Better Accounting Badly Needed for Federal Student Loan Programs

Earlier this August, the Biden White House and Congressional Democrats patted themselves on the back for achieving $300 billion in deficit reduction in the so-called Inflation Reduction Act. But even aside from the fact that this deficit reduction came from tax hikes that will set back the economy, price controls that will harm future medical innovation, and a large budget gimmick that won't have a real world impact, all the claimed savings were more than wiped away in an instant with President Biden’s student loan discharge announcement.

An official cost estimate is not yet available, but the student loan proposal will likely add between $400 to $600 billion to the deficit, per the Committee for a Responsible Federal Budget. The ill-advised policy shows the need for better accountability for the student Direct Loan program. Government estimates reported that student loans would generate billions of dollars for the federal budget, but in practice they have left taxpayers on the hook for massive liabilities. 

The Government Accountability Office (GAO) shed light on this issue last month when it reported on the growth in costs of the student loan program. The Department of Education had originally estimated that the student loans issued over the past 25 years would generate $114 billion in revenues for the federal government. This assumption assumed that most of the loans would be paid back in full, earning net interest that would reduce the deficit.

Unfortunately, the government does not engage in loans and credit activities with the same careful attention to the bottom line as lenders in the private sector do. GAO found that instead of a deficit-reduction windfall, the loans actually cost the government $197 billion — a $311 billion swing from the Education Department's estimate. Add in today’s action, and the ultimate cost of loans issued over the past 25 years rises from $600 to $800 billion.

Nearly 40 percent of that cost swing, $122 billion, resulted from programmatic changes as Congress enacted carve outs and exceptions to the repayment of student loans. For example, deferments for people in the military, loan cancellation for teachers and those in public service, and income-based payments all reduced expected payments.

The largest programmatic cost changes resulted from the CARES Act, which suspended payment requirements, interest accrual, and collections for defaulted loans. Through May 2022, GAO estimates the CARES Act has added $102 billion in Direct Loan costs.

The rest of the $189 billion difference results from re-estimates based on newer assumptions and data. Some of this is attributable to re-estimates as economic conditions impact receipts from borrowers enrolled in income-driven repayment plans, but basically, the government made faulty assumptions about borrower behavior over the lifetime of the loans.

The major factor driving the poor assumptions is that the federal government uses an inaccurate method for determining the risk of loan and credit activities. This issue was not discussed by GAO in this report, though they did note that a “forthcoming report will examine government and private sector estimation methods and Education’s approach to estimating Direct Loan costs.”

Projecting the long-term budgetary impact of credit programs is difficult, since a true evaluation of the costs must take into account future cash flows. Through the Fair Credit Reform Act of 1990 (FCRA), Congress requires that the government assess the risk of loans based on the projected yields of Treasury securities with the same term to maturity. This method underestimates the value of loans or guarantees because it does not accurately consider the market risk of loans. 

An alternative method called fair-value accounting creates estimates based on the market rate of interest that a private entity would have to pay. Fair-value accounting does a better job of incorporating  the risk of defaults.

The Congressional Budget Office (CBO) annually re-evaluates federal credit programs to compare the two methods. They each result in vastly different estimates, comparable to the difference between GAO’s report that Direct Loans will result in a net liability and Education's prediction of net positive revenue. CBO found that using the FCRA method, all of the loans and guarantees issued this year will result in savings of $41 billion. But through the lens of fair-value accounting, CBO sees net liabilities of $51 billion — a $92 billion swing.

Failure to factor in market risk will only make for more inaccurate assumptions moving forward. The CARES Act suspension of interest accrual provided an opportunity for borrowers to pay down their principal, but Biden’s action will discourage responsible repayment behavior going forward. It is also likely to further the gap in projected versus actual receipts by incentivizing borrowers to pay as little as possible so they can take advantage of another bailout from the government.

Good policymaking is far more difficult when budget watchdogs are forced to use inaccurate assumptions. Congress should not allow itself to pretend that the student loan program is profitable for taxpayers — particularly when the president can introduce debt forgiveness curveballs at any given moment.