oversight

Student Loans: Default Rates Need To Be Computed More Appropriately

Published by the Government Accountability Office on 1999-07-28.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                  United States General Accounting Office

GAO               Report to Congressional Requesters




July 1999
                  STUDENT LOANS
                  Default Rates Need to
                  Be Computed More
                  Appropriately




GAO/HEHS-99-135
      United States
GAO   General Accounting Office
      Washington, D.C. 20548

      Health, Education, and
      Human Services Division

      B-282065

      July 28, 1999

      The Honorable William F. Goodling
      Chairman, Committee on Education
        and the Workforce
      House of Representatives

      The Honorable John L. Mica
      Chairman, Subcommittee on Criminal
        Justice, Drug Policy, and Human Resources
      Committee on Government Reform
      House of Representatives

      Two major federal student loan programs, the Federal Family Education
      Loan Program (FFELP) and the William D. Ford Federal Direct Loan
      Program (FDLP), provide funding that is vital to helping students meet their
      postsecondary education costs. FFELP, formerly known as the guaranteed
      student loan program, provides loans through private lenders such as
      banks. These loans are insured against default by state or nonprofit
      guaranty agencies, which are later reimbursed by the Department of
      Education. FDLP, often referred to as the direct loan program, provides
      loans from the federal government through students’ schools. The first
      FDLP loans were made in the fourth quarter of fiscal year 1994. The
      Department uses student borrowers’ experiences with loans from FFELP
      and FDLP for determining a school’s default rate.

      In fiscal year 1998, these programs provided student borrowers with
      nearly 8.4 million loans totaling more than $30 billion. However, when
      borrowers fail to meet their financial obligations by not repaying their
      federal student loans, it is the government that ultimately must pay for this
      failure. For example, the Department of Education paid $2.1 billion in
      default claims from these programs in fiscal year 1998. An accurate
      measure of student loan defaults at colleges, universities, and vocational
      schools is an important means for monitoring the extent of financial risk
      to the Department from its student loan programs. To protect the
      government from the costs associated with extremely high rates of default,
      the Department now excludes schools from program participation if their
      default rate is 25 percent or more for 3 consecutive years.1




      1
       A school’s default rate is the rate at which the school’s FFELP and FDLP student borrowers have
      defaulted on their loans.



      Page 1                               GAO/HEHS-99-135 Computing Default Rates Appropriately
                                     B-282065




                                     You asked us to examine one aspect of the way in which the Department
                                     of Education calculates this default rate. The issue involves borrowers
                                     who have temporary approval through their lenders or loan services
                                     through “deferment” or “forbearance” not to make payments on their
                                     loans.2 In the Department’s calculation of a school’s default rate, these
                                     borrowers are not counted as defaulters, but they do count as a part of the
                                     total number of borrowers. As shown below, the number of borrowers in
                                     default is divided by a number larger than the total number of borrowers
                                     who are actually repaying their loans (see fig. 1). As a result, the default
                                     rate is understated.


Figure 1: Current Method for
Calculating Schools’ Default Rates
                                          Number of borrowers who began repaying during the first
                                          fiscal year of a 2-year cohort a period and defaulted by the
                                         end of the second fiscal year (borrowers granted deferment
                                                   or forbearance status are not included).




                                                                       divided by




                                           Number of borrowers who began repaying during the first
                                            fiscal year of a 2-year cohort period (borrowers granted
                                                 deferment or forbearance status are included).



                                     a
                                      The Department gathers data required for calculating default rates by cohort. Covering a 2-year
                                     period, a cohort constitutes a group of student borrowers who began repaying their loans during
                                     a given fiscal year and also identifies those in the group who defaulted before the end of the next
                                     fiscal year. Although it covers a 2-year period, a cohort is identified by its first fiscal year. For
                                     example, borrowers in the 1996 cohort began repaying during the 1996 fiscal year.


                                     As agreed with your offices, we focused our work on answering three
                                     specific questions with regard to these calculations:



                                     2
                                      Deferment is postponement of payments for such reasons as continued study, inability to find work,
                                     or economic hardship. Forbearance is permission to temporarily suspend payments, make smaller
                                     payments, or extend the time for making payments because of poor health or other acceptable
                                     reasons.



