oversight

Stafford Student Loans: Millions of Dollars in Loans Awarded to Ineligible Borrowers

Published by the Government Accountability Office on 1990-12-12.

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

                                                                                                                          ’:
                - __“~
_.--___-^_l. ..._...          ^. ..-..._ .- . . ‘I’
                     _.___...__....._            - ttil~vi       Sl;tWs
                                                                  --_.---- (;twer;tl
                                                   ..-.I.-.._.- _..__                  Aowtttrt.irtg   Of’f’iw
                                                                                                                           1
GAO                                                12qxwt, to the Cldmm,           Ikrrr~;~nent
                                                   Slll)(lor-tlrnit,t~~eon Investigations,
                                                   Cornmittw on Governrne~tal Affairs,
                                                   I J.S. Sc!nat;c
...__
    .__..      -.-_-_....
       *_.-..._..    -I- __-....._
                             -..___l_l_“..“_”
                                    -____-~~
I)ecY~IIl!)~‘I’I’.)00 .
                                                   STAFFORD STUDENT
                                                   LOANS
                                                   Millions of Dollars in
                                                   Loans Awarded to
                                                   Ineligible Borrowers


                                                                                                                 142994




                                     RESTRICTED--      Not to be released outside the
                                     General Accounting Oftke unless speciacally
                                     approved by the Office of Congressional
                                     Relations.


GAO,‘IMrlW:-91-7
*_ ““”.l”l._-.... ...--“.-.__
                           -....-----_l--__--~-
                   United States
GAO                General Accounting Office
                   Washington,      D.C. 20648

                   Information      Management       and
                   Technology      Division

                   B-240703

                   December 12,199O

                   The Honorable Sam Nunn
                   Chairman, Permanent Subcommittee
                     on Investigations
                   Committee on Governmental Affairs
                   United States Senate

                   Dear Mr. Chairman:

                   In response to your request, we are reporting to you on the Department
                   of Education’s data base on Stafford Student Loans (formerly called
                   Guaranteed Student Loans). This report evaluates the data base and dis-
                   cusses how the Department can use the information in this data base to
                   prevent loan abuses.

                   The Stafford Loan Program1 provides loans to vocational and college
                   students. This program has been identified by the Comptroller General
                   as an area with the potential for mismanagement, fraud, and abuse due
                   to internal control weaknesses. It is estimated that in fiscal year 1990
                   student loan defaults will cost about $2.4 billion, or more than 64 per-
                   cent of total program costs. In response to this problem, we looked at the
                   data base to determine whether students who have defaulted on Staf-
                   ford loans were obtaining new loans and whether students were
                   obtaining loans that exceeded legal loan limits.

                   GAO  reviews are underway on other aspects of the Department’s student
                   loan programs. Details on our objectives, scope, and methodology are
                   included in appendix I.


                   The Department maintains a data base on Stafford Student Loans. How-
Results in Brief   ever, it does not check the information in the data base for completeness
                   or accuracy. Most significantly, the Department does not require that
                   the data base be used to (1) identify student loan defaulters who are
                   trying to obtain new loans or (2) determine whether borrowers have
                   exceeded legal loan limits.

                   Our analysis of the Department’s data base showed that loan defaulters
                   may have obtained about $109 million in new loans, and students have

                   ‘This program includes three kinds of loans: (1) Stafford Loans, (2) Parent Loans for Undergraduate
                   Students (PLUS), and (3) Supplemental Loans for Students (SW). This report discusses Stafford loans
                   only, the largest of the three, in terms of dollars and numbers of participating students.



                   Page 1               GAO/IMTEG91-7      Staff’ord   Student   Loam Awarded   to Ineligible   Borruwera
             B-240708                                                                  ,




             received millions of dollars in loans over the legal loan limits. We were
             unable to determine the exact amount of such loans because the Depart-
             ment’s data are so unreliable. However, it is certain that because of
             inadequate internal controls, significant abuses exist. The abuses are
             costing the federal government millions of dollars annually in federal
             interest subsidies and additional defaulted loans.

             In commenting on a draft of this report, the Department said that the
             data base was not designed to be used to prevent abuses. In fact, how-
             ever, the Department’s own procedures call for using it for that purpose.
             The Department plans to install a new data system in 1993, but until
             then attention must be given to (1) making the current data base as
             accurate and complete as practical, and (2) using the current data base
             to more effectively manage the program.


             The Department of Education administers five student aid programs
Background   under Title IV of the Higher Education Act of 1965, as amended. The
             Stafford Student Loan Program is the largest of these, having under-
             written and subsidized $93 billion in student loans over the last 25
             years. From 1983 to 1989, the amount of loans made and defaults under
             this program increased dramatically. Loans increased from about $7 bil-
             lion annually to about $12 billion- a 7 l-percent increase; defaults rose
             from $445 million to about $1.9 billion-an increase of over 318 per-
             cent. The federal government’s contingent liability for outstanding stu-
             dent loans as of September 30, 1989, is about $48.5 billion. The Stafford
             loan process involves five parties: the student borrower, the school, the
             lender (e.g., banks and credit unions), the guaranty agency, and the
             Department of Education.

             Stafford loans are based on financial need. A student typically obtains a
             Stafford loan application at school and the school determines and veri-
             fies the student’s eligibility based on income and estimated cost of edu-
             cation The student applies to and receives a loan from a participating
             lender. One of 47 state or private nonprofit guaranty agencies guaran-
             tees the loan against default. The guaranty agency is the middleman
             between the Department and the lender, insuring the loan made by the
             lender to the student and assuring that the lender and schools meet pro-
             gram requirements.

             The Department pays the interest due while the student is in school.
             When the student begins repaying the loan, the payment includes
             interest (currently 8 percent) and principal; however, the Department


             Page 2          GAO/JMTEGBl-7   Stafford   Student Loam Awarded to Ineligible   Borrowers
B-240703




continues to pay an interest subsidy to bring the interest rate near
market level. The Department also reimburses guaranty agencies for
most of the defaulted loans they paid to lenders and for some of their
administrative costs.

The Department relies on schools to assure that students are eligible for
loans and do not abuse the system. Federal law and regulations state
that students who have defaulted on a loan may not receive another
loan (unless they arrange to repay their defaulted loans) and cannot
receive more than a certain loan amount each year.

As part of its monitoring of the program, the Department each year col-
lects Stafford loan information from guaranty agencies and consolidates
this information into the Stafford loan data base, commonly referred to
as the tape dump. This data base is maintained by a contractor for
approximately $900,000 a year and is primarily for the internal use of
the Department. The contractor edits the data base’s information to
check whether guaranty agencies are reporting complete data. For fiscal
year 1988, the Department required guaranty agencies to report g&per-
cent-complete data for 10 data fields the Department considered critical,
such as students’ social security numbers and loan amounts guaranteed.
However, these edit checks can only determine data completeness, not
accuracy.

The Department primarily used information in the data base for trend
analysis (Le., the number of students obtaining loans). Department offi-
cials said that information in the data base is not regularly provided to
guaranty agencies because the Department did not believe there was a
serious problem with defaulters obtaining new loans or students
exceeding loan limits.

The 1986 amendments to the Higher Education Act authorized the
Department to develop a National Student Loan Data System (NSLDS)-a
computer system that would make national student loan data accessible
to guaranty agencies. But because the 1986 amendments contained a
provision that the Department could not require guaranty agencies to
use the system before guaranteeing new loans, the system was not
developed. The 1989 Budget Reconciliation Act, however, allows the
Department to require guaranty agencies to use such a system before
approving new loans. The Department plans to develop the new system
by 1993; the guaranty agencies will have to use this system before guar-
anteeing new loans.



