Introduction and summary
Attending college in the United States, more than in any other nation, is a high-risk, high-reward proposition. To a growing extent, students are taking on debt and betting their financial futures on the hope that their gamble on college will pay off in the future.
U.S. taxpayers are making the same gamble—funding grants and loans to put college within reach of more Americans—based on the belief that college is good for both individual students and society.
Unfortunately, this is a gamble that does not always pay off. Students do not always see lucrative returns on their investment in higher education programs. Some students fall victim to institutions that collect tuition dollars but offer little in return, while others face adversity as a result of their own decision-making. Regardless of who is at fault, students and taxpayers are the ones who pay the price. Students cannot get back the time or money they spent pursuing higher education, and taxpayers are on the hook financially when borrowers do not repay their debts. Institutions, however, are largely insulated from paying a price when their students do not succeed.
This imbalance of responsibility has prompted many to argue for a new approach to dividing the costs when federal investments in students do not produce successes. Researchers, advocates, and political leaders from across the ideological spectrum have embraced a concept called risk sharing that would create a new regime in which colleges would share in some portion of the costs generated when federal investments in higher education do not work out.
The concept of risk sharing is simple: put institutions on the hook to cover some of the costs generated when student investments in higher education do not pay off. The theory is that a risk-sharing regime would cause institutions to have more skin in the game, increasing incentives to ensure that students are set up to succeed, at least financially.
Despite its popularity as a concept, there is little consensus about the best way to construct a risk-sharing regime. For example, there are currently three bills in Congress that propose to create some sort of risk-sharing system, all of which take slightly different approaches.1
Recognizing this deficiency—and also Congressional interest in risk sharing—the Center for American Progress coordinated the publication of a body of research on the subject.2 The goal was to develop a set of concrete and detailed risk-sharing proposals that could be used to inform policymaking. The proposals—eight in all— were authored by individuals and organizations from a variety of positions within the higher education policy community, including academic researchers, think tank scholars, and analysts from advocacy and trade organizations. Not only do the proposals’ authors come from different constituencies, they also represent a range of positions on the ideological spectrum. Contributors to the research series were encouraged to carefully examine the implications of their proposals, including the unintended consequences.
The proposals produced are outlined in the following papers:
- “A New Approach to College Accountability: Balancing Sanctions and Rewards to Improve Student Outcomes” by Lindsay Ahlman, Debbie Cochrane, and Jessica Thompson, The Institute for College Access and Success, or TICAS
- “Risk-sharing: An Efficient Mechanism for Funding Student Loan Safety Nets” by Beth Akers, Manhattan Institute and formerly of the Brookings Institution
- “Getting Risk-Sharing Right: Creating Better Incentives for Colleges and Universities” by Kristin Blagg and Matthew Chingos, Urban Institute
- “Designing and Assessing Risk-Sharing Models for Federal Student Aid” by Nicholas W. Hillman, University of Wisconsin-Madison
- “A Risk-Sharing Model to Align Incentives and Improve Student Performance” by Jorge Klor de Alva and Mark Schneider, Nexus Research and Policy Center
- “Sharing the Risk: A Plan for Colleges to Participate in the Costs of Student Loan Failure” by Ben Miller and CJ Libassi, Center for American Progress
- “A Flexible Risk Retention Model for Federal Student Loans” by Barmak Nassirian and Thomas L. Harnisch, American Association of State Colleges and Universities
- “Risk-Sharing and Student Loan Policy: Consequences for Students and Institutions” by Douglas A. Webber, Temple University
The aim of this report is to synthesize the above set of proposals. Rather than summarizing each proposal, this report outlines the concepts that emerged consistently across the various proposals and highlights the more nuanced differences between the recommendations. Table 1 also provides a simple summary matrix of the policy recommendations from each proposal, although it is recommended that interested readers examine the original proposals for more detail and specifics.
Generally, we found that the proposals were shaped by the answers to three key questions:
- Which metrics should be used to measure institutional performance?
- How should the financial sanctions on institutions be calculated?
- Should institutions with vulnerable populations receive differential treatment?
This report intentionally refrains from recommending a particular risk-sharing setup. By laying out clear design questions and possible ways to address them, this report provides a guide for policymakers interested in the subject. The hope is that this detailed discussion of benefits and trade-offs will aid in the creation of risk-sharing policies that are fair, efficient, and effective.
Background: Risk in higher education
The theory underpinning the concept of risk sharing is that bringing institutions’ incentives more in line with the interests of students and taxpayers will result in institutions working harder to generate student success. Improved student outcomes will then in turn lessen the students’ reliance on publicly funded safety nets. The product of this sort of arrangement is that institutions directly bear some of the cost generated when students do not succeed—expenses that are currently borne by students and taxpayers.
