1, Percentage
            of Undergraduates Who Borrowed Among
          Dependent Graduating Seniors, by Amount
           of Parental Income, School Years 1992-
                       93 and 1995-96

                     Percentage of dependent students who
                                  borrowed\a
                    --------------------------------------
Parental income
(constant 1995-96                               Confidence
dollars)                       Percent            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Less than $15,000                 69.2          62.2 -76.2
$15,000 -$29,999                  66.3          61.3 -71.2
$30,000 -$44,999                  64.6          60.0 -69.1
$45,000 -$59,999                42.8\b          38.4 -47.1
$60,000 -$74,999                28.7\b          25.4 -31.9
$75,000 -$99,999                29.4\b          25.8 -32.9
$100,000 or more                16.3\b          13.5 -19.1

1995-96
----------------------------------------------------------
Less than $15,000                 73.3          52.2 -94.4
$15,000 -$29,999                  76.3          62.8 -89.9
$30,000 -$44,999                76.3\c          66.5 -86.1
$45,000 -$59,999                  50.6          38.2 -62.9
$60,000 -$74,999              43.3\b,c          29.7 -56.9
$75,000 -$99,999                34.9\b          22.9 -46.8
$100,000 or more              32.6\b,c          22.7 -42.4
----------------------------------------------------------
\a Financial aid calculations are based on prior-year income; thus,
parental income information for 1991 and 1994 was used for school
years 1992-93 and 1995-96, respectively. 

\b This estimated value differs to a statistically significant extent
from the corresponding estimated value for students whose parents'
income was less than $15,000. 

\c This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                         Table IV.9
          
            Data Supporting Figure IV.1, Average
          Cumulative Amount Borrowed by Dependent
              Graduating Seniors, by Amount of
           Parental Income, School Years 1992-93
                        and 1995-96

                      Amount borrowed (constant 1995-96
                                  dollars)\a
                    --------------------------------------
Parental income
(constant 1995-96                               Confidence
dollars)                        Amount            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Less than $15,000               $9,905      $8,758 -11,052
$15,000 -$29,999                11,086       9,997 -12,176
$30,000 -$44,999                10,565       9,878 -11,252
$45,000 -$59,999                 9,658       8,930 -10,385
$60,000 -$74,999                 9,212        8,453 -9,972
$75,000 -$99,999                10,453       9,453 -11,454
$100,000 or more                11,174       9,328 -13,020

1995-96
----------------------------------------------------------
Less than $15,000               11,540       8,647 -14,432
$15,000 -$29,999              14,529\b      12,409 -16,649
$30,000 -$44,999              14,694\b      12,339 -17,050
$45,000 -$59,999              13,949\b      11,835 -16,064
$60,000 -$74,999              14,348\b      11,357 -17,340
$75,000 -$99,999              15,183\b      10,286 -20,080
$100,000 or more                10,682       7,731 -13,632
----------------------------------------------------------
\a None of the estimated values differ to a statistically significant
extent from the corresponding estimated value for students whose
parents' income was less than $15,000. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 



                        Table IV.10
          
                Data Supporting Table IV.2,
            Undergraduate Borrowing, by Racial/
           Ethnic Group, Percentage of Graduating
          Seniors Who Borrowed in 1 or More Years,
              School Years 1992-93 and 1995-96

                     Percentage of graduating seniors who
                                   borrowed
                    --------------------------------------
Racial/ethnic                                   Confidence
group                          Percent            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
White, non-                       44.6          42.7 -46.5
 Hispanic
Black, non-                     61.0\a          54.8 -67.2
 Hispanic
Hispanic                        56.5\a          50.5 -62.5
Asian/Pacific                   38.1\a          32.9 -43.4
 Islander
American Indian/                64.3\a          47.6 -80.9
 Alaskan Native

1995-96
----------------------------------------------------------
White, non-                       58.6          54.2 -63.0
 Hispanic
Black, non-                   75.7\a,b          64.7 -86.7
 Hispanic
Hispanic                      72.8\a,b          61.5 -84.2
Asian/Pacific                     53.3          38.7 -67.9
 Islander
American Indian/                    \c                  \c
 Alaskan Native
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for white students. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 

\c Information is only available for 1992-93.  Information was
collected in 1995-96 that included American Indians/Alaskan Natives,
but the data were insufficient for analysis. 



