Skip Navigation
| Skip Navigation |
|
|
|
|
|
July 7, 1998
The Honorable Edward M. Kennedy
Ranking Minority Member
Committee on Labor and Human Resources
United States Senate
Washington, DC 20510
Dear Senator:
I am writing in response to your letter of June 4, 1998, which requested the views of the Congressional Budget Office (CBO) on the possibility of using auction-based mechanisms to determine the yields that lenders earn on Federal Family Education Loans. The attached analysis addresses the types of auction-based mechanisms that might be studied and tested and the kind of information that a demonstration or pilot project could provide. Consistent with CBO's mandate to provide impartial analysis, the analysis makes no recommendations.
If you have questions about the analysis or would like further information, please contact me. Your staff may wish to contact Robin Seiler of CBO's Special Studies Division or Nabeel Alsalam of CBO's Health and Human Resources Division.
Sincerely,
June E. O'Neill
Attachment
cc: The Honorable James M. Jeffords, Chairman
INTRODUCTION: PROBLEMS WITH THE FFEL PROGRAM
USING AUCTIONS TO ADDRESS THOSE PROBLEMS
OTHER ISSUES TO CONSIDER IN INCORPORATING AUCTIONS INTO
THE FFEL PROGRAM
The Federal Family Education Loan (FFEL) program, which encourages private lenders to make federally guaranteed loans to postsecondary students and their families, uses a formula to determine the yield that lenders earn on such loans. That formula effectively sets a price for the services that lenders provide. Recent analyses by the Congressional Budget Office (CBO) and the Department of the Treasury have highlighted two basic problems that result from that price-setting mechanism.
First, the government cannot prevent some low-cost lenders from earning persistently above-normal profits. If the interest rate on FFELs is set too high, the government and students pay more than is necessary to get low-cost lenders to participate in the program. (Such lenders may have low costs per loan because they have large portfolios or lend mainly to students with large loan balances.)
Second, the government cannot ensure that all eligible students can borrow from private lenders. If the yield that lenders earn on FFELs is set too low, some firms will not participate in the program or will choose to do business mainly with schools whose students borrow large amounts (and, thus, have low servicing costs per dollar borrowed).
Both of those problems exist because the government does not have information about lenders' costs, and because those costs vary both among lenders and over time. The problems are inherent in any effort to induce private firms to participate in a market by setting a uniform price to be paid for their services. CBO and the Treasury Department have determined that, because of those problems, the FFEL program costs taxpayers more than is required to achieve its policy objectives.
The government recently lowered the yield that FFEL lenders earn on loans originated on or after July 1, 1998. That rate is scheduled to drop further in October 1998. The controversy surrounding those reductions in lenders' yield, and recognition of the problems with the FFEL program discussed above, have led education policymakers to consider using auctions to bring new market incentives into the program. Auctions would allow lenders to compete by bidding for loan business. Competitive bidding would in effect replace the fixed interest rate formula with a pricing mechanism that would base each lender's returns on its specific costs.
Two models of auctions for the FFEL program have been discussed in the education-policy community. The first model would involve "rights auctions," in which the authority to make federally guaranteed loans to students at specific schools would be sold at auction to lenders that made the most favorable bids. Rights might be sold separately for very large schools and for groups of smaller ones organized geographically or, perhaps, by length of program, average cost of attendance, or whether the schools were public or private. Students at each school would be directed to the lender that had won the right to make federally guaranteed loans to them. The interest rate formula and the current federal guarantee of 98 percent of loan principal would not be modified.
The rights-auction model would make only one change in the FFEL program--substituting a rights auction for students' choice of lenders. Recent improvements made by schools, guarantee agencies, and lenders in the delivery system for student loans, as well any disadvantages of that system, would be retained. Large-volume lenders might have a competitive advantage in bidding for the authority and obligation to lend to students at schools with which they already do business. Lenders that do a small volume of business or serve students at small schools might have a competitive disadvantage.
A small-scale demonstration or pilot project could produce valuable information about the strengths and weaknesses of rights auctions and about possible approaches to implementing them. Because such auctions would make little change in the current program, lenders would probably participate at a sufficient level to allow a pilot to produce clear results even if it was carried out on a small scale. However, the schools taking part in a pilot would incur costs associated with changing to a new program, and their students would lose the ability to choose their lenders. Thus, the government might have to create financial incentives for schools to participate in order to secure the desired level of activity.