Currently, students shoulder the risk in obvious ways. They invest time that cannot be recouped and some combination of their own money and creditworthiness on the assumption that future employment opportunities will make the investment worthwhile.
As for taxpayers, substantial state and federal intervention in the higher education market also puts them on the hook for some of the risk of investments when higher education does not pay off. This risk assumption happens through two primary channels: front-end subsidies and a back-end safety net. Front-end subsidies, such as grants to students and funds given directly to institutions, mean that taxpayers have made an investment with the expectation that there will be a return in terms of positive student outcomes. If an educational experience fails to yield the anticipated returns, then the resources expended will have been spent in vain. On the back end, the relationship with risk is more explicit: Federal student loan programs provide borrowers with a range of benefits that allows them to pause payments or have loan balances forgiven when they face consistently low earnings relative to their debt.3 But these programs are not free; when a student has to stop making loan payments or receives loan forgiveness, the taxpayers—again—are on the hook. And of course, borrowers who fall into extreme financial hardship due to their college expenses may make use of other social safety nets, including housing, health care, and other necessities—costs borne by taxpayers.
As already noted, the current system imposes little cost on institutions when the services they provide do not generate the expected return. As detailed in the next section, there is just one situation in which all institutions risk the loss of access to federal financial aid as a result of poor student outcomes: having too many borrowers default on their federal loans. Certain types of career programs can also land in hot water if they fail to maintain a ratio of debt payments relative to earnings for their graduates. Some states, meanwhile, hold their institutions accountable through what is known as outcomes-based funding systems, in which future funding appropriations are partially tied to results achieved. This, however, only affects public institutions, and the amount of funds at risk varies substantially by state.4
Outside of these oversight systems, there is one mechanism through which institutions can pay a price for producing bad student outcomes: reputation. It is often difficult for potential students to assess the value of different programs of study. As a result, students may make decisions about where to enroll based on the reputations of the institutions they are considering attending. It is certainly possible that a history of bad outcomes could tarnish the reputation of an institution, thereby reducing enrollment and its corresponding revenue. However, this effect is mitigated by the fact that enrollment decisions are made with incomplete information. Students often may not know the experiences and concerns of their predecessors. Moreover, the amount of institutional skin in the game is further lessened by the fact that students themselves are protected to a degree from negative outcomes by loan-repayment safety nets. This means that students are less likely to be disgruntled by the failure of their institution to meet their expectations.
This obvious imbalance in who shoulders the downside risk of federal investment in colleges likely explains the popularity of proposals calling for institutional risk sharing. Proposals in this space seek to fill in the missing leg of the stool of shared higher education responsibility composed of students, government, and institutions.
The current system of federal accountability
Under the current system of federal oversight, institutions face no regulatory penalties for poor outcomes unless their performance is so concerning that they fail to meet basic eligibility criteria, in which case they are expelled from the federal aid program.
The primary performance metric used in this system is the cohort default rate, which tracks the percentage of students who attend a given school and default on their loans within three years of entering repayment.5 Institutions whose default rate exceeds 30 percent for three consecutive years risk losing access to all forms of federal financial aid. Those with a default rate of more than 40 percent in one year risk losing access to federal loans.6
Unfortunately, this system leaves much to be desired when it comes to aligning the incentives of institutions with the incentives of students and taxpayers. First, this system has very little bite. That is to say, very few schools end up facing sanctions due to high student loan default rates. Of the 593,000 students who entered repayment on a federal loan in 2013 and defaulted within three years, just 619 students attended a college at risk of losing access to financial aid due to high default rates.7 And even the small number of schools that are facing loss of federal aid eligibility may avoid accountability thanks to a host of appeal options to avoid consequences. Second, the three-year cohort default rate is not an ideal indicator of institutional quality, even if concerns are limited to financial outcomes alone. For example, it fails to capture a range of negative loan outcomes, such as borrowers remaining current on their debt but having the balance increase over time. It also says nothing about whether the benefits of a program of study justify the price a student paid. And lastly, this system only creates an effective improvement incentive for institutions with cohort default rates that are close to the eligibility threshold, since institutions whose default rates are high but not near the eligibility cutoff have no pressing reason to get better.
Proponents of risk sharing tend to believe that this system of accountability offers insufficient protections for both students and taxpayer dollars and that, instead, the introduction of a risk-sharing regime would offer an improvement over the status quo.
Designing a system of risk sharing
The following sections are intended to help guide policymakers through the different decision points in the creation of a risk-sharing system. To that end, there are three design questions that need to be addressed:
- Which metrics should be used to measure institutional performance?
- How should the financial sanctions on institutions be calculated?
- Should institutions with vulnerable populations receive differential treatment?
The first two questions define the process for determining the financial obligation to be paid by the institutions. The third deals with how to address the unintended consequences of risk sharing—namely, the potential adverse impact of risk sharing on access for disadvantaged student populations.