                        Table IV.11
          
                Data Supporting Table IV.2,
             Undergraduate Borrowing by Racial/
          Ethnic Group, Average Cumulative Amount
             Borrowed by Graduating Seniors Who
            Borrowed in 1 or More Years, School
                 Years 1992-93 and 1995-96

                       Average total principal borrowed
                          (constant 1995-96 dollars)
                    --------------------------------------
Racial/ethnic                                   Confidence
group                           Amount            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
White, non-                    $10,314      $9,168 -$9,859
 Hispanic
Black, non-                      9,489        7,991 -9,513
 Hispanic
Hispanic                       8,146\a        6,550 -8,477
Asian/Pacific                    9,613       7,345 -10,387
 Islander
American Indian/                 8,543       4,836 -10,923
 Alaskan Native

1995-96
----------------------------------------------------------
White, non-                   13,076\b      12,056 -14,096
 Hispanic
Black, non-                   14,246\b      10,313 -18,179
 Hispanic
Hispanic                      11,910\b       9,769 -14,051
Asian/Pacific                 16,531\b      12,691 -20,372
 Islander
American Indian/                    \c                  \c
 Alaskan Native
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for white students. 

\b This estimated value differs to a statistically significant extent
from the estimated value for 1992-93. 

\c Information was only available for 1992-93.  Information was
collected in 1995-96 that included American Indians/Alaskan Natives,
but the data were insufficient for analysis. 



                        Table IV.12
          
           Data Supporting Table IV.3, Percentage
           of Full-Time, Full-Year Undergraduates
          Who Worked While Enrolled, School Years
                    1992-93 and 1995-96

                     Percentage who worked while enrolled
                    --------------------------------------
                                                Confidence
Class level                    Percent            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Freshman                          59.5          57.5 -61.5
Sophomore                       63.3\a          61.2 -65.5
Junior                          64.4\a          62.2 -66.6
Senior                          69.1\a          67.6 -70.6

1995-96
----------------------------------------------------------
Freshman                        68.9\b          66.6 -71.3
Sophomore                       73.0\b          69.0 -77.1
Junior                        75.3\a,b          72.2 -78.5
Senior                       77.6\a,\b          74.1 -81.0
----------------------------------------------------------
\a This estimate differed to a statistically significant extent from
the estimate for freshmen. 

\b This estimate differed to a statistically significant extent from
the estimate for 1992-93. 



                        Table IV.13
          
            Data Supporting Table IV.3, Average
          Number of Hours Worked per Week by Full-
            Time, Full-Year Students Who Worked
          While Enrolled, School Years 1992-93 and
                          1995-96

                      Estimated average hours worked per
                             week while enrolled
                    --------------------------------------
                                                Confidence
Class level                      Hours            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Freshman                          22.7          21.9 -23.4
Sophomore                         22.1          21.2 -23.0
Junior                          20.7\a          20.1 -21.3
Senior                            22.4          21.7 -23.0

1995-96
----------------------------------------------------------
Freshman                          23.1          22.4 -23.7
Sophomore                         23.4          21.8 -25.0
Junior                       21.7\a,\b          20.7 -22.6
Senior                          21.1\a          20.1 -22.0
----------------------------------------------------------
\a This estimate differed to a statistically significant extent from
the estimate for freshmen. 

\b This estimate differed to a statistically significant extent from
the estimate for 1992-93. 



                        Table IV.14
          
           Data Supporting Table IV.4, Percentage
             of Full-Time, Full-Year Dependent
              Undergraduates Who Worked While
            Enrolled, by Parental Income Level,
              School Years 1992-93 and 1995-96

                     Estimated percentage of students who
                            worked while enrolled
                    --------------------------------------
1994 parental
income (constant                                Confidence
1995-96 dollars)               Percent            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Less than $15,000                 61.5          56.7 -66.2
$15,000 -$29,999                68.2\a          65.0 -71.4
$30,000 -$44,999                70.2\a          67.5 -72.8
$45,000 -$59,999                70.6\a          67.8 -73.3
$60,000 -$74,999                  61.2          58.6 -63.9
75,000 -$99,999                   56.5          52.9 -60.1
$100,000 or more                53.8\a          50.5 -57.1

1995-96
----------------------------------------------------------
Less than $15,000               70.3\b          63.7 -76.9
$15,000 -$29,999                77.7\b          73.6 -81.8
$30,000 -$44,999              80.5\a,b          77.3 -83.7
$45,000 -$59,999                76.9\b          73.3 -80.6
$60,000 -$74,999                74.9\b          70.5 -79.3
$75,000 -$99,999                74.1\b          70.3 -77.9
$100,000 or more                65.3\b          60.6 -70.1
----------------------------------------------------------
\a This estimate differed to a statistically significant extent from
the estimate for dependent students with parental income less than
$15,000. 