                                     Page 2                                GAO/HEHS-99-135 Computing Default Rates Appropriately
                       B-282065




                   •   Over the past several years, has there been an increase in the number of
                       borrowers who entered repayment but subsequently received deferments
                       or forbearances?
                   •   What would have been the effect on the most recent default rates if
                       borrowers whose loans were in deferment or forbearance had been
                       excluded from the default rate calculation?
                   •   Under this alternative method of calculating the default rate, would any
                       additional schools have exceeded the 25-percent default rate threshold?

                       Appendix I describes our scope and methodology. We conducted our
                       review between August 1998 and May 1999 in accordance with generally
                       accepted government auditing standards.


                       Between 1993 and 1996, the percentage of borrowers with loans in
Results in Brief       deferment or forbearance more than doubled, from 5.2 percent of
                       borrowers who had begun repaying to 11.3 percent. This doubling was
                       consistent across the various types of schools, including 4-year and
                       less-than-4-year public and private schools as well as proprietary schools.
                       According to Department of Education officials, the increase was
                       attributable, in part, to provisions of the 1992 amendments to the Higher
                       Education Act of 1965 that eased the requirements for obtaining
                       deferments and forbearances as a way of helping minimize loan defaults.

                       Excluding borrowers with loans in deferment or forbearance entirely from
                       the calculation of the cohort default rate would have had the effect of
                       increasing the overall default rate from 9.6 percent to 10.9 percent for
                       1996, the most recent cohort year for which data are available. The
                       proportional increases would have been roughly similar for the various
                       types of schools. For example, the rate at 4-year schools would have risen
                       from 6.8 to 7.7 percent, while the rate at proprietary schools would have
                       risen from 18.3 to 20.1 percent.

                       For the 1996 cohort, excluding borrowers with loans in deferment or
                       forbearance from the calculation would have increased the number of
                       schools with rates exceeding the 25-percent threshold (for excluding
                       schools from the loan programs) by 181 schools, from 352 to 533—an
                       increase of 51 percent. Under the law, these schools would have become
                       ineligible to participate in student loan programs if their cohort default
                       rate had exceeded the threshold for 3 consecutive years. Since 1991, the
                       Department has denied participation in the programs to more than 1,000
                       schools because their default rates were too high. Most of the additional



                       Page 3                      GAO/HEHS-99-135 Computing Default Rates Appropriately
             B-282065




             schools that would have exceeded the threshold under the alternative
             calculation method were proprietary schools, but 12 were 4-year colleges
             and universities and 57 were public or private schools with degree
             programs of less than 4 years. Because the number of borrowers with
             loans in deferment or forbearance has been growing, and because the
             exclusion of these borrowers from the calculation could have a substantial
             effect on schools’ default rates, the Congress may want to consider
             requiring an alternative method for computing default rates.


             Default reduction measures were part of the default reduction initiative
Background   that the Department introduced in June 1989 in response to the rising
             default rates in federal student loan programs at that time. According to
             Department of Education officials, these measures apply to all schools
             participating in federal student loan programs. Default reduction
             measures, incorporated into statutes, regulations, and guidance, require
             schools to provide students with loan counseling; take steps to promote
             repayment among delinquent borrowers; and, for schools whose default
             rates exceed certain thresholds, implement a default management plan.
             Such actions, if properly implemented, reduce loan defaults and the
             associated federal costs to pay lenders’ default claims, as anticipated.

             The Department’s efforts to reduce historically high default rates for
             federal student loan programs have paid dividends. Schools’ overall
             default rate hit its high of 22.4 percent in fiscal year 1990 but declined to
             9.6 percent for the 1996 cohort. One tool the Department has used to bring
             greater financial accountability to the programs is the default rate
             threshold, which was authorized in a 1990 amendment to title IV of the
             Higher Education Act of 1965.3 The threshold was instituted as a safeguard
             to protect the government from the costs associated with schools whose
             students consistently had exceptionally high default rates. In 1991, under
             this legislation, the Department began to bar schools from participating in
             federal student loan programs if their cohort default rates exceeded the
             statutory threshold of 25 percent for 3 consecutive years.4

             As provided by section 435(m) of the Higher Education Act of 1965, as
             amended, the Department establishes a default rate for each school by
             creating a cohort consisting of all the school’s students who are expected
             to begin repaying their loans in a given year. The Department then
             determines how many of these students default on their loans in that year

             3
              The amendment was part of the Student Loan Default Prevention Initiative Act of 1990 (P.L. 101-508).
             4
              This threshold decreased from 35 percent to 25 percent in 1991.