Page 3          GAO/IMTEGBl-7   Stafford   Student Loans Awarded   to Ineligible   Borrowera
                                         0240703




                                         By using various computer programs, we analyzed over 30 million files
Defaulters Are                           in the Department’s Stafford loan data base, which contains information
Obtaining New Loans                      gathered since 1965. These data showed that about 32,000 students had
                                         defaulted on loans totaling over $64 million and later received new loans
                                         worth over $109 million, In addition to paying for defaulted loans, the
                                         government could spend up to $65 million in interest subsidies over the
                                         life of these new 10ans.~

                                         Moreover, as figure 1 shows, in the last few years loans to students who
                                         have defaulted on earlier loans increased-from about $800,000 in 1980
                                         to about $40.2 million in 1988.


FIgwe 1: Growth of Dollar Value of New
Loans Obtained by Prior Defaulters
                                         42   Dollam In Milliona

                                         38

                                         30

                                         24


                                         18




                                          1990       1981          1282        1983       1984         1986        1886        1987         1988
                                          Fiscal Yum



                                         Because this information is based on the data contained in the Depart-
                                         ment’s data base, and some of these data are missing or inaccurate, the
                                         actual abuse could differ from these figures. A detailed discussion of the
                                         data problems we identified is contained in appendix II.



                                         2This estimate was developed on the basis of the current interest rate of 8 percent and the special
                                         allowance rate of 3.26 percent and assumes students will stay in school 4 years and take 6 years to
                                         repay their loans.



                                         Page 4                GAO/IMTEGBl-7      Stafford Student Lane Awarded to Ineligible         Borrowers
                     In order to get some estimate of the accuracy of the Department’s loan
                     data, we obtained loan documents for 79 students who the data base
                     showed as in default when they obtained new loans.3 However, of this
                     group, only 64 students (68 percent) were in default on Stafford loans
                     when they were guaranteed new loans.


                     Defaulters obtaining new loans is a significant problem. The problem of
Students Are         students obtaining loans for more than the annual or total limits doesn’t
Obtaining Loans in   appear significant, but still needs to be controlled. Unchecked, this type
Excess of Limits     of abuse has the potential to become a major problem, costing the gov-
                     ernment millions of dollars annually.

                     Title IV of the Higher Education Act of 1966, as amended, states that
                     students are limited as to how much they can borrow in any one aca-
                     demic year and in total. For instance, college freshmen and sophomores
                     can get up to $2,625 a year, while juniors and seniors can get up to
                     $4,000 a year. Graduate and professional students can receive up to
                     $7,600 a year. Total Stafford loan limits are $17,250 for undergradu-
                     ates, and $64,750-including the $17,250 obtained as an undergrad-
                     uate-for graduate and professional students. In addition, students can
                     get more than one loan during a school year if they advance in grade
                     level, for example from being a freshman to a sophomore.

                     We used computer programs to analyze the Department’s data base. Our
                     analyses showed that up to $5 million in loans could have been made to
                     students in excess of their annual limits from September 1987 through
                     August 1988. Consequently, these abuses cost the government money in
                     interest subsidies. At current interest rates, $5 million dollars in excess
                     loans could cost the government $1.8 million in subsidies over the life of
                     those loans.4

                     As in the case of defaulters, the data base is sometimes missing or has
                     inaccurate information. Some examples of abuse, however, are clear cut.
                     For instance, one graduate student received a $7,500 Stafford loan, the
                     maximum amount allowed for one year. That student also received loans
                     from three other guaranty agencies. These loans totaled $27,600 for that

                     3We limited our detailed review of defaulted borrowers to 79 cases because to review a larger sample
                     would have required work at 40 guaranty agencies. Appendix I describes our methodology for
                     selecting the 79 cases.
                     4This estimate was also developed on the basis of the current interest rate of 8 percent and the
                     special allowance rate of 3.26 percent, and assumes students will stay in school 4 years and take 6
                     years to repay their loans.



                     Page 5                GAO/IMTEGBl-7      Stafford   Student   Lcmns Awarded   to Ineligible   Fkmowere
                          5240703                                                                        ,




                          year. This student has already defaulted on $20,000 worth of loans (the
                          $7,600 loan mentioned in this example, as well as $12,500 in earlier
                          loans).

                          The Department’s data show that students exceeding federal loan limits
                          has not yet become a significant problem, but left unchecked this abuse
                          can continue and become worse. The case of the graduate student men-
                          tioned above shows how someone can exceed loan limits by applying to
                          different guaranty agencies. Until NSLDS is developed, giving guaranty
                          agencies access to each other’s data, students can continue to exceed
                          loan limits undetected.


                          Two problems are hampering the ability of the Department and the
The Department Could      guaranty agencies to prevent loan abuses, First, the Department’s auto-
Do More to Prevent        mated systems do not allow guaranty agencies to access the data base
Abuse                     on Stafford loans. Second, the data base is poorly monitored and main-
                          tained; often files on students contain suspect data or data are missing.

                          The Department plans to develop NSLDS and require the guaranty agen-
                          cies to use it before guaranteeing loans. However, the Department esti-
                          mates that the new system will not be operational until 1993. Until then,
                          the Department needs to detect and halt abuse by improving its current
                          data base and requiring the guaranty agencies to use it before they guar-
                      t   antee loans.


Data Are Inadequate       In reviewing the Department’s data base, it was hard to tell how often
                          abuse was occurring because the data were inadequate. These data,
                          which the guaranty agencies provided to the Department, are not thor-
                          oughly checked for accuracy and completeness. For instance, data were
                          sometimes missing, such as the date a loan was declared in default and
                          the date a loan was approved. As a result, we could not determine if
                          30,000 Stafford loans totaling about $71 million, which were obtained
                          by defaulted borrowers, were approved before or after they had
                          defaulted on other Stafford loans. Further, suspect data were included;
                          for instance, several different names were sometimes listed with the
                          same social security number. We found about 6,100 of these types of
                          cases affecting nearly $21 million in loans.

                          The Federal Managers’ Financial Integrity Act (FIA) (Public Law 97-265)
                          requires agencies to establish systems of internal control to ensure that
                          obligations and costs comply with applicable law, and that assets are


                          Page 6          GAO/IMTEG91-7   Stafford   Student   Loans Awarded   to Ineligible   Borrowers
                   B.249708




                   safeguarded against waste, loss, or unauthorized use. Federal agencies
                   are supposed to annually report material weaknesses in these controls
                   and the status of corrective actions until these weaknesses are
                   corrected.

                   The weaknesses and abuses discussed in this report are clear evidence
                   that critical internal controls are missing-students in default can
                   receive new loans and some students receive loans in excess of the limits
                   set forth in the law. The Comptroller General’s internal control stan-
                   dards state that with regard to the recording of transactions and events,
                   such transactions are to be promptly recorded and properly classifiede6
                   The Department is not meeting this standard. The Department’s FIA
                   reports showed that while certain aspects of the Stafford Loan Program
                   have been identified as having weaknesses in internal controls, the data
                   base has not been reported under the act. This needs to be done.


Interim Measures   To improve the internal controls in the program, the Department should
                   correct the information in its data base, identify students who have
                   defaulted on loans, and require guaranty agencies to review their stu-
                   dent loan data before guaranteeing new loans. The Department’s data
                   showed that 80 percent of defaulters wait 1 year or more after
                   defaulting before obtaining new loans. Accordingly, this historical data,
                   even with its shortcomings, could be used to prevent most defaulters
                   from obtaining new loans. Further, by using the address on the new
                   application, agencies may be able to track defaulters who have moved,
                   and try to collect on defaulted loans.