The second model would involve "loans auctions," in which the federal government would mandate and manage the sale of FFELs after they were originated. Currently, many originators sell those loans to secondary-market entities before the borrowers begin repayment. The loans-auction model would replace such voluntary, private transactions with mandatory sales that would occur under close federal supervision. Proponents of that model believe the government could structure the auctions so as to encourage greater competition among lenders and produce more net proceeds to the government than would occur under rights auctions.
The loans-auction model is not fully developed, and significant further refinement would be needed before the model could be tested in a pilot. The model raises two key questions: what entities would originate FFELs and finance them prior to sale, and how would the government determine the amount of compensation those entities received? With respect to the first issue, it would be difficult to use private lenders as originators without overpaying them to some degree, since each firm would have an incentive to overstate its costs and the government would have trouble collecting accurate information. If a new government-sponsored enterprise was created to originate federally guaranteed student loans, it would be likely to have a dominant position in the market. But an absence of competition would dampen the enterprise's incentive to hold down costs. Alternatively, the government could require the student-lending industry to charter a mutually owned corporation that would originate and provide short-term funding of FFELs. The objective would be to structure the ownership and transactions of the corporation so as to give it an incentive to keep costs down, although an absence of competition would make that difficult as well.
With respect to the second issue, it has been suggested that the government would set minimum prices for different bundles of loans to be sold. If the winning bid for a bundle of loans exceeded the minimum price, the seller and the government would each receive a portion of the difference. If all bids were below the minimum price, the seller could retain the loans until the next auction, perhaps with the government providing financing. This approach would require the government to have reasonably good cost information, which would be difficult to obtain from private lenders.
A number of issues related to incorporating auctions into the FFEL program
would deserve further study before a pilot began. Those issues include
the size of the savings expected to be achieved from auctions, what rules
for disclosure of bids would be advisable, the type of auction process
most likely to maximize the government's receipts, and the appropriate
frequency of auctions. Other important topics for study are the potential
effects of auctions on students' access to credit, on their ability to
consolidate loans originated under auctions with those originated outside
auctions, and on schools, guarantee agencies, and different types of lenders.
A final issue that merits consideration is whether possible changes in
the terms of FFELs should be evaluated in the context of a pilot.
INTRODUCTION: PROBLEMS WITH THE FFEL PROGRAM
The Federal Family Education Loan (FFEL) program encourages private borrowing and lending to students and their families to help them pay for postsecondary education. Under the program, participating banks and other financial institutions lend money to students at an interest rate that is specified in law and that fluctuates with market conditions. For loans originated before July 1, 1998, that rate equals the yield on 91-day Treasury bills plus 3.1 percentage points. (That rate applies to students once they have left school and have begun repaying their loans. While they are in school, the interest rate is lower.) Lenders have found that rate attractive because the government assumes nearly all of the risk of default by guaranteeing 98 percent of the principal of the loans. That federal guarantee lowers the interest rate that private lenders would need to charge to earn an adequate profit; it also induces them to lend to students who might otherwise not be able to obtain credit. For loans originated on or after July 1, 1998, the interest rate will be slightly lower--equal to the yield on 91-day Treasury bills plus 2.8 percentage points while borrowers are in repayment.
Research by the Congressional Budget Office (CBO) and the Treasury Department has concluded that the FFEL program costs taxpayers more than is necessary to achieve the policy objective of ensuring that all eligible students have access to private credit on affordable terms to finance their education. Most of the budgetary cost of the program stems from the substantial subsidies that are conveyed to students through the government guarantee, from the ceiling that the government imposes on the interest rate paid by students, and from the government's payment of the interest on some loans (called subsidized loans) while students are in school.(1) Lenders earn profits that are higher than normal (in the sense of being more than adequate to attract capital, given the risks of the business) to the extent that their yield exceeds their costs of originating, funding, and servicing the loans. Although the magnitude of above-normal profits for lenders is not known precisely, they probably account for only a small fraction of the budgetary cost of the FFEL program.
A major source of inefficiency in the program is the government's use of a uniform interest rate formula to set the yield that lenders earn on all FFELs. That formula effectively sets a price for the services that lenders provide. Two basic problems result from the price-setting mechanism. First, the government cannot prevent some low-cost lenders from earning profits that are persistently higher than normal. If the government sets the interest rate on FFELs too high, both it and students end up paying more than is necessary to encourage low-cost lenders to participate in the program. Second, the government cannot ensure that all eligible students can borrow from private lenders. If the yield that lenders earn on FFELs is set too low, some financial institutions will not participate or will do business mainly with schools whose students borrow large amounts (and, thus, whose loans are less costly to service per dollar borrowed).
