This appendix discusses in detail our methodology for addressing two
research questions for the Direct Loan program: (1) How does borrower
participation in Income-Based Repayment (IBR) and Pay As You Earn
(PAYE) compare to available estimates of eligibility, and to what extent
has the Department of Education (Education) taken steps to increase
borrower awareness of these plans? and (2) What is known about Public
Service Loan Forgiveness (PSLF) certification and eligibility, and to what
extent has Education taken steps to increase awareness of this program?
To address these questions, we used data from Education, the
Department of the Treasury (Treasury), the loan servicer that administers
PSLF for Education, and the Department of Labor’s Bureau of Labor
Statistics. We reviewed relevant federal laws, regulations, and
documentation from Education. We also conducted interviews with
officials from Education and three of its loan servicers, Treasury, and the
Bureau of Labor Statistics; representatives of higher education
associations; borrower advocacy groups; researchers; and a
nongeneralizable sample of Direct Loan borrowers.
Page 39 GAO-15-663 Federal Student Loans
1 We conducted this
performance audit from November 2013 to August 2015 in accordance
with generally accepted government auditing standards. Those standards
require that we plan and perform the audit to obtain sufficient, appropriate
evidence to provide a reasonable basis for our findings and conclusions
based on our audit objectives. We believe that the evidence obtained
provides a reasonable basis for our findings and conclusions based on
our audit objectives.
To examine participation and key characteristics of borrowers in IBR,
PAYE, and other repayment plans,2 we reviewed summary data from
Education’s National Student Loan Data System (NSLDS) on 19.3 million
Direct Loan borrowers
Page 40 GAO-15-663 Federal Student Loans
3
(excluding parent PLUS)4 who entered repayment
and had an outstanding loan balance as of September 2014. We chose
these parameters in order to get as close as possible to the eligibility
criteria for IBR and PAYE. To determine participation in IBR and PAYE,
we focused on data for 11.2 million borrowers in active repayment (not in
deferment, forbearance, or default). For borrowers with multiple loans, the
repayment plan was based on the most recent loan that either entered
repayment or was loaded into NSLDS. While borrowers with multiple
loans are able to participate in different repayment plans, we found that to
be the case for only 1 percent of borrowers in our analysis.
In addition, we reviewed available estimates of IBR eligibility from a
Treasury analysis of tax return data and Education’s student loan data for
a random sample of borrowers. These estimates, which are based on
September 2012 NSLDS data for borrowers who entered repayment in
2010 or earlier and Internal Revenue Service tax return data from 2010
and 2011, depending on the most recent available for each borrower, are
the most recent and only available estimates of IBR eligibility we
identified. We were not able to estimate eligibility using data from
Education because only borrowers who apply for income-driven
repayment plans are required to provide information on their income and
family size.
We also analyzed data from the Pennsylvania Higher Education
Assistance Agency, the loan servicer that administers PSLF for
Education, on 146,866 borrowers who voluntarily requested and had their
employment and loans certified for PSLF, as of September 22, 2014.
Page 41 GAO-15-663 Federal Student Loans
5
Specifically, we analyzed the number of certifications over time,
repayment plan participation, and available borrower characteristics (i.e.,
sector of employment, amount of student loan debt, and adjusted gross
income).6
To approximate the percentage of Direct Loan borrowers who may be
eligible for PSLF, we used 2012 Bureau of Labor Statistics data—the
most recent available.7 We calculated the percentage of workers
nationwide who were employed by federal, state, and local government
agencies and 501(c)(3) nonprofit organizations, and applied it to our
summary NSLDS data on 16.3 million Direct Loan borrowers (excluding
parent PLUS) who were in repayment, deferment, or forbearance as of
September 2014. We also applied this percentage to the sub-population
of these Direct Loan borrowers who were participating in IBR, PAYE, or
ICR, the repayment plans more likely to enable borrowers to benefit from
PSLF.