Terminology
This report refers to several measures of institutional performance, often in rather specific detail. To guide readers, the commonly used definitions for these measures are provided below. Please note that many proposals suggest using modified definitions for these measures.
Default rate: This is the percentage of student loan borrowers who have defaulted on their federal student loans within three years of entering repayment. This measure tracks all individuals who entered repayment in the same federal fiscal year.
Repayment rate: This is the percentage of student loan borrowers who, after three years of repayment, have made sufficient payments to reduce the balance they owed upon entering repayment by at least $1 and not defaulted.
Completion rate: This is typically the percentage of full-time students who have never enrolled in another institution—that is to say, did not transfer into the institution—who complete their programs within 150 percent of the expected time to completion. In a bachelor’s degree program, for example, this means the percentage of students who graduate within six years.
Institutional performance metrics included
The first step in setting up a risk-sharing regime is to determine the outcomes on which institutional performance will be judged. This is an important step because the metrics used in determining institutional payments also serve as performance targets for schools. If the system creates strong enough incentives, then institutions will modify their practices to improve outcomes on these dimensions. Judging institutions on a metric that reflects student and taxpayer values will lead to an alignment of incentives, and as a result, institutions will work to achieve the outcomes that students and taxpayers desire. Using a metric that does not reflect those values will lead to unintended consequences, examined in a later section.
There was a striking level of agreement between the proposals on how best to manage the measurement of performance. Almost all of the proposals use student financial outcomes as the primary indicators of institutional performance. They differ, however, in how to quantify that performance and exactly how to define specific measures. Most of the proposals recommend that performance should be measured using some combination of three student outcomes: student loan default rates, student loan repayment rates, and completion, which includes graduation rates. See the text box above for the definitions of these measures, but note that proposals differed on whether they suggested following the traditional definition or not.
Some policy analysts prefer the repayment rate over the default rate because it captures borrowers who are not making progress in paying down their debt but have not been delinquent enough to default. Borrowers who seek a forbearance from the federal government, under which they are not required to make any payments for at least several years, will not be counted in the default rate but will be counted against the repayment rate.
A completion rate measures a different type of outcome—the share of students entering a program who ultimately finish. Unlike repayment or default rates, it need not track only the results for individuals who have federal student loans. The role of dropouts in a completion rate is different as well. Both default and repayment rates typically judge the outcomes of graduates and dropouts together. Because these measures are only concerned with what happens to loan balances, they allow a dropout to still be counted as a success as long as they do not default or fail to repay. By contrast, completion rates treat a noncompleter as a negative outcome.
The most commonly used completion rate in postsecondary education looks at the percentage of students who attended full time, never enrolled at another institution, and completed their programs within 150 percent of the typical expected time to finish—for example, six years for a four-year program. Despite its popularity, this formula is often criticized by many stakeholders because it fails to capture results for part-time students or deal with the issue of transfer students.8
While no two proposals produce identical policy designs, certain themes emerge around the types of measures recommended for use. In particular, six of the eight proposals suggest using some form of a repayment rate, either as the sole measure or in conjunction with other measures. In their description of the rationale for this choice, the authors tend to suggest using repayment rates for one or two main reasons: the limitation of cohort default rates as an effective accountability mechanism and the need for accountability at institutions where large numbers of borrowers may end up having too much debt relative to their income, potentially resulting in student loan forgiveness.
While tracking negative outcomes beyond default is a substantial benefit of repayment rates, other authors raise concerns about whether this measure may be influenced by behavioral choices that are not under an institution’s control. For example, the proposal offered by Blagg and Chingos, “Getting Risk Sharing Right: Creating Better Incentives for Colleges and Universities,” notes that borrowers choosing to prioritize paying other debts over student loans could affect repayment rates. Similarly, borrowers could choose to make the lower payments allowed under income-driven repayment even if they could afford to contribute greater amounts.
Even for those authors recommending the use of student loan-repayment rates, key issues remain. One question is of particular concern: whether the current formula commonly used by the U.S. Department of Education for student loan repayment makes sense. Under that measure, a student is deemed to be successfully repaying as long as they have paid $1 of their principal and not defaulted. The Institute for College Access and Success, in “A New Approach to College Accountability: Balancing Sanctions and Rewards to Improve Student Outcomes,” suggests revisiting the construction of repayment rates and recommends as one of those changes adjusting the repayment rate by the percentage of students who borrow. Hillman, in “Designing and Assessing Risk-Sharing Models for Federal Student Aid,” suggests that the Department of Education disaggregate both its repayment and default figures by repayment plan and status of the borrower, as well as by loan volume and servicer.