One source of those problems is the government's lack of information about the costs of FFEL lenders. Another source is the fact that those costs vary considerably both among lenders and over time. Three factors are responsible for that variation. First, although each lender's origination and servicing costs are relatively fixed, at least in the short run, there are significant differences in how much students borrow.(2) Because an FFEL lender earns the same interest rate on loans to all borrowers, it can lend more profitably to students who take out larger loans. Second, significant differences exist in the efficiency of FFEL lenders. More efficient lenders have lower per-loan origination, servicing, or funding costs, often because they have larger loan portfolios that enable them to achieve economies of scale. Third, lenders' costs vary as they adopt new technologies and as conditions in financial markets change.
In the past, the government appears to have set the yield that lenders earn on FFELs high enough to induce the industry to make loans to virtually all eligible borrowers. But that yield has been set higher than needed to encourage participation by low-cost lenders. As a result, such lenders have earned above-normal profits, which has made the program more costly to taxpayers than it would be if each lender earned a yield that reflected its particular costs. No FFEL lender has an incentive to reveal its costs to the government, however, so the government cannot vary a lender's yield based on those costs.
As noted above, lenders' yield on FFELs has been reduced by 0.3 percentage
points for loans originated on or after July 1, 1998. That yield is scheduled
to drop further--by as much as 1.2 percentage points--in October. Such
adjustments, however, cannot correct the problems associated with using
a uniform interest rate formula. Within limits, lenders can respond to
statutory reductions in the yield on FFELs by lowering their costs. As
CBO's March 1998 analysis showed, lenders can reduce their origination
and servicing costs by doing business mainly with schools whose students
borrow large amounts; they can lower their funding costs by securitizing
the loans they make. (Securitization is the process of selling debt securities
to investors, with groups of loans serving as collateral for the debt.)
To a degree, such behavior lets lenders continue to earn above-normal profits
even after their yield on FFELs has been reduced by law. Moreover, even
if such reductions do curtail the high profits earned by some low-cost
lenders, a sufficiently large reduction might also cause students at some
schools to have trouble obtaining private credit.
USING AUCTIONS TO ADDRESS THOSE PROBLEMS
The shortcomings of the FFEL program have raised interest in introducing market incentives that would stimulate competition among lenders. Competition is already a feature of the program in that lenders compete to do business with students who attend specific schools and to consolidate the loans of borrowers who are in repayment. But although such competition can result in better service for institutions and students, and may also lower the interest rates that some low-risk students pay, it does not reduce the cost of the FFEL program to taxpayers. One possible way to lower that cost is to incorporate auctions into the program, with the aim of reducing the above-normal profits for lenders that result from using a uniform interest rate formula.
The Administration and some Members of Congress have suggested that a study be conducted to identify and evaluate the advantages and disadvantages of auctions (and possibly of other approaches to bringing new market incentives into the program). Such a study could consider how auctions would affect all interested parties: students, schools, lenders, guarantee agencies, and taxpayers. The study could be followed by a small-scale demonstration or pilot project to test alternative approaches to implementing auctions. This analysis by CBO provides an overview of how the government could use auctions to increase market incentives in the FFEL program. It examines in a preliminary way the issues that an in-depth study would analyze in more detail, and it discusses the information that could be obtained through a pilot. The analysis focuses on two types of auctions for federally guaranteed student loans that have been discussed in the education-policy community. Those types differ with respect to what the government would sell at auction and how the lending process would be structured.
"Rights Auctions": Auctioning the Authority to Originate FFELs
One type of auction could be added to the FFEL program without changing the terms of federally guaranteed student loans and without imposing significant new costs on students, schools, guarantee agencies, or the government.(3) That type is known as "rights auctions" because the government would auction the authority to originate FFELs to a particular set of students. Each winning bidder would be obligated to serve all of the eligible students to whom it had won the authority to lend. That authority would be allocated by groups of borrowers who matriculated at eligible postsecondary institutions during a particular school year and would last throughout their attendance at those schools. Allowing the winner of each year's auction for a given school to lend to students throughout their attendance at that school would enable the lender to spread its relatively fixed origination and servicing costs over the largest possible loan balances. It would also allow borrowers to make payments to just one lender for each school they attended and would limit the number of lenders that each school had to deal with.