                   The Department could also analyze its data to develop criminal cases
                   against those who default. The Stafford loan application states that stu-
                   dents who provide false or misleading information are subject to prose-
                   cution under the United States criminal code. Prosecuting these cases
                   could deter others. After completing our review, we turned over to the
                   Department’s Inspector General documents on 50 of the 64 students who
                   had defaulted on loans, but claimed on a later loan application that they
                   had never been in default.6

                   Since 1984, the Department has put out an annual cross match report on
                   students who have obtained loans from more than one guaranty agency.

                   %tandards for Internal Controls in the Federal Government, GAO (Washington, DC.: 1983).
                   6The applications for four of the students did not clearly indicate whether the students had declared
                   their previous defaulted loans.



                   Page 7                GAO/lMTJK91-7      Stafford   Student   Loam Awarded   to Ineligible   Borrowers
              B-249703                                                                    ,

                                                            .




              The report is supposed to help guaranty agencies identify students who
              may have obtained loans in excess of loan limits. The report, however,
              does not contain information on the borrower’s academic year periods,
              loan balances, and whether loan funds were actually disbursed.

              Most guaranty agencies we spoke to said they do not use this report
              because it does not contain complete, accurate, or timely information
              However, one agency said it had used the report to recover $146,000 in
              1989. It said that from 1986 to 1989, it prevented the disbursement of
              $936,000 by identifying students who tried to get or did get excessive
              loans from several agencies. By improving the data base and requiring
              that the guaranty agencies check it before guaranteeing loans, the
              Department could also begin to address the problem of loans being pro-
              vided over the annual limits.


              The Stafford Loan Program, as the largest government program for stu-
Conclusions   dent loans, provides money that is vital to many students’ education.
              But this program does not have an information system in place to ensure
              that the government’s interests are protected. As evidence of this, in
              fiscal year 1990 more than half of the federal funds for this program,
              some $2.4 billion, went to repay defaulted loans. The Department of
              Education has failed to ensure that program abuses are identified and
              prevented.

              The Department is running a multibillion-dollar commercial-type loan
              operation with a data system that contains incomplete, inaccurate, unre-
              liable information. It does not have adequate systems of internal control
              to ensure that federal assets are safeguarded against waste and loss. As
              a result, students who default are able to obtain new loans and some
              borrowers are obtaining more money than they are entitled to. While the
              total impact of such abuses cannot be accurately determined because the
              data are so poor, we believe that a significant problem exists.

              The Department of Education plans to have a new data system running
              in 1993. This system is supposed to detect and prevent abuse. In the
              meantime, however, the Department needs to do much more to ensure
              that the information in its current data base is as accurate and complete
              as practical, and to make this information available to guaranty agen-
              cies. A great deal is at stake. If the abuses remain unchecked, they will
              continue to corrode the financial aid system and students who are eli-
              gible for loans may not be able to get the money they need to finance
              their education.


              Page 8          GAO/IMTEG91-7   St&ford   Student Loans Awarded   to Ineligible   Borrowers
                          E240703




                          To help curb these abuses, we recommend that the Secretary of
Recommendations           Education

                      . direct guaranty agencies to (1) ensure that their student-loan data are as
                        accurate and complete as practical, with special emphasis given to data
                        on defaulted borrowers, and (2) provide updated information to the
                        Department; and
                      l report the ability of ineligible borrowers to obtain loans as a material
                        internal control weakness under the Federal Managers’ Financial Integ-
                        rity Act.

                          Since updating this information will take time, we also recommend that
                          the Secretary take immediate action to

                      . analyze the data it now has and, until NSLDS is developed and imple-
                        mented, provide to guaranty agencies on a periodic basis the data
                        needed to prevent defaulters from obtaining new loans. This informa-
                        tion should also be used to stop and recover federal interest payments
                        on loans obtained by defaulted borrowers; and
                      l develop a more comprehensive annual report that gives guaranty agen-
                        cies the data they need to more efficiently detect loan limit abuses. In
                        addition, require guaranty agencies to use the report.

                          We also recommend that the Office of Inspector General follow up on
                          the data and documents we provided on defaulters who have obtained
                          new loans, and take appropriate action.


                          In its October 17,1990, comments on a draft of this report, the Depart-
Agency Comments and       ment of Education did not disagree that student defaulters were
Our Evaluation            obtaining new loans or that students were obtaining loans in excess of
                          legal loan limits. It plans to report a material weakness under FIA,
                          namely that borrowers can receive loans for which they are ineligible.
                          However, the Department stated that we had misinterpreted the basic
                          concept and intent of the guaranty agency tape dump. It said that the
                          tape dump was initially intended to be used primarily as an annual
                          source of data for analysis of program trends.

                          We do not believe that we have misinterpreted the concept and intent of
                          the tape dump. While the tape dump may not have been initially
                          designed to prevent ineligible borrowers from receiving loans and to
                          track down defaulters, the Department’s own procedures clearly show
                          that the intent and use of the tape dump has changed. For example, its


                          Page 9         GAO/IMTEG91-7   Stafford   Student   Loans Awarded   to Ineligible   J3orrowera
B-249703




1988 tape dump procedures state that tape dump information is neces-
sary so that the Department and guaranty agencies can

“identify possible violations of loan limits by borrowers who obtain loans guaran-
teed by more than one agency, [and track down] defaulted borrowers by obtaining
information from other agencies who have guaranteed loans for the same
borrowers.”

The Department generally disagreed with our recommendation that it
require guaranty agencies to report complete and accurate data. We rec-
ognize the Department’s concern that correcting millions of automated
loan records could cost a great deal of time and money. However, our
report cites examples in which significant amounts of critical loan data
are missing, incomplete, and inaccurate. Since these data are going to be
the foundation for the Department’s new system, we continue to believe
that it is important for the most critical data in the tape dump to be as
accurate as possible.

The Department also generally disagreed with our recommendations to
use the tape dump to prevent defaulters from obtaining new loans and
detect loan-limit abuses. The Department states that the tape dump is
not an effective vehicle for preventing these loan abuses. We agree that
NSLDS should be better able to prevent loan abuses than the current tape
dump process. However, the Department plans to wait until NSLDS is
developed in 1993 before taking aggressive corrective action on these
abuses. In the meantime, the Department needs to identify borrowers
who are illegally receiving new loans in order to protect the govern-
ment’s interest.

Detailed Department of Education comments and our evaluation are
contained in appendix III.


As agreed with your office, unless you publicly announce this report’s
contents earlier, we plan no further distribution until 30 days from the
date of this letter. At that time we will send copies to the Secretary of
Education and other interested parties, and will make copies available
to others upon request. This work was performed under the direction of




Page 10           GAO/IMTEG91-7   Stafford   Student Loana Awarded   to Ineligible       Borrowers



                         ‘/
                                                                                     .
Frank Reilly, Director, Human Resources Information Systems, who
can be reached at (202) 275-3195. Other major contributors are
listed in appendix IV.