\b This estimate differed to a statistically significant extent from
the estimate for 1992-93. 



                        Table IV.15
          
            Data Supporting Table IV.4, Average
             Hours Worked per Week by Dependent
              Undergraduates Who Worked While
            Enrolled, by Parental Income Level,
              School Years 1992-93 and 1995-96

                      Estimated average hours worked per
                                     week
                    --------------------------------------
1994 parental
income (constant                                Confidence
1995-96 dollars)                 Hours            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
Less than $15,000                 21.1          19.7 -22.5
$15,000 -$29,999                  20.2          19.1 -21.2
$30,000 -$44,999                  20.9          20.0 -21.8
$45,000 -$59,999                  21.8          20.9 -22.7
$60,000 -$74,999                  20.5          19.7 -21.2
$75,000 -$99,999                  19.1          17.9 -20.3
$100,000 or more                18.3\a          17.2 -19.3

1995-96
----------------------------------------------------------
Less than $15,000                 20.0          18.6 -21.3
$15,000 -$29,999                  21.8          20.1 -23.4
$30,000 -$44,999                  22.0          20.5 -23.5
$45,000 -$59,999                  20.0          19.1 -20.9
$60,000 -$74,999                  21.3          20.1 -22.6
$75,000 -$99,999                  21.5          19.3 -23.6
$100,000 or more                21.2\b          19.5 -22.8
----------------------------------------------------------
\a This estimate differed to a statistically significant extent from
the estimate for dependent students with parental income less than
$15,000. 

\b This estimate differed to a statistically significant extent from
the estimate for 1992-93. 



                        Table IV.16
          
           Data Supporting Table IV.5, Percentage
           of Full-Year Undergraduates Who Worked
          While Enrolled, by Racial/Ethnic Group,
              School Years 1992-93 and 1995-96

                    Estimated percentage who worked while
                                   enrolled
                    --------------------------------------
Racial/ethnic                                   Confidence
group                          Percent            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
White, non-                       65.3          63.9 -66.7
 Hispanic
Black, non-                     58.1\a          54.0 -62.3
 Hispanic
Hispanic                        60.3\a          55.6 -64.9
Asian/Pacific                   47.8\a          43.8 -51.8
 Islander
American Indian/                  61.4          51.7 -71.0
 Alaskan Native

1995-96
----------------------------------------------------------
White, non-                     73.7\b          71.9 -75.5
 Hispanic
Black, non-                     74.4\b          70.5 -78.4
 Hispanic
Hispanic                        70.3\b          62.2 -78.4
Asian/Pacific                 64.1\a,b          57.6 -70.6
 Islander
American Indian/                  62.0          47.8 -76.2
 Alaskan Native
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for white, non-Hispanic. 

\b This estimate differed to a statistically significant extent from
the corresponding estimate for 1992-93. 



                        Table IV.17
          
            Data Supporting Table IV.5, Average
          Number of Hours Worked per Week by Full-
             Time, Full-Year Undergraduates Who
          Worked While Enrolled, by Racial/Ethnic
          Group, School Years 1992-93 and 1995-96

                      Estimated average hours worked per
                                     week
                    --------------------------------------
Racial/ethnic                                   Confidence
group                            Hours            interval
------------------  ------------------  ------------------
1992-93
----------------------------------------------------------
White, non-                       22.1          21.6 -22.6
 Hispanic
Black, non-                     23.7\a          22.5 -24.9
 Hispanic
Hispanic                        23.0\a          21.6 -24.5
Asian/Pacific                   18.0\a          16.5 -19.6
 Islander
American Indian/                  25.2          20.5 -29.8
 Alaskan Native

1995-96\b
----------------------------------------------------------
White, non-                       22.6          21.9 -23.2
 Hispanic
Black, non-                     24.5\a          23.0 -25.9
 Hispanic
Hispanic                          24.8          22.4 -27.2
Asian/Pacific                   18.7\a          16.7 -20.7
 Islander
American Indian/                  22.8          20.3 -25.4
 Alaskan Native
----------------------------------------------------------
\a This estimated value differs to a statistically significant extent
from the corresponding estimated value for white, non-Hispanic. 

\b None of the estimates for 1995-96 differed to a statistically
significant extent from the corresponding estimate for 1992-93. 


*** End of document. ***