This model of FFEL auctions would be relatively easy to put in place, would require little change in the current delivery system for loans, and would involve minimal additional investment by lenders or the government. The government would continue to rely on the private sector to organize and manage the lending process. However, lenders would compete for business on the basis of the price they paid or received from the government rather than on the basis of their level of service to schools or the terms they offer to borrowers, as is the case today. To monitor lenders' performance, the government could establish a system in which schools and students reported on the quality of service they received.
Bidding in Rights Auctions. One relatively simple approach to taking bids in rights auctions would be to let lenders offer one-time payments to, or receive one-time payments from, the government in exchange for lending authority. Lenders would offer to pay the government for such authority if they expected loans to have relatively high average balances and, therefore, low servicing costs per dollar borrowed. Conversely, lenders that expected loans to have high servicing costs would offer to assume the obligation to lend in exchange for payments from the government. That approach to bidding would allow auctioned FFELs to have exactly the same payment terms as ones originated before auctions began. Thus, all of the loans would be as liquid and marketable in the secondary market as they are today. Other approaches to bidding in rights auctions are also possible and merit further study.
Two features of rights auctions might tend to lower the bids offered. First, lenders would be uncertain about the volume and characteristics of the loans that would be originated under each auction. Second, firms that were new to student lending might find that acquiring the capacity to originate FFELs was so costly as to deter them from bidding.
Some incumbent FFEL lenders could have a competitive advantage, and others a disadvantage, in bidding for the authority to lend to students at schools with which they now do business. If the authority to lend to students at a particular school was auctioned by itself (rather than with the authority to lend to students at other schools), a lender that now does a large volume of lending to students at that school might be in the best position to make a winning bid. It would already have the capability to originate a large number of loans and also have information on the loan histories of students at the school. Conversely, a small lender that made loans to students at a small school might be at a disadvantage in bidding for the authority to lend to new students at that school if such authority was bundled with the authority to lend to students at many other schools and the lender did not have the additional origination or short-term funding capacity. However, those lenders could serve as contract originators for winning bidders, if such arrangements proved cost-effective. Bidders could plan to use the secondary market to resell loans and to separate servicing and funding from loan origination, as some lenders do under the current program. Over time, the lowest bids would tend to reflect the costs of the institutions that were the most efficient servicers and providers of long-term funding.
Implementing Rights Auctions. The government could implement rights auctions in a number of ways. However, the trade-offs among the different approaches to implementation are not yet clear, which suggests that determining which approach would be most likely to achieve the program's goals is a practical issue that could best be settled through experimentation. Thus, if the Congress wanted to pursue rights auctions in the FFEL program, it would be sensible to conduct a small-scale pilot program to discover the strengths and weaknesses of such auctions, and possible approaches to setting them up, before broadly implementing a full-scale auction system. A pilot could evaluate different designs in terms of how well they promoted the goals of minimizing above-normal profits earned by FFEL lenders, on the one hand, and ensuring borrowers' access to federally guaranteed student loans, on the other.
One issue to consider in designing a rights-auction pilot is whether students who received loans originated by lenders that had submitted winning bids would have the right to consolidate those loans with loans originated outside the pilot. When a borrower consolidates several FFELs into a single loan, any lender holding a loan that is prepaid loses the ability to expense its origination costs over the remaining life of the loan. In the current program, all lenders are on the same footing with respect to that risk, and they can hope that such losses will be balanced by gains from making consolidation loans. However, lenders that had originated loans as part of a pilot would not be on the same footing as lenders that had not: by having bid for and won the right to make loans, they would effectively incur different origination costs than other lenders. Thus, if students retained the ability to consolidate loans originated through a rights-auction pilot, the net proceeds to the government from the pilot could be somewhat lower than what would be expected under a full-scale auction system.
A pilot could evaluate several key issues related to setting up rights auctions. One issue is whether it would be advantageous for the government to solicit separate bids for a right of first refusal to originate loans to borrowers at particular schools, and for the obligation to originate loans to all borrowers remaining after a right of first refusal had been exercised. Some schools might have a mix of borrowers whose loans were expected to produce positive and negative rates of return. Separating auctions of rights from obligations to lend could encourage lenders that specialized in high- or low-return borrowers to bid for authority to lend at such schools. That could result in greater competition and higher net proceeds for the government.