             Page 4                               GAO/HEHS-99-135 Computing Default Rates Appropriately
B-282065




or by the end of the following year. For a school with 30 or more
borrowers5 beginning to repay their loans, the default rate is the
percentage that results from dividing (1) the number of students who
begin to repay in a given fiscal year and default in that year or before the
end of the next fiscal year (the numerator) by (2) the number of students
who are supposed to begin repaying in that given fiscal year (the
denominator). For example, if 100 students from a school were scheduled
to begin repaying their loans in fiscal year 1996 and 25 defaulted on their
loans by the end of fiscal year 1997, the school’s 1996 default rate would
be 25 percent.

The criterion for determining when a borrower has defaulted for the
purpose of being placed in the numerator of the cohort default calculation
varies by loan program. For FFELP, a borrower is considered to be in
default only if the guaranty agency has paid a default claim to the lender
on the borrower’s loan. The date the guaranty agency reimburses the
lender for the defaulted loan (the “claim paid” date, which is generally
after the borrower has been delinquent for over 270 days on a loan payable
in monthly installments) is the basis for determining when a borrower is
placed in the numerator of the default calculation.6 For FDLP, borrowers
are considered to be in default on the 271st day of delinquency and are to
be placed in the numerator of the default calculation at that time.

Each year, the Department assesses a school’s eligibility to continue
participating in FFELP and FDLP on the basis of the school’s default rates for
the most recent 3 consecutive years for which data are available. In fiscal
year 1999, for example, eligibility is based on default rates for fiscal years
1994, 1995, and 1996. A school remains eligible if its rate is below the
25-percent threshold in at least 1 of these years. Most schools become
ineligible if their default rate equals or exceeds the default threshold in all
3 fiscal years.7 Some of the student borrowers placed in a cohort neither
default nor make payments on their loans during the full cohort period.
These borrowers include the following:



5
 If a school has fewer than 30 borrowers entering repayment, the Department calculates a 3-year
average default rate (see app. I).
6
 Before the enactment of the Higher Education Amendments of 1998 (sec. 429 of P.L. 105-244, effective
Oct. 1, 1998), sec. 435(l) of the Higher Education Act of 1965, as amended, defined default as including
failures to repay that had existed for only (1) 180 (rather than 270) days, in the case of a loan that was
repayable in monthly installments, or (2) 240 days, in the case of a loan that was repayable in less
frequent installments.
7
 The Higher Education Act exempted historically black colleges and universities, tribally controlled
institutions, and Navajo community colleges from the threshold requirement through June 1999.



Page 5                                 GAO/HEHS-99-135 Computing Default Rates Appropriately
    B-282065




•   Some borrowers are allowed to defer payment for an additional period
    because, for example, they are pursuing an approved course of study,
    trying but unable to find full-time work, or otherwise experiencing
    economic hardship. People having trouble finding work or experiencing
    other economic hardship may defer payment for up to 3 years, if they
    borrowed for the first time on or after July 1, 1993.
•   Other borrowers may receive forbearance, which generally involves
    temporarily ceasing payment.8 Borrowers may receive forbearance if, for
    example, because of poor health or other acceptable reasons, they cannot
    make scheduled payments. In certain circumstances, forbearance may be
    administratively granted by the Secretary of Education without requiring
    documentation from the borrower. For example, administrative
    forbearance may cover a period of national military mobilization or other
    local or national emergency. In other circumstances, forbearances are
    mandatory. For example, a lender must grant forbearance when a
    borrower is serving in a medical or dental internship or residency program
    and has exhausted his or her eligibility for deferment, or when a
    borrower’s monthly loan payments are equal to or greater than 20 percent
    of total monthly income. Generally, a borrower may be granted
    forbearance for up to 1 year at a time.

    If the number of borrowers who have been provided deferments or
    forbearances becomes substantial, default rates can be affected in two
    ways. First, because these borrowers are not removed from the cohort, the
    default rate is lowered. This happens because they are counted as part of
    the total number of borrowers in the cohort who began repayment, even
    though they are not making payments on their loans. Second, because
    these borrowers are never placed in a subsequent cohort, they are never
    included in calculations of a school’s default rate, even if they default on
    their loans after the deferment or forbearance period is over.

    Department officials said they do not attempt to remove students whose
    loans are in deferment or forbearance from a cohort because it is
    Department policy to view such borrowers as a part of the repayment
    population. This policy is consistent with section 435(m) of the Higher
    Education Act, which defines the term “cohort default rate.”




    8
     Forbearance can also involve extending the payment period or making smaller payments. However,
    according to officials from five large loan servicing organizations, nearly all forbearance cases
    currently being processed involve temporarily stopping payment.