Three other proposals suggest judging repayment success based upon a tougher bar: whether students are on track to repay their loans on time. “A Flexible Risk Retention Model for Federal Student Loans” by Barmak Nassirian and Thomas Harnisch from the American Association of State Colleges and Universities and CAP’s paper “Sharing the Risk” both suggest that institutions should be responsible for some share of the difference between what students should have repaid in order to retire their debt on time and what they have actually paid. In “Risk-Sharing: An Efficient Mechanism for Funding Student Loan Safety Nets,” Akers follows a similar approach, albeit with different framing. She suggests that risk-sharing payments should reflect the cost imposed on taxpayers by borrowers’ utilization of loan-repayment safety nets. Although not presented directly as a repayment rate, the concept is similar: Students who do not repay in a timely manner would generate a risk-sharing payment for a school.
Default rates
The proposals that chose to use default rates in their risk-sharing calculation noted that these rates are a valuable accountability metric because the consequences associated with this outcome are so severe for borrowers. However, they also noted that the default rate may not capture all worrisome outcomes due to the use of income-driven repayment, deferments, and forbearances, which allow borrowers to avoid default without necessarily making progress in paying down their loans.
Of the four proposals that recommend using a default rate, only the one published by Nexus Research and Policy Center, “A Risk-Sharing Model to Align Incentives and Improve Student Performance,” included it as the sole measure of student loan performance. By contrast, CAP and Hillman recommend using a combination of default and repayment rates, with the latter proposal also calling for adjusting rates by the percentage of students who borrow. TICAS, meanwhile explores using either default or repayment rates, in both cases adjusting for a school’s borrowing rate without making a choice between the two.
Other measures
Completion of some form is the only type of measure that also showed up in more than one proposal, albeit in very different forms. The Blagg and Chingos proposal rejects typical risk-sharing structures that establish payments based on student outcomes several years after leaving an institution. Instead, it suggests that institutions should make a risk-sharing payment when a student drops out in the middle of a term. By contrast, the Nexus Research and Policy Center proposal suggests partially tying risk-sharing payments to the graduation rate for Pell Grant recipients. Although CAP does not have a direct measure for completion, it does suggest calculating risk-sharing payments separately based on results for graduates and for students who did not complete.
Calculating institutional payments
Selecting performance measures is only the beginning of a risk-sharing system. The next step requires translating an institution’s results into a calculated payment. This rests on two key questions: Should all institutions make payments or only those that fall short of a performance benchmark? And what is a reasonable payment size for an institution to make?
The majority of proposals agree that institutions should not be asked to repay any portion of loans in good standing. This means that risk-sharing payments would be calculated as a percentage of loans in default, nonrepayment, or both, depending on the proposal. For example, TICAS and Nexus Research and Policy Center suggest that risk-sharing payments be based on a share of the balance of an institution’s loans in default.9 Alternatively, Webber, in his paper “Risk-Sharing and Student Loan Policy: Consequences for Students and Institutions,” and Nassirian and Harnisch suggest that institutions should be on the hook for a share of the loans not being repaid on time. CAP, meanwhile, employs a hybrid approach that calculates payments based on loans in both categories.
There was less agreement among the proposals about exactly how to determine the proper payment levels, although three broad approaches emerged: create a preset payment schedule based on performance; require institutions to pay a share of each loan with a negative outcome; or require institutions to pay a portion of projected taxpayer losses from nonrepayment. Each is explained in greater detail below.
Preset payment schedule based on performance
This approach adopts the idea that institutions should make risk-sharing payments only if they fail to meet a predetermined performance standard. For example, Hillman recommends that only institutions whose default or nonrepayment rate is more than one standard deviation worse than the typical result for two-year or four-year schools should make a risk-sharing payment. Similarly, Nexus Research and Policy Center and TICAS both recommend risk-sharing payments for schools that fall within certain ranges of performance on measures such as default. Nexus Research and Policy Center also suggests a payment schedule based on Pell Grant graduation rates.
The proposals tied to performance targets often also vary by how much a school would pay based upon where the institution falls within that range of performance. An example from Nexus Research and Policy Center’s proposal shows how this idea works. Nexus Research and Policy Center recommends that colleges face a risk-sharing payment of up to 20 percent of defaulted loan balances based upon the institution’s cohort default rate. The percentage a school would pay depends on where it falls within a set schedule of payments. For example, schools with a cohort default rate between 15 percent and 17 percent would pay only 2 percent of defaulted loan balances, while those with a cohort default rate between 17 percent and 19 percent would pay 3 percent, and so on.
This type of set schedule approach has several noteworthy features. First, it sends a strong signal that risk sharing should only be focused on institutions that have particularly poor results. Second, it provides institutions with an easy-to-follow schedule that may help them predict payment amounts. That said, it does have the potential drawback of creating cliff effects, where institutions that fall barely on one side or the other of a threshold pay disparate amounts. For example, in the Nexus Research and Policy Center proposal, an institution with a default rate of 18.9 percent pays a rate that is 40 percent less than the amount a school with a default rate of 19.1 percent pays.
