Statement of
STEVEN A. MCNAMARA
ASSISTANT INSPECTOR GENERAL FOR AUDIT SERVICES
OFFICE OF INSPECTOR GENERAL
U.S. DEPARTMENT OF EDUCATION

Before the
SUBCOMMITTEE ON CRIMINAL JUSTICE,
DRUG POLICY AND HUMAN RESOURCES
Committee on Government Reform
United States House of Representatives

Regarding
STUDENT LOAN PROGRAMS

JUNE 17, 1999

Mr. Chairman and Members of the Subcommittee:

Thank you for the opportunity to discuss issues, costs and the efficiency of the Federal Direct Loan Program (FDLP). My name is Steven A. McNamara, and I am the Assistant Inspector General for Audit Services at the Office of Inspector General, Department of Education. I am representing the Office of the Inspector General today because our new Inspector General, Lorraine Lewis, was just sworn in Monday, June 14, 1999. Ms. Lewis regrets not being here today to provide our testimony, but she has not had a sufficient opportunity to become thoroughly familiar with the details of our recent report, "Study of Cost Issues, Federal Family Education Loan Program and Federal Direct Loan Program," which is the focus of my testimony today. I offer a copy of the report for the record.

In the study that led to the report, the Office of Inspector General (OIG) objective was to study and compare cost issues of the two student loan programs. We reviewed the Department's actual costs for Fiscal Years 1996 and 1997, along with audits, information reports, congressional testimony and other related documents bearing on the issues that affect FDLP and the Federal Family Education Loan Program (FFELP) costs. Our goal was to arrive at a reasonable estimate of the costs of the programs rather than precise figures.

Before discussing what the study did say, let me put to rest some misconceptions about it that have come to our attention. First, our study did not conclude that one program is inherently cheaper than the other. Second, we did not conclude that eliminating the FDLP will save the government money. Third, we did not state that the inefficiencies affect only one of the programs. Finally, our study did not state that private lenders making student loans are more efficient than the government contractors serving the FDLP.

In October of 1996, we began to study the costs of the FFELP and the FDLP as an internal OIG project and for future use if we were asked about the issue. As part of our initial analysis, we reviewed various prior cost studies, but found them inadequate to address the complex and changing state of the student loan programs. Several factors in particular made it difficult to perform a conclusive analysis of program costs, including the newness of the FDLP and the possible long-term affects of income contingent repayment.

We proceeded to a second phase to attempt to arrive at a reasonable cost estimate of the two programs based on the best available data. This was a study, not an audit, of either program.

We obtained the incomplete study results and workpapers of Macro International Inc., a private contractor that the Department had retained to study the costs of the two loan programs. We expanded on Macro???s previous work on FY 1996 costs by collecting cost and other financial data from the Department for FY 1997 for the loan programs.

Consistent with the Credit Reform Act of 1990, we segregated costs associated with all Student Financial Assistance Programs into two primary categories -- subsidy and administrative costs -- and addressed them separately in our study. Subsidy costs include interest expense, loan origination fees, default costs and other fees. Subsidy costs constitute by far the majority of FDLP and FFELP costs. The Department has limited control over subsidy costs, because the economy and Congress exert the greatest influence on these costs. For example, changing economic conditions can result in interest rate volatility that will make subsidy costs rise and fall. Subsidy costs are also subject to yearly re-estimations that make them fluctuate, sometimes significantly. Congress affects subsidy costs by its legislative decisions, such as setting borrower and lender origination fees. Administrative costs are those that the Department incurs in managing the FFELP and FDLP, including such expenses as contracting, personnel, rent, travel, communication and others. The Department can largely control these costs through effective management.

