Analysis of the Proposal

The 60 bp fee seems lower than the weighted average of the spreads on Stafford (CP + 1.19% - (CP + 0.50%) = 0.69%) and PLUS (CP + 1.79% - (CP + 0.50%) = 1.29%). Assuming 14% of loan volume under ECASLA is PLUS and 86% is Stafford, the weighted average spread under ECASLA is 0.77%. The timing of the fee payment occurs sooner than under ECASLA (where accrued but unpaid interest is paid to the lender when the loan is put to the Department), but duration of the fee is shorter. Overall this 60 bp fee would be less expensive to the federal government than the current ECASLA spreads even on a net present value basis. It also would be better for lenders from a cash flow perspective.

The Department would contract with lenders for servicing through a competitive bidding process. Winning bidders would be able to continue servicing the loans they originated. But there's also a provision that allows lenders that don't win a contract to retain the servicing of their loans if the lender agreed to comply with the terms of the Department's servicing contract, including risk sharing and price. Otherwise each college would get to choose the servicer of loans made to its students.

In effect, this lets lenders have the right to continue servicing the loans they originate, albeit under the terms set by the competitive bidding process. The bidding process would be more competitive if the lenders did not have a lock on servicing the loans they originate and had to fight for the right to service loans. Moreover, since the US Department of Education would be comparing performance in preventing defaults among the servicers, loans should be assigned randomly to servicers to ensure that default rates were not skewed by the mix of schools in each lender's loan portfolio. The need to compare servicers on an apples-to-apples basis would likely outweigh the benefits of providing continuity of servicing to borrowers.

If a borrower defaults within four years of entering repayment, the servicer who serviced the loan for the two consecutive years immediately preceding the default would reimburse the Department for 3% of the loan balance (unpaid principle plus accrued but unpaid interest). This would give servicers an incentive to reduce default rates, or at least push them beyond the four-year window by encouraging the use of deferments, forbearances and income-based repayment. (Income-based repayment will function much like a deferment or forbearance when a borrower's income is less than 150% of the poverty line. Approximately half of borrowers in the income-based repayment plan will have a zero monthly payment.)

The proposal also allows colleges to choose the lenders available to their students. A college could choose a single lender under this proposal. This limits borrower choice. However, 100% direct lending would also limit borrowers to a single lender, the direct loan program. Competition between the FFEL and direct loan programs has been healthy for both, yielding benefits for both borrowers and taxpayers. It would be better to increase competition between the two programs, perhaps by adding a direct-to-consumer version of the direct loan program.

Potential Savings

Overall, this looks like it would provide similar savings to 100% direct lending, assuming that the servicing and origination proposals are replaced with something more competitive and less restrictive on borrower choice. To the extent that the ECASLA liquidity facilities are at no net cost to the federal government, this proposal would appear to also be at no net cost to the federal government. I would likely provide about 90% or more of the savings that would be provided under President Obama's proposal for 100% direct lending, assuming that the CBO analysis of the potential savings is accurate. (The CBO analysis probably overstates the potential savings by a factor of two. The Sallie Mae proposal would achieve approximately 80% of the savings under the more realistic OMB analysis.)

The additional revenue to the government through 100% direct lending by avoiding the 60 bp spread fee and the $75 put fee would be approximately $650 million to $1 billion per year over the next ten years (and about $9 billion in total, or roughly 10% of the CBO savings estimate), with about 1/5 of it coming from the spread fee and 4/5 from the put fee. Servicing costs should be similar in the two proposals.

This savings, however, would be partially offset by the costs associated with the federal government's need to establish an alternative originator for making direct loans to students at small schools that do not have the administrative staff necessary to manage the direct loan program. 100% direct lending also carries other costs, including a reduction in income tax revenue paid by education lenders and the loss of tens of thousands of jobs industry-wide. It would also require replacing some of the functions currently performed by guarantee agencies, such as default aversion, outreach and financial literacy education. On the other hand, maintaining the FFEL program also carries additional costs for guarantee agencies and program administration. There is sufficient data to provide a realistic estimate of the financial impact of these additional factors.

Transition Risk

Sallie Mae argues that their proposal would also avoid the potential disruption and delays associated with a massive transition from the FFEL program to the direct loan program and would permit the savings to be realized sooner. The Obama administration is confident that it can manage the transition efficiently. The direct loan program handled a 40% increase in loan volume from hundreds of additional colleges in 2008 without problems. However, eliminating FFELP would result in an order of magnitude greater increase in loan volume (300%) from thousands of additional colleges.