Sincerely yours,




Ralph V. Carlone
Assistant Comptroller General




Page 11            GAO/IMTJZGSl-7   Stafford   Student   Loam Awarded   to Ineligible   Borrowers
,,,,,,,
     ,,




          Letter                                                                                                                1

          Appendix I                                                                                                        14
          Objectives, Scope,and    Methodologies for Identifying Stafford Loan Abuses                                       14
          Methodology
          Appendix II                                                                                                       19
          Problems Related to      Missing Data
                                   Questionable and Erroneous Data
                                                                                                                            19
                                                                                                                            19
          Missing, Questionable,
          and Erroneous Data
          Appendix III                                                                                                     22
          Comments From the        GAO Comments                                                                            28
          Department of
          Education
          Appendix IV
          Major Contributors to
          This Report
          Related GAO Products                                                                                             32

          Tables                   Table I. 1: Examples Where the Data Base Indicated That                                  17
                                       Four Students Had Potentially Exceeded Annual
                                       Loan Limits
                                   Table II. 1: Examples Showing Same Social Security                                       19
                                       Numbers for Borrowers With Different Names

          Figure                   Figure 1: Growth of Dollar Value of New Loans Obtained
                                       by Prior Defaulters




                                   Page 12        GAO/IMTEG91-7   Stafford   Student   Loam Awaxded to Ineligible   Borrowers


                                                        :         ,.,             ;_   ,d’
    Content+3




    Abbreviations

    FIA         Financial Integrity Act
    GAO         General Accounting Office
    IMTEC       Information Management and Technology Division
Y
    NSLDS       National Student Loan Data System
    PLUS        Parent Loans for Undergraduate Students
    SE3         Supplemental Loans for Students
    SSN         social security number


    Page 13          GAO/IMTEG91-7   Stafford   Student   Loans Awarded   to Ineliglble   Eorrowera
 Plx!

&$ktives, Scope,and Methodology


                       Because of recent interest in the Department of Education’s student loan
                       programs, the Chairman, Permanent Subcommittee on Investigations,
                       Senate Committee on Governmental Affairs, asked that we report to him
                       on our ongoing review of the Stafford Student Loan Program. Our
                       review objectives were to determine the extent to which students who
                       have defaulted on Stafford loans were obtaining new loans and the
                       extent to which students were obtaining loans that exceeded legal loan
                       limits. We performed our work at the Department of Education’s head-
                       quarters and Office of Inspector General in Washington, D.C. We also
                       contacted 11 guaranty agencies and obtained loan documents and poli-
                       cies and procedures for guaranteeing Stafford loans.

                       We obtained and reviewed documentation related to the Department’s
                       automated data base. To achieve our objective, we relied extensively on
                       computer-processed information contained in that data base, Our review
                       of the information in the data base and the results of our detailed
                       sample analysis showed data problems (i.e., questionable and incorrect
                       data) that cast doubt on the data base’s validity. However, when these
                       data are viewed in context with other available evidence, we believe the
                       opinions, conclusions, and recommendations in this report are valid.

                       We obtained and reviewed our prior reports, those of the Office of
                       Inspector General, and Departmental reports that identified deficiencies
                       in the Department’s automated data base. We also reviewed Education’s
                       Federal Managers’ Financial Integrity Act reports. We obtained and ana-
                       lyzed over 30 million computerized Stafford loan records to identify
                       potential abuses, namely defaulted borrowers obtaining new loans and
                       students exceeding loan limits. Our work was performed in accordance
                       with generally accepted government auditing standards, from Sep-
                       tember 1989 to June 1990.


                       To identify defaulted borrowers who obtained new loans and borrowers
Methodologies for      who had exceeded loan limits, we obtained and analyzed the Depart-
Identifying Stafford   ment’s fiscal year 1988 computerized data base for Stafford loans, com-
Loan Abuses            monly referred to as the tape dump. The tape dump is a cumulative
                       record of the active loans guaranteed by each guaranty agency since
                       November 1966, which the Department requires the agencies to update
                       annually. We used the data base as of September 30,1988 (fiscal year
           Y           1988 tape dump) because it contained the latest information available.




                       Page 14        GAO/IMTECOl-7   Stafford   Student   Loans Awarded   to IneligIble   Borrowem
                      4m*           1
                      Objectives,       Scope, and   MetlwdologY




Defaulted Borrowers
                      defaulted, written off or compromised;                      (3)
                      The Department directed guaranty agencies to classify borrowers who
                      have defaulted on their loans as either (1) defaulted, unresolved; (2)
                                                                   defaulted, paid in full; (4)
                      defaulted, in repayment; or (5) defaulted, assigned to Education, Each
                      one of these classifications was identified by a loan status code, such as
                      DF for defaulted, unresolved.

                      Initially we compiled data on all borrowers who were classified as
                      defaulted, unresolved; defaulted, written off or compromised; and
                      defaulted, assigned to Education for collection because these borrowers
                      were presumably not eligible to obtain new Stafford loans. This totaled
                      about 2 million students. However, federal regulations do not clearly
                      state whether defaulted borrowers who are classified as written off or
                      compromised are ineligible to obtain new loans. Because students who
                      have repaid their defaulted loans and students who are repaying their
                      defaulted loans are generally eligible to obtain new loans,’we excluded
                      these loans from our data base of defaulted borrowers.

                      After we had built our defaulter data base, we used the defaulters’
                      social security numbers (SSN) to identify other Stafford loans these bor-
                      rowers had also obtained, about 230,000 loans. Then, using the loan
                      status dates for defaulted loans (the dates the loans were declared in
                      default) and the loan guaranteed dates for other loans (the dates the
                      loans were guaranteed) we identified all borrowers who had obtained
                      new loans after they were declared in default. From this process we
                      identified 31,977 students who obtained 44,130 new loans totaling about
                      $109 million. Already, 6,700 of these borrowers who defaulted on their
                      first loans have obtained and defaulted on subsequent loans totaling
                      about $19 million.

                      We initially planned to obtain and review loan documents for statisti-
                      cally valid samples of these borrowers in order to project the accuracy
                      of the tape dump data. That is, we planned to verify whether borrowers
                      were actually in default when they obtained a new loan, by doing two
                      samples. In one sample of 300 students, we looked for borrowers who
                      had the same SSN and last names for both their defaulted loan and new
                      loan, and in one sample of 100 students we looked for borrowers who
                      had the same SSN, but different last names for both loans. We found,
                      however, it would have required substantial time, staff resources, and
                      travel costs to complete both samples because we would have been
                      required to contact 40 guaranty agencies.




                      Page 15                   GAO/IMTEC91-7      Stafford   Student   Loans Awarded   to Ineligible   Borrowers
                           Appendix    I
                           Object&n,       fkope, and Methodology




                           We therefore limited our samples to three guaranty agencies: two multi-
                           state agencies and one single-state agency. These three collectively held
                           about 44 percent of the defaulted loans and new loans in the above sam-
                           ples and were the designated guaranty agencies for nine states, the Dis-
                           trict of Columbia, and the Pacific Islands. These agencies are (1) the
                           Higher Education Assistance Foundation, which is the designated guar-
                           anty agency for Kansas, Minnesota, Nebraska, West Virginia, Wyoming,
                           and the District of Columbia; (2) United Student Aid Funds, which is the
                           designated guaranty agency for Arizona, Hawaii, and Nevada, and the
                           Pacific Islands; and (3) the state of California.

                           We requested and obtained from these three agencies loan documents
                           for 79 students-66 names from the sample of 300 borrowers and 13
                           names from the sample of 100 borrowers. To determine if these bor-
                           rowers were in default on Stafford loans and later obtained new loans,
                           we asked the three guaranty agencies to provide documents on these
                           borrowers. For defaulted loans we requested copies of the borrowers’
                           loan application forms, default claim forms filed by the borrowers’
                           lenders, and copies of any default notices and collection letters sent to
                           the borrower. For new loans, we requested copies of the borrowers’ loan
                           application forms, notices of loan guarantees by the guaranty agencies,
                           and checks or other documentation showing that loan funds were actu-
                           ally disbursed.