To examine how IBR, PAYE, and PSLF may affect total loan costs for
borrowers with various characteristics, we used summary data from
Education’s NSLDS, specifications from November 2014 for a calculator
on Education’s website that allows borrowers to estimate loan payments,
and program requirements based on federal laws and regulations. We
developed repayment scenarios by assigning selected levels of adjusted
gross income and loan debt to a set of hypothetical borrowers to simulate
their total payments under IBR, PAYE, and 10-year Standard plans, and
under PSLF.
Page 42 GAO-15-663 Federal Student Loans
8
In each of our scenarios, we assumed:
· each borrower is single with no dependents;
· borrowers’ initial annual adjusted gross incomes will increase 5
percent annually, consistent with the assumption Education uses for
its loan calculator;
· the poverty threshold will increase at an average annual rate of 2.3
percent, which is based on the Congressional Budget Office’s inflation
rate projections from 2014 through 2024, and is consistent with the
assumption Education uses for its loan calculator;
· all loans have an interest rate of 6.8 percent, the rate for certain
borrowers with federal student loans disbursed from July 1, 2006
through June 30, 2013;
· all Direct Loans are subsidized. While borrowers may have a
combination of subsidized and unsubsidized loans, making this
assumption allowed us to show the effect of Education paying the first
3 years of interest if an IBR or PAYE borrower’s payments do not fully
cover the interest owed on a subsidized loan.9 This assumption
means we might understate total loan costs for IBR and PAYE
borrowers with unsubsidized loans whose payments do not fully cover
interest. Education’s loan calculator assumes that all Direct Loans are
unsubsidized and therefore does not account for this potential interest
benefit for IBR and PAYE borrowers with subsidized loans. Education
officials told us this assumption would have a slight effect on total loan
costs, and that the department plans to revise the calculator by
Changing the assumptions explained above would change the monthly
and total loan costs for borrowers in our scenarios. To the extent
possible, we validated our results against Education’s loan calculator and
worked with Education officials to resolve discrepancies. These scenarios
are intended for illustrative purposes only; they do not incorporate
experiences that could affect individual borrowers’ eligibility for incomedriven
repayment or their payment amounts. For example, individual
borrowers could experience periodic unemployment or job promotions,
and get married or form families. These and other experiences could
change income levels or household size, which help determine the
applicable poverty threshold for monthly income-based payments.
We determined that data from each of these sources were sufficiently
reliable for the purposes of this report by reviewing existing information
about the data and the systems that produced them, and by interviewing
knowledgeable agency officials.
To understand program terms and eligibility requirements, we reviewed
relevant federal laws, regulations, and documentation from Education. To
determine the extent to which Education has taken steps to raise
awareness of IBR, PAYE, and PSLF, we reviewed program information
Education makes available to borrowers on its website, including fact
sheets; a loan repayment calculator; and entrance, exit, and financial
awareness counseling tools. We also reviewed information about
Education’s targeted efforts to raise awareness of IBR and PAYE,
including documentation of borrower email campaigns and partnerships,
press releases, and memoranda from the President. We compared
information on Education’s efforts to criteria outlined in contract
requirements applicable to Education’s 11 Direct Loan servicers related to
communication with borrowers and the goals and objectives in the Office
of Federal Student Aid’s Fiscal Year 2012-2016 Strategic Plan.
To examine IBR and PAYE participation and eligibility, PSLF certification
and eligibility, and determine the extent to which Education has taken
steps to raise awareness of the programs, we interviewed officials from
Education, Treasury, and the Bureau of Labor Statistics. In addition, we
interviewed representatives of higher education associations, borrower
advocacy groups, and researchers about student loan repayment and
forgiveness, including factors that may affect borrowers’ decisions about
repayment. We also interviewed representatives of, and reviewed
documentation for, 3 of Education’s 11 loan servicers, which serviced
about half of all recipients of loans owned by Education.
Page 44 GAO-15-663 Federal Student Loans
10 In addition,
during April and May 2015, we interviewed a nongeneralizable sample of
14 randomly selected borrowers about their awareness of income-driven
repayment plans. Using a random sample of Direct Loans from a 4-
percent random sample of loans from the NSLDS, we identified 4,000
borrowers who, as of January 2014, were in active repayment, deferment,
or forbearance. Education sent emails to these borrowers inviting them to
email us to participate in interviews.