    Page 6                              GAO/HEHS-99-135 Computing Default Rates Appropriately
                                       B-282065




                                       Between cohort years 1993 and 1996, the percentage of borrowers who
Student Loans in                       were granted a deferment or forbearance for their loans more than
Deferment and                          doubled, rising from 5.2 percent to 11.3 percent—or from about 96,000
Forbearance Have                       borrowers to about 227,000. As figure 2 shows, the increase in borrowers
                                       granted deferments or forbearances was relatively consistent across
Increased for All                      various school types.
Types of Schools
Figure 2: Borrowers in Deferment or
Forbearance as a Percentage of Total   14.0   Percentage
Borrowers in Repayment, by School                                           13.0
Type, 1993 and 1996 Cohorts
                                       12.0            11.4                                                       11.3


                                       10.0
                                                                                                 9.1


                                        8.0


                                        6.0     5.7                  5.7
                                                                                                          5.2


                                        4.0                                               3.7



                                        2.0


                                         0

                                                  4-Year Public        Less Than            Proprietary     All
                                                  and Private          4-Year Public
                                                                       and Private
                                                 Type of School



                                                              1993

                                                              1996




                                       Department officials attributed this rise in deferments and forbearances, in
                                       part, to various changes in the law instituted by the 1992 amendments that
                                       eased the requirements for obtaining deferments and forbearances. These
                                       changes included the simplification of deferment by reducing the number
                                       of deferment categories from 13 to 3, the provision for mandatory and
                                       administrative forbearances, and a broadened definition of economic
                                       hardship for deferments. For example, under the provision for mandatory
                                       forbearances, guaranty agencies are no longer given discretion in deciding



                                       Page 7                                      GAO/HEHS-99-135 Computing Default Rates Appropriately
B-282065




whether to grant forbearances for certain conditions; they are required to
grant forbearances, for example, to students who demonstrate a
willingness to pay but are currently unable to do so, and to students who
have exhausted medical and dental internship deferments.

Because proprietary schools have historically had more difficulty
remaining under the default rate threshold, we examined whether their
students might have become the borrowers most likely to seek loan
deferments and forbearances as an alternative to default. However, in both
cohort years 1993 and 1996, the percentage of students with deferments or
forbearances was significantly less for proprietary schools than for other
types of schools.

Although both deferments and forbearances increased between cohort
years 1993 and 1996, the number of forbearances grew more. Overall, the
number of borrowers who obtained deferments nearly doubled, from
about 80,000 to 148,000, but the number who obtained forbearances
quintupled, from about 16,000 to 80,000. These increases were generally
similar across all types of schools.

We also attempted to determine what differences, if any, could be
discerned between borrowers in FDLP and FFELP. Comparisons were
limited, because FDLP is relatively new. No borrowers with FDLP loans were
included in the 1993 cohort, but FDLP borrowers constituted about
3.9 percent of total borrowers in repayment for the 1996 cohort. For that
year, FDLP borrowers in repayment were nearly three times more likely
than FFELP borrowers in repayment to have had a deferment or
forbearance. Although these early data indicate a potentially significant
difference between the two loan programs, it is premature to conclude
that a large difference will persist in future cohorts. Department officials
said they believed this difference between FDLP and FFELP is due, in part, to
the Department’s Direct Loan Servicing Center’s active use of deferments
and forbearances as tools to facilitate the resolution of delinquencies to
help minimize loan defaults. Additionally, the servicing center commonly
uses administrative forbearances when correcting erroneous information
generated by schools, the Department, or itself to ensure that borrowers
are not unfairly penalized.




Page 8                      GAO/HEHS-99-135 Computing Default Rates Appropriately
                                       B-282065




                                       Calculating default rates using an alternative methodology that excluded
Default Rates Rose                     borrowers with loans in deferment or forbearance resulted in higher
When Borrowers With                    default rates. Using the current methodology for calculating cohort default
Deferments and                         rates, the overall rate for all schools in the 1996 cohort was 9.6 percent.
                                       When recalculated using the alternative method, the rate increased by
Forbearances Were                      1.2 percentage points to 10.9 percent.9 The increases were proportionately
Excluded                               similar across the different types of schools (see fig. 3).