                           The above samples are not statistically valid for projecting to the total
                           universe of about 32,000 defaulted borrowers who were shown to have
                           obtained new loans totaling about $109 million. We believe, however,
                           the results indicate that the Department has a significant problem.


Borrowers Exceeding Loan   We used the present Stafford total loan limits of $17,250 for undergrad-
Limits                     uate students and the $64,750 overall total to identify all students who
                           potentially exceeded total loan limits regardless of when they obtained
                           their last loan. This methodology most likely understates the number of
                           students who exceeded total loan limits because before October 1986
                           total loan limits were substantially less. We chose this methodology
                           because total loan limits have changed several times since the student
                           loan program began in 1966. Identifying students who exceeded pre-
                           vious total limits, on the basis of the dates of their last loans, would
                           have required costly and complex computer analyses.

                           To determine students who potentially exceeded annual loan limits, we
                           first identified all students who obtained loans from September 1, 1987,


                           Page 16                 GAO/IMTEG914     Staff’ord   Student   Loans Awarded   to Ineligible   Borrowers
                                                Apwndlx I
                                                Objectivea,    Scope, and Methodology




                                                through August 31,1988, where (1) the loans were for classes (academic
                                                year or semester) with the same ending dates, and (2) the loans
                                                exceeded the students’ loan limits based on their academic level. We
                                                chose this period because it represents a typical academic year,
                                                including summer school, for most colleges and universities.

                                                Using the data in the tape dump we performed further analyses. Table
                                                I. 1 shows examples of four students who potentially exceeded annual
                                                loan limits based on the above criteria.’


Table 1.1: Examples Where the Data Base Indicated That Four Students Had Potentially Exceeded Annual Loan Limits
                                                                            Loan         Loan
                           Class      Class Quaranteed        Amount        Status      Status    Total      Loan
Student      AcadE%       Beoan      Ended         Date Guaranteed          Code          Date   Loans       Limit                                      Excess
                                                                                                                                                        --
1              -._- .-- 2           09/87   05188             10187        $2,625           ID                   06188
                -         2
                        -..--       09;87   OS;88             02;88        $1,500           ID        --         06/88       $4,125 -      $2,625        $1,500
                                                                                                                                                         -__
2     ..      I.. .~.__-- 1         09187   05/88             10187        $2.625           ID                   06188
                          2         09;87   05;88             02;88        $1,500           ID                   06;88       $4,125
                                                                                                                              --           $2,625 --- $1,500
3                         3         09187   05188             08187        $2,900           PF                   03188
             .            3
                         ..-_-.--   09/87   05188             10187        $1,100           ID                   II/88
                          3         09187   05188             04188        $1,450           ID                   II/88
                          3         09187   05188             05188        $1,450           ID                   11/88       $6,900        $4,000      --- $2,900
4                         4         09/87   05188             05/87        $4,000           ID                   11188
                          4         09187   05/88             06s          $2,000           ID                   II/88       $6,000        $4,000        $2,000
                                                                                                                                                            __


                                                According to the Department’s tape dump instructions, the beginning
                                                and ending class dates should refer to the specific school period covered
                                                by the loan, for example one year or one semester. Accordingly, we ini-
                                                tially identified all students such as examples 1, 2, 3, and 4 as students
                                                who potentially exceeded annual loan limits because they obtained loans
                                                that exceeded their loan limits for what appears to be the same aca-
                                                demic year.

                                                In the above table, student 1 appears to have exceeded the annual loan
                                                limits. However, student 2 was entitled to obtain a second loan during
                                                the same class period because this student had progressed to a higher
                                                academic level during the loan period. Student 4 may have been entitled


                                                ‘The annual loan limit for academic levels 1 and 2 (i.e., freshman and sophomore) is $2,626, and for
                                                levels 3 and 4 (i.e., junior and senior), $4,000. Guaranteed date refers to the date a loan was guaran-
                                                teed. Loan status code refers to students’ status, such as still in school (ID) or the loan has been paid
                                                in full (PF).



                                                P&e   17               GAO/IMTEG918        Stafford        Student   Loans Awarded    to Ineligible   Borrowers
Am’@*         1
Objectives,       Scope, and Methodology




to obtain a second loan during the same class period because the defini-
tion of one academic year depends on several variables that are not
included in the information available in the data basea We eliminated
cases such as these in determining students who exceeded annual loan
limits.

We also found that several guaranty agencies had based annual loan
limits on the students’ unpaid principal balance of all loans obtained
during the loan period. Student 3 is an example where this apparently
occurred. The table shows that this student exceeded the loan limit by
$2,900. However, using the unpaid principal balance criteria, the stu-
dent did not exceed loan limits, because the first loan of $2,900 was paid
off before obtaining the third and fourth loans. A Department official
said that the unpaid principal balance rule applies to the total loan
limits but not annual loan limits. Therefore, we did not eliminate cases
such as these in determining students who exceeded loan limits.

To find students who potentially exceeded total loan limits, we identi-
fied all students who obtained loans that exceeded the undergraduate
and total loan limits based on their academic levels.




2Definitions of academic year are cited in 34 C.F.R. 668.2 and 34 C.F.R. 682.401(b)(2Xiii).



Page 18                   GAO/IMTEC91-7    Stafford   Student   Loans Awarded   to Ineligible   Jh-rowera
ProblemsRelated to Missing,Questionable,and
Erroneous Data

                                        The number of defaulted students who obtained new loans could be sub-
Missing Data                            stantially different than indicated by the Department’s data. First, only
                                        about 30 million of the 43 million Stafford loans made by the Depart-
                                        ment are stored in computerized format. As a result, we were able to
                                        analyze only about 70 percent of the total Stafford loans made.

                                        Second, the Department records contained 30,000 loans totaling about
                                        $71 million obtained by defaulted borrowers for which we could not
                                        determine whether the loans were obtained before or after the students
                                        were in default. This is because one of two data fields was missing,
                                        either the loan status date, which identifies the date a loan was declared
                                        in default, or the loan guaranty date, which identifies the date a loan
                                        was obtained.


                                        On the basis of the poor quality of data we examined, there is some
Questionable and                        question whether about 6,100 of the 32,000 students had in fact
Erroneous Data                          defaulted and later obtained new loans. For these 6,100 students, who
                                        obtained new loans totaling about $21 million, the SSN was the same for
                                        the defaulted and new loan, but the students’ last names were not the
                                        same for both loans. The examples in table II. 1 illustrate this condition.

Table 11.1:Examples Showlng Same
Social Security Number8 for Borrowerr   Social
Wlth Dlfl ofont Names                   et;;;
                                                      Last Name     First                Birthdate       Loan Status and Date
                                        111-11-1111   Jones         Marv                 10-18-66        Default Og-8-84
                                        111-11-1111   Jones-Smith   Mary                 10-18-66        In Schoo109-9-87
                                        222-22-2222   Johnson       John                 02-05-40        Default Og-II-83
                                        222-22-2222   Baker         Nancy                02-19-45        In School09-20-86


                                        The examples show both the defaulted and new loans were obtained
                                        under the same SSNS, 11 l-l l-1 11 and 222-22-222, but under different
                                        last names. We believe for the first example the two individuals are
                                        most likely the same person, on the basis of (1) an assumption that Mary
                                        Jones married after obtaining and defaulting on her first loan and
                                        obtained the new loan under her married name, and (2) the first name
                                        and date of birth are the same. However, for the second example, there
                                        is no reason to believe the two parties are the same person because
                                        while the social security number is the same, the personal data are sig-
                                        nificantly different.




