The Federal Credit Reform Act of 1990 may not impose enough discipline, as the latest Congressional Budget Office report on federal credit programs shows. The CBO examined three of the biggest ones: student loans, the FHA and the Export-Import Bank.
For each, the CBO calculated cash flow for the next decade – interest and principal payments, fees, expected defaults – and then calculated what those dollars would be worth today, based on a notional interest rate known as the discount rate. All three will make money for Uncle Sam between 2015 and 2024, the CBO concluded. However, when the agency ran the numbers using "fair-value" accounting – i.e., using a higher interest rate more reflective of the real-world risks assumed by the government – it found that all three programs will require big, implicit subsidies: $14 billion over 10 years for the Export-Import Bank, $63 billion for the FHA and $135 billion for student loans.
Alas, the law requires the budget to reflect the rosier of these scenarios. The policy rationale is that government, unlike banks, is relatively indifferent to credit risk; its social priorities, such as educating the populace and promoting homeownership, override getting every last dime back for the taxpayers. If you forced Congress to account for the market risks of its portfolio, that would limit its ability to spend on other worthy purposes.
To us, such arguments tend to confirm that government credit programs can be a way to evade budgetary discipline. It's not self-evident that government should be indifferent to risk on behalf of taxpayers, since not all taxpayers benefit equally, if at all, from the credit programs.
The point isn't that Congress should get rid of student loans, much less any other particular program; it is that the past few years have taught us that credit of all kinds is riskier than it may appear. Whatever decisions the government makes, its books should reflect their actual costs fully and realistically.