Figure 3: Student Loan Default Rates
Calculated Using Current and           25.0      Percentage
Alternative Methodologies, by School
Type, 1996 Cohort
                                                                                                    20.1
                                       20.0
                                                                                             18.3


                                                                               15.4
                                       15.0
                                                                        13.4


                                                                                                                     10.9
                                       10.0                                                                  9.6

                                                        7.7
                                                 6.8

                                           5.0




                                            0

                                                   4-Year Public          Less Than            Proprietary     All
                                                   and Private            4-Year Public
                                                                          and Private
                                                   Type of School



                                                              Current Method

                                                              Alternative Method



                                       Note: The current method includes borrowers with deferments or forbearances in its denominator;
                                       the alternative method does not.




                                       The change in the default rate shows the effect of the alternative
                                       calculation methodology, which excludes borrowers with loans in


                                       9
                                        The percentage point increase is 1.2, rather than the difference between the rounded percentages 10.9
                                       and 9.6.



                                       Page 9                                         GAO/HEHS-99-135 Computing Default Rates Appropriately
                       B-282065




                       deferment or forbearance from the denominator (all borrowers who
                       entered repayment) of the calculation. Because virtually no borrowers
                       with loans in deferment or forbearance have to make loan payments, these
                       borrowers do not, by definition, go into default and thus are excluded
                       from the numerator (borrowers in default) under both the current and the
                       alternative methods for calculating the default rate. The current
                       methodology, however, retains all borrowers with loans in deferment or
                       forbearance in the denominator, even though these borrowers are no
                       longer making payments on their loans. As the percentage of borrowers
                       with loans in deferment or forbearance increases, the difference that
                       occurs in the rates computed under the two methodologies also increases.


                       Excluding borrowers with loans in deferment or forbearance from the
Under the              default rate calculation also had the effect of increasing the number of
Recalculated Default   schools with default rates above the 25-percent threshold. Under the
Rate, More Schools     current methodology, 352 schools (out of 4,320) had default rates equal to
                       or greater than 25 percent for 1996. When we excluded borrowers with
Exceeded the           loans in deferment or forbearance from the calculation entirely, the
25-Percent Threshold   number of schools exceeding the 25-percent threshold rose by 181
                       (51 percent) to a total of 533 schools. As figure 4 shows, proprietary
                       schools represented the greatest part of this increase: 112 of the additional
                       181 schools were proprietary schools. However, the percentage increase
                       for proprietary schools was slightly less than the percentage increase for
                       the other types of schools. The additional proprietary schools represented
                       a 41-percent increase, compared with a 43-percent increase for 4-year
                       schools and a 110-percent increase for less-than-4-year schools.




                       Page 10                     GAO/HEHS-99-135 Computing Default Rates Appropriately
                                           B-282065




Figure 4: Number of Schools Whose
Default Rates Equaled or Exceeded 25       600   Number of Schools
Percent When Calculated Using
                                                                                                                     533
Current and Alternative
Methodologies, by School Type, 1996        500
Cohort
                                           400                                                     384
                                                                                                             352


                                           300                                               272



                                           200


                                                                                109
                                           100
                                                                         52
                                                          40
                                                 28

                                             0

                                                     4-Year Public         Less Than          Proprietary      All
                                                     and Private           4-Year Public
                                                                           and Private
                                                     Type of School



                                                               Current Method

                                                               Alternative Method



                                           Note: The current method includes borrowers with deferments or forbearances in its denominator;
                                           the alternative method does not.




                                           Department officials told us they did not favor changing the current
                                           method for calculating schools’ default rates because the national default
                                           rate has fallen each year since 1991, and, at the same time, more than 1,000
                                           schools have been removed from the programs because their default rates
                                           were too high. Department officials also said changing the method would
                                           create the following problems:

                                       •   The Department would have to modify its computer program at significant
                                           cost.
                                       •   Schools wanting to check the Department’s calculations would create
                                           increased workloads because the schools would request additional
                                           information from the Department, lenders, and loan servicers.




                                           Page 11                                    GAO/HEHS-99-135 Computing Default Rates Appropriately
                  B-282065




              •   The number of schools that would challenge or appeal the default rate
                  calculation would increase overwhelmingly. The Department estimated
                  that the number of challenges and appeals would rise from the current
                  level of 850 schools a year to at least 2,550 a year. The added cost of
                  handling these challenges and appeals, Department officials estimated,
                  would be more than $1 million.