Institutions pay a share of each loan with a negative outcome
Instead of a set schedule, several proposals suggest having institutions repay a portion of affected loan balances each time a negative event occurs. This type of approach takes several forms. One is to require institutions to repay a set percentage of a loan balance each time. Webber, for example, suggests that institutions should repay 5 percent of each loan that is not making progress toward repayment after two years. Meanwhile, Blagg and Chingos suggest that institutions return 50 percent of all aid used for tuition and fees if a student drops out before the midpoint of a semester and 25 percent if they leave before the end of a semester.
Setting a fixed share of each loan or aid amount that a school returns has several noteworthy features. First, it creates a clear calculation formula that institutions can easily understand. Second, it extends the question of risk sharing to a broader set of institutions, since it means that schools would end up making payments even if their overall performance was not among the worst. This approach also avoids any issue related to cliff effects, since the total amount paid simply increases each time something bad occurs. At the same time, approaches such as the ones suggested by Webber and Blagg and Chingos do not increase the share of each loan that must be repaid as performance worsens. For example, an institution with very high dropout or nonrepayment rates is still on the hook for the same payment per bad outcome as a school with very good results.
Institutions pay a share of projected taxpayer losses from nonrepayment
Three proposals took an approach to determining payments that is driven entirely by institutional results and does not rely on a predetermined standard of performance. As a consequence, these systems include neither a set bar for performance targets nor a uniform percentage of loan balances to be repaid.
For example, the CAP proposal suggests that institutions should repay a share of their defaulted and nonrepaid loans based upon the rate at which those problems manifest. Under this approach, if a school has a default rate of 10 percent, it would repay 10 percent of defaulted loan amounts. The result is that the actual share of balances repaid by a school go up or down as its performance worsens or improves.
Similarly, the Nassirian and Harnisch proposal suggests that institutions be on the hook for some share of the loan balances that are not being repaid in a timely manner. There is not, however, a set assessment rate. Rather, the Department of Education would calculate for each borrower how much should have been paid down if the loan was going to be retired on a reasonable schedule versus how much was actually paid down. Institutions would be responsible for the sum of this difference across all borrowers but only after making adjustments for several factors, including the makeup of the student body and spending on instruction.
The Akers proposal, meanwhile, suggests that a school’s assessment rate reflect the extent to which its former borrowers need to make use of income-driven repayment and student-loan forgiveness benefits. While this is not as clear cut as charging institutions a set share of balances, it does allow for substantial variation in the amount repaid based upon a school’s past results.
Like the other approaches, a performance-based assessment regime corrects for some problems while introducing others. On one hand, it avoids the problems that the first two solutions present: cliff effects and not increasing payments as performance worsens. It also presents a more individualized payment rate for each school. On the other hand, these systems are less predictable for institutions than other approaches. It would be harder for schools to project what they are likely to repay, which could create some uncertainty in adjusting to the new risk-sharing system.
Payment calibration
Regardless of how a risk-sharing payment is calculated, it is important that the payment amount be properly calibrated. If the liability is too small, it will have no effect on institutional behavior. If it is too big, it could cause good institutions to stop serving risky students or even shutter their doors.
Most of the proposals suggest some sort of payment calibration on a sliding scale, such that the lowest-performing institutions pay a liability equal to a larger fraction of their defaulted loan balances than high-performing institutions. For example, Nexus Research and Policy Center suggests that institutions should make payments of up to 30 percent of defaulted loan balances. TICAS, meanwhile, provides a hypothetical payment system of up to 14 percent of defaulted loan balances. Webber’s structure ensures that schools would pay 5 percent of the balances of loans not being paid down.
While data limitations make it difficult to definitively know how large the payments based on defaulted loan balances may end up being in terms of the total loan dollars received by a school, several papers estimated answers to this question. For example, the CAP proposal provides some estimates of its predicted payments based on default and nonrepayment rates, and suggests that institutions would end up being responsible for somewhere on the order of 1 percent to 5 percent of all loan volume received in a year. By contrast, Hillman estimates payments would range from 5 percent to 15 percent of all loan balances.
There is some precedent for risk sharing in federal loans in roughly the range of 5 percent of balances or lower. For example, the now defunct bank-based loan system required lenders to absorb 3 percent of a loan’s balance as a loss when it defaulted—although the exact amount lost would be less when taking into account the provision of additional loan subsidies.10
Considerations for vulnerable populations
Most of the proposals discussed in this report recognize that a risk-sharing system carries the potential of dissuading institutions from enrolling greater numbers of students who are likely to struggle with debt. Relatedly, several proposals also acknowledge that institutional performance on default and repayment would be at least partly affected by the demographics of each institution’s student body.