Because the Department lacks a cost accounting system, as recently required by Federal accounting standards, it does not allocate administrative costs to the various financial aid programs. Consequently, we reviewed and analyzed each of the Department???s reported administrative costs and allocated them to the particular loan program in light of the activities and services actually performed, regardless of when the Department received billing invoices or what funding source it used to pay them. As an attempt to benchmark these costs, we compared the Department's cost to administer FDLP loans to what we calculated, based on a U.S. Department of Treasury study, it would cost large lenders to administer these same loans.

During the years covered by our study, the FDLP was only in its third and fourth years of existence, meaning that FDLP administrative servicing costs had not yet reached maturity. For example, servicing costs will rise over time as more borrowers enter repayment status. Consequently, it was necessary that we project FDLP administrative costs to reflect a mature program.

We reached two principal conclusions. First, in any given year the FFELP or FDLP total costs (administrative and subsidy) may be greater, given the affect of prevailing economic conditions on subsidy costs. Since costs may be higher or lower at any one point in time, a total cost figure for any one year does not definitively answer the question of whether the FFELP or FDLP is more expensive.

Second, we concluded that inefficiencies likely affect the Department's administrative costs of both loan programs. We based this conclusion on cost calculations that we made in this study and the reviews of our audit reports and those of the General Accounting Office.

The two-year average cost per loan to administer FDLP and FFELP is $24 and $17, respectively. While the Department's FDLP administrative costs per loan were higher than its FFELP administrative costs per loan, this is due to the additional functions performed by the Department for FDLP. To assess the reasonableness of the FDLP portfolio management costs, which are loan origination, servicing and consolidation, we reduced the FDLP $24 administrative cost by $7, which is the cost to perform inherently governmental functions; the remaining $17 represents the Department's costs of managing the FDLP loan portfolio. We compared our estimate of the Department's cost to manage the FDLP - $17 per loan - to the benchmark average cost of $13, that we derived based on a U.S. Department of Treasury study of lender servicing costs that large private lenders might incur. We believe that a significant portion of the $4 difference may be due to inefficiencies. These inefficiencies can largely be controlled by effective management and include such matters as access to reliable information; having qualified technical and contract management; and using compatible automated data processing systems. We recognize that some of the differences are due to such uncontrollable factors as federal procurement policies and personnel rules.

We were unable to estimate what portion of the FFELP administrative costs result from inefficiencies, because no private sector entity performs a similar oversight function.

I want to be perfectly clear on an essential point: we are not taking the position that either program, over an extended period, is cheaper than the other. What our report does conclude is that in any given year, depending upon prevailing economic conditions that affect subsidy costs, either FDLP or FFELP may be cheaper. However, the Department can achieve significant savings in administrative costs through improved administration of both loan programs. A Performance-Based Organization (PBO) was recently established to do just that.

We submitted this report to the Department for comment before it was made public. They did not disagree with our primary conclusions. While they did disagree with some of our administrative cost allocation decisions for the two loan programs, we concluded that the disagreements would result in less than a penny per loan difference in administrative costs. We consider our results to be a reasonable estimate of costs, not a precise answer.

To improve its administration of the loan programs (and the other Student Financial Assistance programs) and to comply with accounting standards and legislative mandates, our report suggests that the Department:

• institute an activity-based costing system;
• institute interim procedures to track employees' time to the program and activity they work on;
• develop models to predict borrower behavior, loan volume projections, and the cost effects of management decisions; and
• consider and take appropriate actions to address possible reasons for cost inefficiencies as discussed in the report.

The Department has begun efforts to develop a managerial cost accounting system, and the OIG is working with them as they go forward. Further, the PBO has initiated several actions to address areas where we have found inefficiencies in our past audits. For example, plans for modernizing their information technology systems are being developed. It is too soon to tell the ultimate effect of this and other actions.

This concludes my statement and I would be happy to respond to any questions that you or other members of the Subcommittee may have on this issue or other work products of the OIG.

NOTE: To read a copy of the report, please see the web site for the Office of Inspector General, Department of Education - http://www.ed.gov/offices/OIG/misc/s1370001.pdf

Back to Congressional Testimony