                  We do not believe that the Department’s objections to changing the
                  method of calculating the cohort default rate are compelling. The reasons
                  cited are mainly administrative in nature and appear to be overstated. For
                  example, those schools that would likely have a compelling reason to
                  challenge or appeal on this basis are the ones that would move above the
                  default rate threshold specifically because students in deferment or
                  forbearance were excluded from the calculation. It is important to note
                  that schools do not lose their eligibility to participate in FFELP and FDLP on
                  the basis of 1 year’s cohort default rate. A school becomes ineligible if its
                  default rate remains at or above 25 percent for each of the 3 most recent
                  years. Our calculations for the 1996 cohort showed that 181 schools would
                  fall into this category in 1 year, and it is unknown how many schools might
                  exceed the default rate threshold for 3 consecutive years. Whatever the
                  actual number, it would be far less than the additional 1,700 challenges or
                  appeals the Department estimates would occur as a result of changing the
                  method of calculating the default rate. In addition, these 181 schools had
                  over 12,000 borrowers who had defaulted on their student loans. Assuming
                  an average loan size of $3,500, the cost to the government to pay lenders’
                  claims for these defaulted loans could exceed $40 million. Even if only a
                  modest number of these schools were disqualified from participating in
                  federal student loan programs because their default rates exceeded the
                  threshold for 3 consecutive years, the potential savings to the Department
                  in reduced default claims could well exceed the costs of administering the
                  change in the default rate calculation.


                  By definition, a borrower with a loan in deferment or forbearance is
Conclusions       generally not required to make loan payments and, therefore, has no
                  exposure to default during the time the deferment or forbearance is in
                  place. Currently, these borrowers are included in the denominator when
                  computing schools’ cohort default rates as if they were still making
                  payments. Excluding such borrowers from the denominator of the default
                  rate is appropriate because, using the current calculation methodology,
                  should these borrowers default later, the defaults will not be factored into
                  subsequent cohort calculations. Reliance on the current calculation



                  Page 12                     GAO/HEHS-99-135 Computing Default Rates Appropriately
                       B-282065




                       method could allow schools to remain under the 25-percent threshold and
                       to maintain their student loan program eligibility even if they would be
                       ineligible if the default rate were calculated using the alternative
                       methodology. Further, if the trend of an increasing number of borrowers’
                       obtaining deferments and forbearances continues beyond the 1996 cohort,
                       the impact of shielding high-default schools from exceeding the threshold
                       will be greater in the future. This, in turn, could result in these shielded
                       schools’ continued participation in federal student loan programs and
                       increased costs to the federal government as it continued to pay lenders’
                       claims for loans that defaulted.

                       Although the Department believes that administering a change in the
                       default rate calculation method would result in substantially increased
                       costs, we believe the Department’s estimates of increased costs are too
                       high. Additionally, there could be net savings if, as we believe is likely,
                       reduced default claims exceeded the anticipated increases in
                       administration expenses. Consequently, we believe that borrowers with
                       loans in deferment or forbearance should not be viewed as borrowers in
                       repayment when the Department calculates schools’ cohort default rates
                       but should instead be removed from the cohort before calculating default
                       rates. Also, these borrowers should be added to a subsequent default
                       cohort in the year in which their deferments or forbearances end and they
                       have begun repaying their loans.


                       The Congress may wish to consider amending section 435(m) of the
Matters for            Higher Education Act of 1965, as amended, to entirely exclude from the
Consideration by the   annual calculation of school default rates borrowers who are not in
Congress               repayment by the end of a default cohort period because they have loans
                       in deferment or forbearance. Additionally, the Congress may wish to
                       require the Secretary to develop and implement procedures to ensure that
                       borrowers excluded from a cohort’s default rate calculation because of an
                       authorized deferment or forbearance are included in a future cohort after
                       they have resumed making payments on their loans.


                       The Department of Education provided comments on a draft of this report.
Agency Comments        The Department agreed that default prevention measures should be
                       revisited and stated that our report provided a helpful discussion of ways
                       to measure student loan defaults. It also said that its Office of Inspector
                       General is currently exploring an alternative method for calculating the
                       cohort default rate, and that the Department plans to review both our



                       Page 13                     GAO/HEHS-99-135 Computing Default Rates Appropriately
B-282065




suggestions and those of the Inspector General to determine if the use of
the cohort default rate as a default management tool can be improved. In
this regard, the Department offered reasons for maintaining the current
calculation in the interim, issues to consider if a change in calculation
method is implemented, and possible new strategies to reduce default
costs. The Department’s comments appear in appendix II.


Copies of this report will be provided to the Honorable James M. Jeffords,
Chairman, Senate Committee on Health, Education, Labor, and Pensions;
the Honorable Richard W. Riley, Secretary of Education; and other
interested parties. We will also make copies available upon request.