Six of the proposals provide recommendations for dealing with the issue of an institution’s demographic makeup in one of two main ways: create upfront adjustments of performance metrics or offer bonuses to reward schools that do well. These two choices were not mutually exclusive, and some of the proposals suggest doing both. It should be noted that the proposals by Blagg and Chingos and by Akers did not recommend either of these approaches. In the former case, this is because their proposal suggests that institutions immediately return aid to the Department of Education when a student drops out so that there is no risk of unexpected future costs that could create incentives for enrollment changes. As for Akers, she recommends that the savings generated by her proposal be used to fund grant aid aimed at reducing the access gap for disadvantaged students.
Upfront adjustments of performance metrics
One approach to dealing with the potential effects of student body demographics on institutional results is to adjust the measures up front. Known more commonly as input adjustment, this requires an analysis that determines the extent to which observed student outcomes can be fairly attributed to an institution’s contribution rather than explained by its students’ intrinsic likelihood of success. A given school’s performance level can then be adjusted to reflect only the portion of its results that can be explained by the quality of the service it provides. For example, input adjustment would modify a school’s graduation rate to account for the demographic makeup of its students as a way of recognizing that enrolling a large portion of low-income students will likely lower the graduation rate. We refer to this up-front adjustment to highlight that it differs from the current system in which institutions can only lobby the Department of Education for special consideration after the agency publishes final performance metrics.
Some proposals argue that adjusting an institution’s results up front to account for its demographics can provide several benefits. First, it eliminates the reputational risk that comes with having unadjusted low results made public. It also eliminates the need for a backend appeals process where schools try to show that their raw results are better than they appear.
Input adjustment, however, is not easy to get right. It requires making difficult choices about what variables should be controlled for and how. Input adjustment also carries the risk that comparing institutions to their peers sets the bar too low. For example, if every school within a given type of colleges has poor results, then an adjustment process that only compares these schools to each other may excuse results that overall are not good enough.
Two proposals ultimately recommend some form of adjustment based on institutional characteristics. Nexus Research and Policy Center recommends adjusting the Pell Grant graduation rates based upon student risk factors identified by the Department of Education such as delaying enrollment, attending part time, or working while enrolled.11 It also recommends reducing a school’s risk-sharing payment if more than half of its students received need-based federal financial aid; the school spent a high percentage of its expenditures on education, career services, and retention; and the school was carrying out a debt management plan.
The Nassirian and Harnisch proposal also calls for several types of up-front adjustment, with the overall goal of determining what portion of the lending can be attributed to the institution instead of student characteristics or broader economic conditions. First, Nassirian and Harnisch recommend adjusting results based on the percentage of low-income students. Second, they suggest a safe harbor provision in which institutions with higher percentages of their expenditures going to instructional purposes would face lower risk-sharing payments.
Bonuses
Institutional bonuses were also a common approach to account for a school’s student body composition, with half of the proposals recommending such a solution. Under a bonus system, institutions could have their financial liabilities associated with risk sharing reduced if they undertook reforms to benefit students. Just as with up-front adjustment of accountability metrics, two different types of bonus systems emerged: rewards for good performance and opportunities for institutions to reduce their risk-sharing payments.
Regardless of the approach chosen, three themes emerged among all bonus proposals. First, all of the proposals recommending a bonus system included success with vulnerable populations as part of the bonus determination. The goal of these provisions was to mitigate the negative impact of the risk-sharing regime on the enrollment of disadvantaged students and to encourage increased access for low-income students where possible. Second, all the bonus systems tie rewards to an institution’s performance enrolling Pell Grant recipients, the outcomes of Pell Grant students, or both. Third, all the bonus systems take scale into account when rewarding institutions. This is most direct in the proposals from TICAS, Webber, and CAP, all of which recommend making payments on a per-student basis.
It should be noted that four proposals did not recommend any bonus system, although most did suggest some use for the money received by the Department of Education from a risk-sharing system. For instance, Nexus Research and Policy Center recommends that the federal government put any funds paid in a risk-sharing system into a competitive grant program open to institutions where at least half the students receive need-based assistance from the Department of Education. Hillman, meanwhile, suggests that funds go into a pool used for default prevention. Akers suggests that any savings from risk sharing that are not tied to covering the costs of student loan forgiveness be returned to students in the form grant aid designed to reduce enrollment gaps. Because the Blagg and Chingos proposal has institutions return funds as soon as a student drops out, there is no risk-sharing payment made, and thus no need to plan for additional spending.
Among the papers recommending a bonus system, the structure depended in great measure on whether the rest of the risk-sharing system used input adjustment. The more risk-sharing payments are based on results that are not adjusted up-front for institutional performance, the more important it is that the potential to earn bonuses be available to all schools. By contrast, a system with substantial up-front adjustments could adopt a simpler bonus system based on the fact that it has already taken external factors into account.
