About CBO's Alternative Student Loan Cost Estimate

Last week, the U.S. House of Representatives’ Committee on Education and Labor voted 30-17 to pass a bill (PDF) that increase the Pell Grant and also establishes new programs to help community colleges, increase college completion rates, and improve early childhood learning.

These initiatives would be paid for by eliminating subsidies that are currently given to private lenders so they will offer loans to students through the Federal Family Education Loan (FFEL) Program. The loans are nearly identical to ones offered by the government through the Direct Loan Program, and lenders making them receive federal insurance to cover 97 percent of any losses they sustain if a loan defaults.

Not surprisingly, the loan companies aren’t thrilled with this proposal and are fighting back.

The latest salvo over this proposal concerns how much the Congressional Budget Office (CBO) thinks the government would save by eliminating FFEL. Last week, CBO released a cost estimate (PDF) that put these savings at $86.8 billion over the next 10 fiscal years by having the Department of Education issue all federal student loans. Yesterday, the same organization sent to a letter to Sen. Judd Gregg (R-N.H.) saying a different calculation would yield savings of $47 billion over 10 years—a difference of roughly $4 billion a year.

Certainly, savings of $47 billion over 10 years should not be dismissed outright. That figure is actually greater than the savings estimated by the Office of Management and Budget when President Obama first proposed the end of FFEL in February (PDF, Page 23). Nearly $5 billion a year could fund a lot of interesting and creative programs for access and success.

But that’s assuming the alternative CBO estimate isn’t another iteration of a somewhat misleading budgeting tactic lenders have used in the past to try and make the FFEL Program seem cheaper than Direct Loans.

This tactic is known as market risk, and attempts to measure the costs of the two student loan programs by treating them as if they were products offered without a government guarantee on the private market. (My former colleague Jason Delisle wrote a paper on this subject last October, which can be found here.)

Market risk matters in this case because federal student loans are accounted for as a net present value—a process that compares the current cost of making a loan by discounting future cash flows. Operating off of the principle that $1 in the future is worth less than $1 today, the net present value method thus estimates the total cost of a loan at the time it is disbursed.

The net present value method places a great deal of importance on the discount rate—a number that reflects the value of a future dollar versus one today. If it is easy to obtain money for the loan and it is likely to be repaid, then the discount rate is likely to be low. But if there is a high default risk or money is hard to borrow, then future payments become less valuable, resulting in a higher discount rate.

The current discount rate used by CBO does not, however, reflect any of these factors. Instead, it is legally required to use a discount rate equal to the yields on Treasury securities—a number so low it is basically a risk free rate.

If future borrower payments are treated as risk-free, then their present value remains high and the net present cost of a direct loan seems to be fairly small. It may even appear to have a negative cost, meaning it seems to make money for the government.

FFEL loans, however, look more expensive under a risk-free discount rate. This is because CBO’s estimates of FFEL loans only include subsidies, fees, and default payments between lenders and the government. (Loan disbursement and borrower payments are not measured because they occur only between the lender and borrower.) Since all of these government expenses occur in the future, a low discount rate means they have a high present value. The sum of these future costs to the government thus drives up the net cost of a FFEL loan.

For years lenders have claimed that using a risk-free discount rate is unfair since student loans are in fact risky investments. (They even paid a previous head of CBO to put out a paper [PDF] arguing in favor of using market costs.) Instead, of the risk-free Treasury rate, lenders argue that CBO should use a higher discount rate that reflects the financing and default risk the private market would assign to issuing student loans.

Asking for market-based rates in cost estimates makes sense, but just increasing the discount rate just reverses problems with the current system. Direct loans cost more when using a higher, market-based discount rate because future borrower payments would be worth less in present dollars. FFEL loans have the opposite effect, as future government subsidy or default payments appear cheaper. (Just as $1 paid to you is worth less in the future than it is in the present, so too is $1 owed.)

Simply substituting the risk-free Treasury rate for a higher market cost rate thus presents the Direct Loan Program in a more costly (and realistic) light without affording the same treatment to FFEL loans.

It’s important to have a good sense of the savings from reforms to the student loan programs, but it is disappointing to see discussions of how best to spend savings on student access and success get caught up in additional political wrangling.