This report was prepared under the direction of Cynthia M. Fagnoni,
Director, Education, Workforce, and Income Security Issues, who may be
reached at (202) 512-7215 if you or your staff have any questions. Other
staff who made key contributions to this report include Joseph J. Eglin,
Jr.; Daniel C. Jacobsen; and Edward H. Tuchman.




Richard L. Hembra
Assistant Comptroller General




Page 14                    GAO/HEHS-99-135 Computing Default Rates Appropriately
Page 15   GAO/HEHS-99-135 Computing Default Rates Appropriately
Contents



Letter                                                                                           1


Appendix I                                                                                      18

Scope and
Methodology
Appendix II                                                                                     21

Comments From the
Department of
Education
Figures             Figure 1: Current Method for Calculating Schools’ Default Rates              2
                    Figure 2: Borrowers in Deferment or Forbearance as a                         7
                      Percentage of Total Borrowers in Repayment, by School Type,
                      1993 and 1996 Cohorts
                    Figure 3: Student Loan Default Rates Calculated Using Current                9
                      and Alternative Methodologies, by School Type, 1996 Cohort
                    Figure 4: Number of Schools Whose Default Rates Equaled or                  11
                      Exceeded 25 Percent When Calculated Using Current and
                      Alternative Methodologies, by School Type, 1996 Cohort




                    Abbreviations

                    FDLP      William D. Ford Federal Direct Loan Program
                    FFELP     Federal Family Education Loan Program
                    NSLDS     National Student Loan Data System


                    Page 16                   GAO/HEHS-99-135 Computing Default Rates Appropriately
Page 17   GAO/HEHS-99-135 Computing Default Rates Appropriately
Appendix I

Scope and Methodology


             We obtained most of the data used to address our report objectives from
             the Department of Education’s National Student Loan Data System (NSLDS)
             through the Default Management Division. Specifically, we obtained Loan
             Record Detail Report information (formerly referred to as backup data)
             for all schools with 30 or more borrowers in repayment for both the 1996
             and 1993 default rate cohorts. We used the 1996 cohort because it contains
             the most recently published default data available on the extent to which
             borrowers in repayment were using student loan deferments and
             forbearances. The 1993 cohort was selected because it permits a 3-year
             time difference for comparison with the 1996 cohort and was readily
             available for our use.

             The formula the Department uses for calculating a school’s cohort default
             rate depends on the number of student borrowers from that school
             entering repayment in a given fiscal year. For a school with 30 or more
             borrowers entering repayment, the cohort default rate is the percentage
             that results from dividing the number of students who entered repayment
             in a given fiscal year and defaulted before the end of the next fiscal year
             (the numerator) by the number of students who entered repayment in that
             given fiscal year (the denominator). If a school has fewer than 30
             borrowers entering repayment, the Department calculates an average
             cohort default rate. This average rate is the percentage that results from
             dividing the number of students who entered repayment in the 3 most
             recent fiscal years and defaulted before the end of the fiscal year
             immediately following the fiscal year in which the loan entered repayment
             (the numerator) by the number of students who entered repayment in the
             3 most recent fiscal years (the denominator). Because of the complexities
             involved in re-creating the Department’s average cohort default rate
             calculations for schools with fewer than 30 borrowers in repayment, we
             limited our analyses to schools with 30 or more borrowers entering
             repayment in both the 1993 and 1996 cohorts. We estimated that over
             99 percent of the borrowers identified by the Department as being in
             repayment for these cohort years were included in our analyses.

             The loan record report contained school identification numbers for each
             borrower but did not contain data on school level (that is, whether it was a
             4-year, 2-year, or less-than-2-year school) or control (that is, whether it
             was a public, private nonprofit, proprietary, or foreign school). By using
             school identification numbers, we were able to obtain school level and
             control data from the Department’s “FY 1994-1996 Cohort Default Rates”
             report, its Integrated Postsecondary Education Data System, and its




             Page 18                     GAO/HEHS-99-135 Computing Default Rates Appropriately
Appendix I
Scope and Methodology




Postsecondary Education Participant System database. Foreign schools
were excluded from our analysis.

Our analysis of the number of schools that exceeded the 25-percent
statutory threshold was based on 1996 cohort data only. We did not
estimate the number of schools that could become ineligible to participate
in federal loan programs under the alternative methodology because such
a determination would have to be based on rates for 3 consecutive cohort
years. We did not have loan records for the needed additional 2 years
immediately before or after 1996 (neither 1994 and 1995 nor 1997 and
1998).