Bonuses for good performance
Two proposals—TICAS and Nassirian and Harnisch—present bonuses as a way to reward institutions that demonstrate better-than-expected performance. In the former case, these bonuses would go to institutions with repayment or default rates that, when combined with borrowing rates, met certain target thresholds. For Nassirian and Harnisch, bonuses would go to institutions where students repaid more than expected after adjusting for student body composition.
In both these proposals, bonuses act as a carrot that makes the risk-sharing system more palatable for institutions by rewarding the schools that do well. This process also means there is no overlap between schools that make a risk-sharing payment and those that receive a bonus.
Opportunities for risk-sharing payment reduction
The bonus systems in CAP’s and Webber’s proposals suggest a different approach that does not restrict these rewards only to institutions with good results. This is partly because neither of these proposals suggest doing any up-front adjustments for demographics at an institution. Instead, they rely on the bonus system to help offset potential payments for schools.
Webber’s bonus proposal is the more straightforward of the two. He suggests that institutions should receive a bonus for each at-risk student, such as a Pell Grant recipient, who graduates from a school and repays their loans. The goal is to only direct rewards for success with at-risk students, which in turn creates a strong financial incentive for institutions to enroll these individuals and do well by them.
The CAP bonus proposal takes a more complex approach. It determines bonus eligibility by generating individual repayment rate targets for each institution. These figures are based on an expected repayment rate, which is calculated by adjusting for an institution’s demographics, state and local unemployment rates, and whether it is a minority-serving institution, among other factors. Institutions that exceed the expected rate by a reasonable amount receive a bonus for each student who performs better than anticipated. The CAP proposal also runs a second calculation just for Pell Grant recipients who borrow. It suggests that bonuses for successful Pell Grant repayment should be double the size of bonuses provided for overall success.
Conclusion
Because of the substantial federal investment in higher education—more than $120 billion each year—it is imperative that there is an effective system of oversight in place that aligns the incentives of institutions with those of students and taxpayers.12 Establishing a system in which institutions share in a greater portion of the risk when federal loans go wrong is a promising way to accomplish this goal.
Admittedly, risk sharing is not a simple concept. Different design choices will yield quite different systems and results. And the potential for unintended consequences—particularly around the admission of low-income or vulnerable students—is high. However, by laying out the key questions that must be answered along the way, as well as options for addressing them, this report hopefully provides a path for policymakers to chart a course through complex choices. The end goal is a risk-sharing system that is fair, thoughtful, and brings needed reform to the oversight of institutions of higher education.
Appendix: Descriptions of proposals in the risk-sharing series
The Center for American Progress commissioned seven proposals for a federal student loan risk-sharing system. CAP also produced its own proposal on the topic. The external authors were afforded complete editorial independence, and their ideas and suggestions do not necessarily reflect the opinions or views of CAP. Below is a short summary of each of the proposals. A link to each, where greater detail can be found, is provided in the endnotes.
Title: “A New Approach to College Accountability: Balancing Sanctions and Rewards to Improve Student Outcomes”
Authors: Lindsay Ahlman, Debbie Cochrane, and Jessica Thompson, The Institute for College Access and Success
Proposal: TICAS proposes an accountability system, including risk sharing and rewards, that phases in over time based on a borrower-weighted debt outcome measure. Colleges that perform well would receive financial rewards based on their low-income student enrollment and/or nonfinancial rewards. Colleges that struggle would make progressively higher risk-sharing payments—5 percent to 14 percent of the volume of defaulted loans—depending on performance. Colleges with unacceptably poor outcomes would lose access to federal financial aid. Performance thresholds would be fixed, not relative, meaning that all schools would be held to the same standards. College outcomes could be measured using one of two options: either the Student Default Risk Indicator, which is a school’s cohort default rate multiplied by its borrowing rate, or the Student Nonrepayment Risk Indicator, or SNRI, which is a school’s nonrepayment rate multiplied by its borrowing rate.13
Title: “Risk-sharing: An efficient mechanism for funding student loan safety nets”
Author: Beth Akers, senior fellow at the Manhattan Institute and former fellow at the Brookings Institution’s Center on Children and Families
Proposal: Akers proposes a system of risk sharing in which institutions pay a premium to the Department of Education to cover the cost of providing loan-repayment safety nets to their former students. Akers argues that risk sharing should be used to correct the perversion of incentives created by repayment safety nets in the federal lending program and that risk sharing as a punitive mechanism would be ineffective.14
Title: “Getting Risk Sharing Right: Creating Better Incentives for Colleges and Universities”
Authors: Kristin Blagg and Matthew Chingos, Urban Institute