                                        Page 19          GAO/IMTEG91-7      Stafford   Student   Loans Awarded   to Ineligible   Borrowers
                           Appendix II
                           Problema Related to Missing, Questionable,
                           and Erroneous Data




                           Our analysis of the 6,119 cases showed that for 3,300 cases the stu-
                           dents’ first names and birthdates were the same, and for 1,119 cases
                           both the first names and the birthdates were different. For the
                           remaining 1,700 cases, either the first names were the same and the
                           birthdates were not, the birthdates were the same but the first names
                           were not, or data for one of these fields were missing. For the 3,300
                           cases where the birthdates and first names were the same, both bor-
                           rowers were most likely the same person, and for the 1,119 cases where
                           the birthdates and first names differed, two different individuals may
                           have been involved.

                           Included in the 32,000 defaulted borrowers were about 3,800 who
                           obtained new loans, totaling about $12.3 million, from the same guar-
                           anty agencies who had guaranteed their defaulted loans. Several guar-
                           anty agencies said they first review their computer loan data to
                           determine if loan applicants have ever defaulted on loans they have
                           guaranteed before giving the students new loans. Therefore, these stu-
                           dents may not have actually been in default when they obtained their
                           new loans.


Detailed Sample Analysis   In order to get some estimate of the accuracy of the Department’s com-
                           puterized loan data, and how much abuse has occurred, we obtained
                           loan documents for 79 students who the Department’s data showed
                           were in default when they obtained new loans. On the basis of our
                           review of these loan documents, we concluded that of the 79 students:

                           54 (68 percent) were definitely in default on Stafford loans when they
                           were guaranteed new loans.
                           19 (24 percent) were eligible to obtain new loans because they had
                           either settled or arranged to repay their defaulted loans.
                           One obtained the new loan before defaulting on the first loan.
                           Three students had not defaulted on loans and obtained new loans
                           because although the defaulted loans and the new loans were obtained
                           under the same SSN, they were apparently obtained by different
                           individuals.

                           For the remaining two students, we could not determine if they had
                           defaulted and obtained new loans because the guaranty agency holding
                           the loans did not provide us with the documents needed to make this
                           determination.




                           Page 20              GAO/IMTEG91-7     Stafford   Student   Loans Awarded   to Ineligible   Borrowers
.
    Appendix II
    Problems Related to Missing, Questionable,
    and Erroneous Data




    Although we randomly selected the 79 students, the sample is not large
    enough to accurately project the actual number of defaulted borrowers
    who obtained new loans. Nevertheless, our review verified that 68 per-
    cent of students in our sample were actually in default when they
    obtained new loans, and clearly indicates both that substantial abuse
    has occurred and that the Department’s data base is unreliable.




    Page 21             GAO/IMTEC91-7      Stafford   Student Loam Awarded   to Ineligible   Borrowers
Appendix III                                                                                                                                 .

CommentsFrom the Departmentof Educations


Note: GAO comments
supplementing those in the
report text appear at the
end of this appendix.                               UNITEDSTATESDEPARTMENTOF
                                                                         EDUCATION
                                        OFFICEOF   THE   ASSISTANT     SECRETARY          FOR    POSTSECONDARY        EDUCATION




                             Ralph V. Carlone
                             Assistant Comptroller General
                             Information Management and Technology Division
                             United States General Accounting Office
                             Washington, DC 20548
                             Dear Mr. Carlone:
                             Thank you for the opportunity to review GAOdraft report, "Stafford
                             Student Loans: Millions     of Dollars of Ineligible Loans are Being
                             Awarded," GAO/IMTEC 90-92, dated September 14, 1990.
                             Attached      is the Department of Education's                                   response to the draft
                             report.
                             If you have any questions,   please contact                                         Valerie          Hurry     of the
                             Division of Quality Assurance on 708-9453.
                                                                                                                  Sincerely,



                                                                                                                  Leonard L. Haynes III




                     Y

                                                     401, MARYLAI(ID    AVE.,   S.W     WASHINGTON,      DC   10101




                                 Page 22                  GAO/IMTTX%l-7               Stafford    Student Loans Awarded           to Ineligible   Borrowers
       Append& III
       Comments Fromthe Department
       ofEducation




                     Department of Education Response
                                     to
     General    Accounting Office  Draft Report GAO/IMTEC 90-92 --
                          Stafford Student Loans:
     Millions    of Dollars of Ineligible  Loans are Being Awarded

Prior to responding to the specific       recommendations of this
report, we would like to make some general observations regarding
the findings.    A number of findings     and recommendation$ appear to
indicate that GAO has misinterpreted       the basic concept and intent
of the guarantee agency tape dump. Although the GAO report
refers extensively     to a Stafford Student Loan "Data Base," to
ED'8 knowledge, the tape dump was the only system reviewed by
GAO.    Further, clarification     is needed regarding the name of the
Higher Education Act, Title IV, Part B programs.         ED defines
those programs as the Guaranteed Student Loan Programs, which
include the Stafford Loan Program, the PLUS Program, the
Supplemental Loans for Students (SLS) Program, the Federal
Insured Student Loan Program, and the Consolidation        Loan Program.
GAO appears to refer to these programs as the Stafford Student
Loan Programs.
The guarantee agency tape dump is a system for collecting
selected information       on each Stafford,     PLUS, and SLS loan which
is currently     outstanding or which was closed (cancelled,        paid in
full,   or filed as a death, disability         or bankruptcy claim) in the
fiscal    year covered by the tape dump. The tape dump, developed
in the late 1970'8, was initially         intended to be used primarily
as an annual source of data for analysis of program trends, and
the data elements requested reflect          that purpose.    ED did not
expect that every guarantee agency would have historically
collected    all the data elements requested, because each guarantee
agency's system was designed to meet the needs of that individual
guarantee agency.
ED collects    information     on approximately    30 million     loans each
year from the guarantee agencies through the tape dump. The
information    is received on computer tape because the receipt,
review, and analysis of 30 million          paper forms containing the
same information       is neither practical     nor feasible.      The tape
dump system is a way of collecting          information;    and the detailed
information    is consolidated onto one set of tapes and stored in a
tape library.      It is not itself     a database system which is
available   for interactive      review and correction,       though summary-
level databases can and have been designed to process some of the
information    from it.     It is not a financial      system used in the
collection    or payment of Federal funds and therefore does not
fall under the provisions of the Federal Managers' Financial
Integrity   Act.     It is a system which was developed to provide
summary program data and was therefore not designed to be used to
prevent the making of loans to ineligible            borrowers.      It is also
                                       1




       Page23
                              ApPendixLII
                              CommentsFrom the Department
                              of JSducation




                       ED   Response to GAO Draft      Report GAO/IMTEC 90-92,   Stafford   Loans



                       used currently     as the basis for calculation of the annual cohort
                       default rates     for the Secretary's Default Reduction Initiative.