A methodological concern regarding forbearance status borrowers was
addressed before we analyzed the numbers of borrowers for their
deferment or forbearance status or calculated default rates that excluded
deferred or forborne status repayers. The concern involved our inability to
determine from both the 1993 and 1996 cohort loan records the type of
forbearances that were represented by the borrowers in repayment.
Depending upon the type of forbearance a borrower chooses, loan
payments will temporarily cease or will continue in some form. To the
extent that forbearances are the type that require no payments during the
forbearance period, they prohibit a borrower from defaulting.
Forbearances that do not require payments were the type that we intended
to exclude from the denominator in making our alternative default rate
calculations. Consequently, we devised a means for estimating the extent
to which forbearances found within the 1993 and 1996 cohorts were the
type that did not require loan payment during the forbearance period.

We contacted officials from five large student loan servicing organizations
that, combined, according to the executive director of the Student Loan
Servicing Alliance, collect and service about 70 percent of FFELP loans and
100 percent of FDLP loans. We asked these officials, given their companies’
experiences as servicers of federal student loans, what percentage of
borrowers in forbearance make no payments during the forbearance
period. Responses from all five loan servicing organizations were
essentially the same; nearly all forbearances, over 90 percent in one case
and nearly 100 percent in four cases, were estimated to be the type that
require no loan payments during the forbearance period. On the basis of
these officials’ experience, we determined that our methodology for
analyzing forbearances could reasonably presume that all borrowers with
forbearance status in the Department’s 1993 and 1996 cohorts did not have
to make loan payments during the forbearance period.



Page 19                    GAO/HEHS-99-135 Computing Default Rates Appropriately
Appendix I
Scope and Methodology




In calculating an alternative default rate, we excluded from the
denominator all borrowers in repayment who had a loan in deferment or
forbearance. The scope of our work did not include a consideration of
when these deferments or forbearances might end, thereby causing the
loans to reenter repayment.

In addition to contacting loan servicers, we contacted various Department
officials and an NSLDS computer specialist contracted by the Department,
and reviewed laws, regulations, and Department procedures associated
with the management and production of cohort default rates for
postsecondary schools. Relying on Department procedures for ensuring
data integrity, we did not validate the information and data obtained and
used in our analyses. Schools, guaranty agencies, lenders, and FDLP
services are among the primary organizations that provide data to NSLDS
for use in calculating cohort default rates. These same organizations are
afforded the opportunity to review draft cohort default rates before they
are officially released to the public and to take action through adjustment
or appeal requests to correct loan records in the NSLDS system that they
believe are incorrect.




Page 20                    GAO/HEHS-99-135 Computing Default Rates Appropriately
Appendix II

Comments From the Department of
Education




              Page 21   GAO/HEHS-99-135 Computing Default Rates Appropriately
           Appendix II
           Comments From the Department of
           Education




(104940)   Page 22                      GAO/HEHS-99-135 Computing Default Rates Appropriately
Ordering Information

The first copy of each GAO report and testimony is free.
Additional copies are $2 each. Orders should be sent to the
following address, accompanied by a check or money order
made out to the Superintendent of Documents, when
necessary. VISA and MasterCard credit cards are accepted, also.
Orders for 100 or more copies to be mailed to a single address
are discounted 25 percent.

Orders by mail:

U.S. General Accounting Office
P.O. Box 37050
Washington, DC 20013

or visit:

Room 1100
700 4th St. NW (corner of 4th and G Sts. NW)
U.S. General Accounting Office
Washington, DC

Orders may also be placed by calling (202) 512-6000
or by using fax number (202) 512-6061, or TDD (202) 512-2537.

Each day, GAO issues a list of newly available reports and
testimony. To receive facsimile copies of the daily list or any
list from the past 30 days, please call (202) 512-6000 using a
touchtone phone. A recorded menu will provide information on
how to obtain these lists.

For information on how to access GAO reports on the INTERNET,
send an e-mail message with "info" in the body to:

info@www.gao.gov

or visit GAO’s World Wide Web Home Page at:

http://www.gao.gov




PRINTED ON    RECYCLED PAPER
United States                       Bulk Rate
General Accounting Office      Postage & Fees Paid
Washington, D.C. 20548-0001           GAO
                                 Permit No. G100
Official Business
Penalty for Private Use $300

Address Correction Requested