Proposal: Risk-sharing proposals traditionally focus on long-term repayment metrics, but Blagg and Chingos expand the concept to include short-term metrics focused on completion. Specifically, they propose a risk-sharing system that modifies current rules around returning aid when students drop out in the middle of a term. Under their proposal, institutions would return to the Department of Education 50 percent of federal student aid used to pay for tuition and fees for students who dropped out before the middle of a term and 25 percent for those who dropped out before the end of the term. In addition, institutions would lose access to federal aid programs if a large share of their students receiving these funds ended up earning poverty wages after leaving the institution. They also propose changes to the disbursement of financial aid for living expenses so that students who drop out do not end up with as much debt.15
Title: “A Flexible Risk Retention Model for Federal Student Loans”
Authors: Thomas L. Harnisch and Barmak Nassirian, the American Association of State Colleges and Universities
Proposal: The authors propose a new federal student loan risk-retention model. They caution that a punitive risk-retention framework only works as a fail-safe mechanism against extreme institutional subpar performance, and argue for pairing any risk-retention policy with a positive incentive for enhancing quality, such as an institutional bonus system for better-than-expected outcomes. Their model distributes progressively larger financial liabilities to institutions that produce increasingly poorer borrower repayment patterns. The authors call for identifying and separating the social-policy costs of student lending—such as promoting access and opportunity—from costs that are reasonably attributable to program quality. Harnisch and Nassirian advocate for allocating policy costs to the federal government, while assigning a portion of quality costs—as measured by repayment patterns—to the institutions themselves. Their proposal would provide an allowance for institutions that enroll large proportions of low-income students. They suggest that institutions only face risk related to spending for noninstructional purposes, helping institutions that devote larger portions of their budgets on instruction.16
Title: “Designing and Assessing Risk-Sharing Models for Federal Student Aid”
Author: Nicholas Hillman, associate professor in educational leadership and policy analysis at the University of Wisconsin–Madison
Proposal: Hillman proposes a system that uses measures of student loan default and nonrepayment, each adjusted by the percentage of students at a school who borrow. Institutions would be asked to repay between 5 percent and 15 percent of the balance of loans received if they had results far outside the norm of similar schools. His paper also provides a detailed look at the characteristics of the institutions that would make risk-sharing payments under his system versus other models suggested by members of Congress.17
Title: “A Risk-Sharing Model to Align Incentives and Improve Student Performance”
Authors: Jorge Klor de Alva and Mark Schneider, Nexus Research and Policy Center
Proposal: Klor de Alva and Schneider propose a system that would be applicable to any institution where more than 25 percent of students borrow. Institutions would make a risk-sharing payment of up to 30 percent of the value of loans in default based on a combination of their three-year cohort default rate, up to 20 percent, and the graduation rate of students receiving Pell Grants, up to 10 percent. Institutions where at least half of students receive need-based federal assistance; spend a lot on education, retention, and career services; and make progress on a student loan management plan would be able to reduce their payment as much as 50 percent.18
Title: “Risk-sharing and student loan policy: Consequences for students and institutions”
Author: Doug Webber, assistant professor in the economic department at Temple University and a research fellow at the Institute of Labor Economics
Proposal: Webber proposes a risk-sharing system where institutions would be responsible for 5 percent of the balances for borrowers who are not making progress repaying their loans within two years of entering repayment. He also proposes a bonus payment for each student who successfully graduates and repays their loans. His proposal also models the interaction between risk-sharing payments and tuition levels.19
Title: “Sharing the Risk: A Plan for Colleges to Participate in the Costs of Student Loan Failure”
Author: Ben Miller, senior director for postsecondary education at the Center for American Progress, and CJ Libassi, policy analyst at CAP
Proposal: CAP proposes a system of risk-sharing payments and bonuses. The payments would be based upon an institution’s default and repayment rates. Institutions would repay an amount of affected loan dollars equal to the rate at which an unwanted outcome occurred. In other words, a school with a 10 percent default rate would repay 10 percent of defaulted loan dollars. Institutions would receive bonus payments for every student who repayed beyond an expected level.20
About the authors
Beth Akers is a senior fellow at the Manhattan Institute. Before joining the Manhattan Institute, she was a fellow in the Brookings Institution’s Center on Children and Families. Akers previously held the position of staff economist with the president’s Council of Economic Advisers, where she worked on federal student lending policy, as well as other education and labor issues. She is an expert on the economics of education, with a focus on higher education policy. She is the co-author of Game of Loans: The Rhetoric and Reality of Student Debt. Akers received a Bachelor of Science in mathematics and economics from the University of Albany-SUNY and a Ph.D. in economics from Columbia University.
Ben Miller is the Senior Director for Postsecondary Education at the Center for American Progress. He was previously the research director for higher education at New America, as well as a senior policy adviser in the Office of Planning, Evaluation and Policy Development at the U.S. Department of Education. Miller’s work focuses on higher-education accountability, affordability, and financial aid, as well as for-profit colleges and other issues. Miller’s work has appeared in The New York Times, Los Angeles Times, The Chronicle of Higher Education, and Inside Higher Ed, among other outlets. He holds a bachelor’s degree in history and economics from Brown University.