See comment   1.       GAO:                     Direct guaranty agencies to (1') ensure
                       their student loan data are accurate and complete with special
                       emphasis given to data on defaulted borrowers   and (2) provide
                       updated information to the Department.
                       DePartmen+                            Contrary to the suggestion in
                       the report that ED performs only'limited     editing on the tape
                       dump, ED edits all fields on a guarantee agency's tape dump
                       submission for accuracy and completeness.      The statement in the
                       report that ED edits only the ten data fields       it considers
                       critical  is not accurate.  After the edits are completed,
                       guarantee agencies are provided with a report of ED's analysis
                       and are required to take steps to correct future tape dump
                       submissions.
                       Each year, ED reviews the tape dump in light of annual program
                       goals and needs, and identifies       fields on the tape dump from
                       which accurate data would contribute         the most to meeting these
                       goals and needs. ED concentrates         its correction    efforts   on these
                       few fields.     For the fiscal    year 1989 tape dump, ED concentrated
                       on the ten data fields     referred to in the GAO report.          These
                       fields concerned data needed for the operation of ED's School
                       Default Initiative,    one of ED's top priorities       for the past few
                       years.    For FY 1989, agencies which had data in these selected
                       fields which was not at least 95 percent accurate were required
                       to correct and resubmit the data until          it was acceptable.     Thus,
                       guarantee agencies have been, and will continue to be, required
                       to ensure their tape dump data is accurate and complete.              Where
                       necessary, agencies are required to resubmit annual tape dump
                       submissions until acceptable data is provided.
                       It is ED's goal to have as accurate data as possible.       However,
                       in ensuring acouracy, ED must consider other factors such as the
                       time and cost involved in correcting   millions    of automated loan
                       records which date back over 20 years.      In order to guarantee
                       absolute accuracy, an agency would be required to review each
                       system record against paper or microfilmed      documents.  This would
                       be an effort  of massive proportions,  and it would not be cost
                       effective.
See comment 2          dAO.Recommendation:            Report the lack of accurate and complete
                       data in the Stafford       loan data base as a material   internal

                   Y                                          2




                              Page24           GAO/IMTRG91-7StaffordStudent.LoanaAwardedtoIneUgibleRorrowers
                           Appendix III
                           Commenta From the Department
                           of Education




                 ED Response to        GAO   Draft   Report GAO/IWTEC 90-92,   Stafford   Loans



                 control        weakness under the Federal Managers' Financial        Integrity
                 Act.
                             of Education            We do not concur with this
                 recommendation.   Tape dump data huality has not had sufficient
                 negative impact on the Utivitiga   for ui.t     was Med         to &
                 Yaaa to alastaify it as a material weakness.
                 We   will, however, report a material weakness in Section 2 of the
                 F’MFIA  which describes the ability     of borrowers to receive loans
                 for  which they are ineligible     due to prior   defaults or having
                 reached statutory     loan limits.    ED has long realized that this is
                 a weakness of the loan programs, but was prohibited        by statute
                 from requiring    edits of new loans through a national system until
                 the law was changed in December,       1989.   The development of the
                 National Student Loan Data System will be listed as the
                 corrective   action for this material weakness.
See comment 3.   GAO:                        Analyze the data that it (ED) now has
                 and, Until the National Student Loan Data System is developed and
                 implemented, provide to guaranty agencies on a periodic basis the
                 data needed to prevent defaulters      from obtaining new loans.  This
                 information    should also be used to stop and recover Federal
                 interest    payment8 on loans obtained by defaulted borrowers.
                                                          ED has, in the past, reduced
                                                        xcess borrowing by providing
                 tape dump data to each guarantee agency on their borrowers who
                 also have loans guaranteed by other agencies.          ED does this on
                 an annual basis in its tape dump Vross match report subsystem.V8
                 If a guarantee agency requests its cross match data on computer
                 tape, it receives all tape dump data about their borrowers which
                 other guarantee   agencies   submitted.     If the agency requests a
                 paper printout,    it receives matching information     only on those
                 borrowers for whom the guarantee agency guaranteed loans in the
                 most recent fiscal    year.   All matching records and all tape dump
                 data  is not provided on the printout      because the printout would
                 be too long to be of practical      use.
                 The cross match report has not been fully effective          in
                 eliminating      excess and ineligible    loans in part because it is
                 based on static tape dump data which only reflects         the GSL
                 Programs at a single point in time, September 30 of one fiscal
                 year.      It takes at least 6 months for the guarantee agencies to
                 submit acceptable tape dump data, for ED to use it to prepare and
                 distribute     the cross match report, and for guarantee agencies to
                 incorporate      the information   into their computerized loan
                                                            3




                              Page25          GAO/IMTEG91-7
                                                          St&ordStudentLoansAwaxdedtoIneligibleRorrowers
                                                                                                   ,
                       AppendixIll
                       Comments From the Department
                       of Education




                ED Response to GAO Draft        Report GAO/IMTEC 90-92, Stafford               Loans



                guarantee processing systems.            During this time,    the status of
                many borrowers changes and the entire file is generally out of
                ;;:Tcnd       therefore unreliable      as a loan guarantee screening
                              These changes include:      (1) Repurchases of defaulted
                loans iy lenders, which routinely           occur as a result of'processing
                errorls    or other    circumstances:    (2) Defaulters who subsequently
                meet the repayment criteria           and resume eligibility     for Title ZV
                financial      aid; (3) Rehabilitation      of loans, under which a
                defaulted loan or loans for which a borrower has made 12
                consecutive monthly payments may be sold to a participating
                lender, in which case the borrower is no longer considered to be
                a defaulter.        Using the tape dump data to sdreen loan guarantee
                applications       would result in the rejection       of large numbers of
                guarantees      which would, upon further investigation,          be found to
                be valid based on actions which had occurred after the tape dump
                data was provided to ED. As no correction              or update mechanism is
                available      for the tape dump, the guarantors would continue to
                have erroneous data until the next year's submission--a period of
                approximately       18 months since the data was current.
                To  have an effective  screening program which only rejects
                borrowers who are truly ineligible,    the data provided to the
                guarantee agencies must be current.     Since the tape dump is not                      a
                dynamic system, it is not an effective    vehicle for such an
                activity.
                ED agrees, however, that screening loan guarantee applications       to
                prevent ineligible  borrowers and excess borrowing is an
                appropriate program objective,  and to meet this objective,    ED is
                developing the National Student Loan Data System (NSLDS). The
                NSLDS will be designed to permit immediate, interactive     updates,
                so that corrections  can be made as errors are discovered.
See comment 4   CAO:                     Develop a more comprehensive annual
                report that gives guaranty agencies the data they need to more
                effectively detect loan limit abuses. In addition,   require
                guaranty agencies to use the report.
                $                                 : The static tape dump data is
                not the most appropriate vehicle to identify    ineligible   borrowers
                and excess borrowing.    However, in the cross match report, a
                guarantee agency can receive all the information      ED has available
                on an agency's borrowers who have loans guaranteed by other
                agencies.   ED agrees that more information   could be collected and
                effectively  used. This is why ED is developing the NSLDS. ED
                has, and will continue to issue appropriate guidance, including
                necessary regulations,  to guarantee agencies on the use of GSL
                                                         4




                       Page 26           GAO/IMTEC91-7   Stafford Student Loans Awarded to Ineligible   Borrowers
                            AQQQ*    Ul
                            CommenmPromtkeDePartment
                            ofEducatIon




                     ED Response to GAO Draft        Report GAO/IMTEC90-92,   Stafford     Loans



                     data from other guarantee agencies which is, or becomes
                     available, through the tape dump or the NSLDS.
See comment 5.       OAO:                    The Office of Inspector General follow                up
                     on the data and documents we (the General Accounting Office)
                     provided on defaulters who have obtained new loans, and take
                     appropriate actions.
                                 of m                     The Office of Inspector
                     General is currently  following  ub on the data and documents
                     provided to it by GAO. All the documentation has been provided
                     to the applicable regional offices for attention   as warranted.