A second implementation issue would be how to bundle together loan-origination authority for sale. Bundling authority from multiple schools by state or geographic region might encourage lenders to serve students at schools that were small or whose students had low average levels of indebtedness. Another option would be to bundle origination rights on the basis of whether the loans were similar or diverse in terms of loan or borrower characteristics. For instance, lenders might offer to pay more for bundles with similar characteristics. Alternatively, it might be advantageous to allow lenders in each year's auctions to tender bids for bundles of their choosing; that way, the bids would reveal the aggregation of origination authority in that year that would maximize the government's net proceeds.
A realistic objective of a pilot program would be to test approaches in order to identify those that seemed most likely to make the transition to a full-scale system as smooth as possible. Obtaining information about how a full-scale auction would affect students, schools, and guarantee agencies would also be desirable. However, the government might need to create financial incentives for schools to participate in pilots in order to secure the desired level of activity, since rights auctions would expose schools to the possibility of having to deal frequently with new lenders. Because rights auctions would make minimal changes in the existing loan-delivery system and require little added investment, even modestly sized pilots would probably indicate fairly clearly the magnitude of the savings to the FFEL program that could result from full-scale implementation. If a number of different approaches to implementing rights auctions were tested through pilots, it might be necessary to auction a significant proportion of the program's loan volume to learn about each approach.
"Loans Auctions": Federally Managed Auctions of Recently Originated FFELs
Increasing the competition among FFEL lenders through auctions can give those lenders more incentive to hold down their costs and induce more efficient lenders to share their savings with the government through the auction process. Some analysts believe, however, that the current structure of the markets for originating, servicing, and funding federally guaranteed student loans might prevent rights auctions from significantly reducing the cost of the FFEL program. Those analysts argue that because new entrants into student lending must incur costs to develop the capacity to originate FFELs, rights auctions might not produce much competition for the authority to lend to students attending many schools, particularly small ones. They also argue that because economies of scale exist in student lending, the segment of the industry that serves schools with large student bodies and whose students borrow large amounts would be likely to consolidate into a small number of lenders that might eventually be able to avoid competing vigorously in the auction process. As noted above, another concern about rights auctions is that lenders would be unsure about the volume and characteristics of loans that would be originated under each auction, which would tend to lower their bids.
Those concerns have led analysts at the Department of Education to develop a second model for FFEL auctions. Under that model, known as "loans auctions," the government would manage the sale of federally guaranteed student loans that had been recently originated by other parties. In managing the sales, the government would aim to structure the markets for originating, servicing, and funding FFELs so as to encourage greater competition among lenders than would otherwise occur. In the industry's current structure, many lenders provide short-term financing of the loans they originate but then, after borrowers leave school but before they begin repayment, sell the loans to secondary-market entities, such as Sallie Mae (the Student Loan Marketing Association), that provide long-term funding of the loans. That process allows banks and other lenders that deal directly with students to specialize in originating FFELs; it also lets secondary-market entities specialize in borrowing large amounts of money and in servicing loans in repayment according to federal regulations. Currently, all sales of federally guaranteed loans are private, voluntary transactions.
The loans-auction model is not fully developed, and significant further study and refinement would be necessary before the model could be tested in a pilot. The model raises two key questions: what entities would originate FFELs and finance them prior to sale, and how would the government determine the amount of compensation those entities received? With respect to the first issue, existing FFEL lenders could perform the origination and short-term funding functions, but it would be difficult to avoid overpaying them to some degree, since each firm would have an incentive to overstate its costs, and the government would have difficulty collecting accurate information.
An alternative would be to give a new government-sponsored enterprise (GSE) the authority to originate federally guaranteed student loans, fund them up to the point at which borrowers started repaying, and then sell the loans at auction under terms specified by the government. The government's implicit guarantee of the GSE's debt would convey subsidies that would probably enable the enterprise to dominate the market for originating FFELs. However, if such a GSE had a dominant market position, the absence of competition would dampen its incentive to hold down costs. This alternative would lack a key feature of the FFEL program--the use of many private lenders--and would somewhat resemble the federal direct student loan program, in which the government originates loans itself and finances them with Treasury funds. It would differ from the direct loan program, however, in that loans would be sold to the private sector, which would finance them over their remaining life.
Another option would be for the government to require the student-lending industry to charter a mutually owned corporation that would originate and provide short-term funding of FFELs. How that would work is unclear, but the objective would be to structure the entity's ownership and transactions so as to give it an incentive to keep down its costs. Achieving that objective would be difficult without competition, however.
