Opacity in Checking Government Loan Cost Accuracy

The U.S. government’s ubiquitous role providing credit—mostly to students and homeowners—in the post-financial crisis economy is unmistakable but the ultimate costs to taxpayers can be far from obvious. While nearly all of the deliberation and debate about the price of extending government credit is directed at expected costs before loans are made, the process used to reassess costs once loans are on the books is undeniably opaque with multibillion dollar implications for taxpayers.

Budgetary rules that have been in effect for nearly 25 years require government officials to determine the price of extending credit by considering all of the cash anticipated to flow to and from the government over the life of loans made for a program during a given fiscal year (a “cohort” in U.S. budget parlance). Although concerns have been raised about a severe deficiency in that process, another shortcoming stems from the process used to true up those initial estimates as cohorts age and observed loan performance is factored into cost reestimation.[i]

The reestimate process allows credit agencies to tap directly into the general fund of the Treasury to cover any shortfalls in expected loan cohort performance without further Congressional action. If, however, performance happens to exceed expectations, the process does not provide a way to reprogram funds (or even, in certain situations, return overpayments to borrowers) to accommodate initial misestimates. Moreover, the reporting of such reestimates is typically buried deep in supplementary materials included as part of the President’s annual budget and can be difficult both to find and decipher.

FHA loan guaranties. A recent case in point involves the loans made by the Federal Housing Administration (FHA) through its Mutual Mortgage Insurance Program. In the years leading up to and during the financial crisis, FHA’s primary loan program was forecast to generate net inflows to the government (sometimes inaccurately referred to as “profits”) of $15 billion from 2000 to 2009, but officials now contend that loans made over that period will cost $37 billion—a swing of $52 billion.

This particular misestimate enabled billions of dollars in additional spending elsewhere, along with commensurate increases in the national debt, because the forecast—and now unrealized—“profits” have already been spent on other programs as uncertain future returns were swapped for certain spending through the Congressional appropriations process.

With the legal authority to go directly to Treasury to cover loan program shortfalls, agencies are able to avoid the typically arduous and acrimonious process of obtaining supplemental appropriations from Congress. FHA, however, has even greater flexibility. FHA-specific provisions in the Omnibus Budget Reconciliation Act of 1990 (OBRA-90)—the same act establishing current credit program budgetary procedures known as credit reform—created a complex funding regime for FHA enabling the agency to capitalize and use a loan loss reserve fund that can obfuscate the agency’s financial position.

As noted above, the credit reestimation process lacks transparency but at least it is reported in supplemental materials submitted with the budget. In the case of FHA reserves, however, the picture is murkier. Unlike other credit agencies—which do not have loan loss reserve funds—FHA is able to mask its financial condition in the event of a significant downturn in loan performance. Given its ability to draw money from a reserve fund held in its behalf, FHA only taps Treasury’s general fund when its reserves are depleted. The use of those reserves has resulted in a perception that FHA fared reasonably well during the financial crisis—with the exception of a relatively small $1.7 billion Treasury infusion needed in 2013. That is the number that garnered headlines–rather than the true cost of tens of billions of dollars relative to amounts initially budgeted. [ii] [iii]

Since the early 2000s, tens of billions of dollars in forecast profits were recorded in FHA’s reserve fund. If those amounts were reserved in a true rainy-day fund and not used to pay for other spending elsewhere in the budget, FHA could have appropriately used them to get it through the darkest days of the financial crisis. However, those receipts did impact other parts of budget and were largely spent elsewhere, leaving any FHA usage as essentially a double spending of the funds.

Student loans. The rise of the direct student loan program at the U.S. Department of Education was enabled by cost estimation models showing the price of direct lending to be much lower than the guaranteed student loan program it replaced. Most recently, the President’s 2016 Budget shows the government will realize nearly $14 billion in net inflows from the 2016 cohort of direct loans, while the guaranteed program had historically resulted in net outflows for the government. But while taxpayer costs have fallen by shifting to direct lending, the resulting net inflows have been decried as “obscene profits” by some elected officials who contend the government is funding itself on the backs of students.

As noted above, both the Center for Finance and Policy and the Congressional Budget Office have expressed concern about the current rules used to estimate cost and the failure to consider some costs borne by the government such as fully accounting for the risk of default, especially in hard economic times. Holding that matter aside for purposes of this post, there are also significant shortcomings in the cost reestimation process for student loans.

That fact was illustrated earlier this year when the 2016 budget was issued. Buried in the supporting materials appended to the submission were details showing that expected loan costs were expected to rise by nearly $22 billion due to a number of policy changes—both in place and planned—by the Administration. Given the budget rules described above for credit programs, Education is able to incur and absorb a cost of this magnitude without seeking authorization from Congress. Nevertheless, transparency was certainly lacking. The cost was not reported in a manner commensurate with spending of that magnitude but was ascertained by at least one intrepid reporter analyzing a series of numbers included deep in the supplemental materials of the budget. [iv]

Conclusion. Policymakers in both the executive and legislative branches have little incentive to pay close attention to credit program cost reestimates. Getting the program funded upfront tends to be the primary focus: once accomplished, the task of monitoring performance falls to career civil servants responsible for administering the government’s many credit programs. But a strong case can be made that policymakers should continue to engage on loan performance because there can be huge swings in performance—and taxpayer costs—over the life of a loan cohort compared with the assumptions used to estimate initial costs. Unexpected shifts in macroeconomic conditions, changes in administrative (e.g., loan servicing) policies, and a variety of other factors can have a profound impact on how loans fare over time.  As a result, individual loan cohorts can ultimately carry a far different price than initially anticipated—information that is vitally needed by policymakers for budgetary decisions going forward.

[i] See Deborah Lucas, MIT Center for Finance and Policy, Budgeting for Federal Credit Programs: The Case for Fair Value, http://brook.gs/1IYFxOU. Also see Congressional Budget Office, Fair-Value Accounting for Federal Credit Programs, March 2012, http://1.usa.gov/1hXLC8w.

[ii] See Jim Puzzanghera, “FHA to get $1.7 billion in its first taxpayer-funded bailout,” Los Angeles Times, September 28, 2013, http://lat.ms/1UHq6RM.

[iii] See Congressional Budget Office, Modeling the Budgetary Costs of FHA’s Single Family Mortgage Insurance: Working Paper 2014-05, September 11, 2014. “As of the end of fiscal year 2013, CBO estimates that the loan guarantees issued between 1992 and 2013 contributed $3 billion to FHA’s capital reserves, $73 billion less than the amount those loan guarantees would have contributed based on their originally estimated subsidy rates,” http://1.usa.gov/1UBJzJ9.

[iv] See Mike Grunwald, “The College Loan Bombshell Hidden in the Budget,” Politico, February 5, 2015, http://politi.co/1i2pEBN.