                     Additional:
                     The report     should reflect    the following--
See comment 6.       1. For rreveral years Congress failed to enact ED-proposed
                        legislation    to repeal the restriction  on the NSLDS.
                        Legislation    was not passed until December 1989. The December
                        1989 legislation    requests the Department to reach consensus on
                        the design of the NSLDS with all GSL constituencies     prior to
                        NSLDS implementation,    which is a time-consuming process.
See comment 7.       2. Page 2 of the draft report should reflect     a range of         $74~$109
                        nillion    as the possible amount of new loans received          by
                        ineligible    students.




                                                             5
                 Y




                            Page27            GAO/IMTEG9l-7StaffordStudentLoaneAwardedtoIneUgibleRot~owem
                                                                                                 I,
               Appendix ]ILI
               Chnment8 From the Department                                                                *
               of Education




               1. Our report did not intend to suggest that the Department only edits
GAO Comments   ten data fields on a guaranty agency’s tape dump. Our report has been
               clarified to state that all data fields are edited. However, these data
               fields are edited only to determine whether the field contains data, not
               whether the data are accurate. As discussed in our letter, we found
               incomplete data fields on 30,000 loans totaling about $71 million, which
               were obtained by defaulted borrowers. In addition, loan documentation
               reviewed for 79 students reported as defaulted showed that 19 bor-
               rowers were eligible to obtain new loans when, in fact, the tape dump
               listed them as ineligible. It may be costly to ensure that all loan data are
               accurate and complete, but we believe that, at a minimum, the Depart-
               ment needs to ensure loan data on all defaulted borrowers are as accu-
               rate and complete as practical. Because the tape dump will be the
               foundation for the Department’s planned NSLDS, it is imperative that the
               Department ensure that these data are accurate and complete.

               2. Since the Department had not reported a material internal control
               weakness regarding ineligible borrowers, it was our intention that it do
               so. As a result, we concur with the Department’s decision to report this
               weakness. Our recommendation was clarified accordingly.

               3. We found that the major reason the cross match report has not helped
               prevent excess and ineligible loans is that most guaranty agencies we
               contacted were not using the report. It is true that using the tape dump
               data, upon which the cross match reports are based, to screen loan
               applicants could delay approval of new loans for eligible applicants who
               have repaid their defaulted loans. Our analysis of Department data
               showed, however, that about $40.2 million in loans may have been given
               to ineligible students in 1988. Moreover, our review of 79 defaulted bor-
               rowers who later obtained new loans showed about 68 percent were in
               default when they obtained new loans. Therefore, because NSLDS will not
               be developed until 1993, we continue to believe that the Department
               must act now to use the tape dump data and give guaranty agencies the
               data needed to identify ineligible applicants.

               4. We agree with the Department that the tape dump is not the best way
               to identify ineligible borrowers or excess borrowing. Nevertheless, it is
               the only method available until NSLLX is developed. The Department’s
               cross match report has proved useful in preventing loan limit abuses.
               For example, one guaranty agency said it used the report to prevent the
               disbursement of $936,000 in excess loans. The Department agreed that
               more information could be collected and used. We, therefore, continue to



               Page 28           GAO/IMTECOl-7   Stafford   Student   IKMLM Awarded   to Ineligible   Borrowers
AQQendh m
Commenti From the Department
of Education




believe that the Department should develop a more comprehensive
annual report until NSLDS is developed.

6. The Office of Inspector General is taking the recommended action.

6. Although until last year the Department couldn’t require guaranty
agencies to screen loans before approving them, the act and federal reg-
ulations did not prohibit the Department from requiring guaranty agen-
cies to review loans after they were approved. Screening loans after
approval would have identified defaulted borrowers who had obtained
loans and borrowers who had exceeded legal loan limits. Then the loan
could be cancelled and immediate repayment demanded.

The 1989 legislation asks the Department to consult with a representa-
tive group of guaranty agencies, eligible lenders, and eligible institutions
to develop a mutually agreeable proposal for the establishment of NSLDS.
Because of this, we recognize that developing NSLDS could be a time-con-
suming process. This is why we recommend that the Department use the
tape dump to prevent the types of abuses this report identified until the
new system can be developed.

7. Our report accurately states that the Department’s data base showed
that ineligible students received $109 million in additional loans. The
report further recognizes that the actual amount of loans obtained by
defaulted borrowers could be greater or less than $109 million. For
example, in our letter, we state that the amount could be greater than
$109 million because, due to missing data, we could not determine
whether $71 million in additional loans received by defaulted borrowers
were obtained before or after they had defaulted on their other Stafford
loans.




Page 29           GAO/JMTEG91-7   StatIord   Student Loans Awarded   to Ineligible   Borrowers
Appendix IV
                                                                                                 * n
Major Contributors to This Report


                      Douglas D. Nosik, Assistant Director
Information           Gail M. Shedlick, Senior Evaluator
Management and        Paula N. Denman, Staff Evaluator
                      Teresa M. Schlee, Writer-Editor
Tech&logy Division,   James F. Loschiavo, Systems Analyst
Washington, DC,

                      John B. Mollet, Evaluator-in-Charge
Kans& City Regional   Donald L. Ficklin, Programmer/Analyst
Office                Kath@n A. Smith, Staff Evaluator
                      Linda K. Simpkins, Staff Evaluator




                      Page 30        GAO/IMTlW91-7   Staff’ord   Student Loam Awarded to Ineligible   Borrowem
a




    Page 31   GAO/IMTEG91-7   Stafford   Student   Loans Awarded   to Ineligible   Bomowers
belated GAO Products                                                                         . L


              GAO Views on the Stafford Student Loan Program (GA&-HRD-00-13,                        Feb.
              20,199O).

              Guaranteed Student Loans: Analvsis of Student Default Rates at 7.800
              Postsecondary Schools (GAOjHRD-iO-63BR, July 6, 1989).

              Defaulted Student Loans: Preliminary Analysis of Student Loan Bor-
              rowers and Defaulters (GAO/HRD~~-~~~BR, June 14, 1988).

              Guaranteed Student Loans: Potential Default and Cost Reduction
              ODtiOnS(GAO/HRD-8%62BR. Jan.7. 1988).

              Guaranteed Student Loans: Legislative and Regulatory Changes Needed
              to Reduce Default Cost (GAOIHRD-W-76, Sept. 30, 1987).

              The Guaranteed Student Loan Information System Needs a Thorough
              Redesign to Account for the Expenditure of Billions (GAO/HRD-81-139,
              Sept. 24,1981).




(610478)      Page 32        GAO/IMTFC91-7   Stafford   Student   Loans Awarded   to Ineligible   Borrowers
-   -   &   -._-.   “lf”.   ._.ll--l”.--I.   -._..   I   ._-.-.....___ --_.-_-   ____l__.___.___-__




                                                                       ‘I’ht* first, f”lve copies of each GAO report are free. Additional   copies
                                                                       art’ I2 each. Orders should be sent to the following address, accom-
                                                                       pauied by a check or mmey          order made out to the Suyerint~ndent
                                                                       of Documf~nts, when necessary. Orders for 100 or more copies to be
                                                                       mailed to a single address are discountSed 26 percent..

                                                                       I1.S. Gt~neral Accounting Office
                                                                       I’.(). Hox 6015
                                                                       Gaithc*rsburg, MD 20877

                                                                       Orders may also be placed by calling (202) 275-6241.
1I   Permit   No. GlOO   1
                         J