With respect to the second issue, it has been suggested that the government, in managing loans auctions, would set minimum prices for different bundles of loans to be sold. If the winning bid for a bundle of loans exceeded the minimum price, the seller and the government would each receive a portion of the difference. If all bids were below the minimum price, the seller could retain the loans until the next auction, perhaps with the government providing financing. This approach would require the government to have reasonably good cost information, which would be difficult to obtain from private lenders.
Proponents argue that a potential advantage of loans auctions is that originators could keep the responsibility of servicing loans that were later sold at auction. This feature, they claim, could increase auction proceeds by attracting bidders that do not have servicing capacity but are interested in receiving interest income and loan principal. But it is not clear how the government could separate servicing and long-term funding of student loans through loans auctions more efficiently than the private sector does now. Servicing and long-term funding are separated today when secondary-market entities securitize student loans they have purchased from originators and when banks securitize loans they have originated. Some originators and secondary-market entities also hire outside servicers to manage their student loan portfolios.
CBO's analysis of the loans-auctions model assumes that, both in a pilot and after a transition to a full-scale system, FFELs sold at auction would have the same terms as ones originated before auctions. That could require the government to keep the bids tendered by specific lenders confidential. If those bids were made public, they would convey information about how much firms were willing to pay for loans sold in the secondary market outside auctions, which could undermine the firms' negotiating position in that market. How keeping those bids confidential could affect the government's net proceeds from loans auctions deserves further study.
Rights auctions, whether as a pilot or a full-scale system, would be
less subject to concerns about confidentiality because there is no current
market for the right to originate FFELs (since students choose their lenders).
The information that bids for lending rights would convey would be only
loosely connected to the value of existing FFELs.
OTHER ISSUES TO CONSIDER IN INCORPORATING AUCTIONS INTO THE FFEL PROGRAM
The government's experience with auctions conducted by the Treasury Department, the Federal Communications Commission (FCC), and other agencies clearly indicates that details in the design of auctions can have a significant effect on their proceeds. That suggests that studying design issues for FFEL auctions in more depth before commencing a pilot would be worthwhile. In preparing a study, it might be useful to run computer simulations or conduct experiments to explore the implications of various design features, as the FCC and the National Telecommunications and Information Administration did with the FCC's auctions.
CBO has identified a number of other issues related to incorporating auctions into the FFEL program that would merit further study before a pilot began operating. First, it would be useful to estimate the magnitude of the savings that auctions could be expected to achieve in the program. Although a small-scale pilot would probably shed light on that issue, it would be helpful to make an estimate of the potential savings before a pilot was launched.
A second issue concerns the possible effects that auctions could have on the objective of giving all eligible students access to federally guaranteed student loans. In particular, could reducing lenders' returns lead some current lenders to leave the industry? Would certain types of schools be likely to shift from the FFEL program to the federal direct student loan program because of poor service from private lenders whose profits had been reduced by the use of auctions?
A third issue is the potential effects of auctions on private lenders that specialize in different lending functions or that do small or large volumes of business. How would introducing auctions tend to change the roles of different types of lenders? In particular, how would auctions affect lenders who do small amounts of lending and tend to sell their loans in the secondary market? What business structures would be likely to develop between small and large lenders, or other firms, under auctions?
Additional questions include, What information from auctions would it
be advisable for the government to make public or keep confidential? How
would alternative disclosure rules affect the willingness of lenders and
secondary-market investors to participate in the industry? How far before
the beginning of a school year or term would it be advisable to conduct
rights auctions? How many rights auctions could practically be conducted
at one time? How often should loans auctions be conducted? And what type
of auction process would be most likely to maximize the government's receipts?
1. Whenever the yield that lenders are entitled to earn on FFELs under the program exceeds the interest rate paid by students, the government makes special allowance payments to make up the difference. In recent years, those payments have accounted for a small portion of the program's cost.
2. For example, a March 1998 analysis by CBO estimated that the average level of indebtedness of students leaving four-year colleges (about $12,500 in 1996) is less than half that of students leaving graduate and professional schools (about $26,000).
3. Another approach to bringing new market incentives into the FFEL program to lower its cost would be to modify the terms of the loans. Possible changes include removing the ceiling on the interest rate paid by students, allowing each student's rate to be determined by market conditions, and reducing the proportion of loan principal guaranteed by the federal government. A thorough analysis of such changes would have to examine their potential effect on the policy objectives of the FFEL program; it would also need to consider general issues about the appropriate magnitude and targeting of federal subsidies to students and the appropriate roles of federal credit assistance and grants as means of conveying those subsidies. Such a discussion is beyond the scope of this preliminary CBO analysis of auctions.