Strengthening Federal Student Aid: Reforming the Student Loan Repayment System


Strengthening Federal Student Aid: Reforming the Student Loan Repayment System

February 3, 2016


Amid concerns about student debt and default, many experts are calling for an income-driven student loan repayment system that will better protect students and significantly reduce defaults. Our review of detailed proposals related to income-driven repayment in the Strengthening Federal Student Aid report reveals how alternative designs would lead to different outcomes for borrowers and for taxpayers, suggesting some important guidelines. The program should provide insurance for borrowers with unforeseen difficulties in repayment. Rather than minimizing payments, the design should ensure that most borrowers eventually repay and that incentives for institutions to overcharge and for students to overborrow are minimized.

Full Publication

Untitled Document

The loan repayment system for the 40 million Americans holding federal student debt is difficult to understand, creates bureaucratic hurdles, and leaves many borrowers with unmanageable required payments.1 As concerns about the impact of student debt on borrowers and the economy mount and the visibility of loan default increases, many voices are calling for a student loan repayment system that will better protect borrowers from financial hardship, make repayment easier, and significantly reduce defaults. Proposals for income-driven repayment (IDR) would provide insurance to students against the risk that their investments in postsecondary education do not pay off as well or as quickly as they had hoped. IDR uses a payment system that is based on income rather than a fixed payment schedule, so borrowers make higher payments when their incomes rise and lower or no payments when their incomes fall. Payment continues until borrowers fully repay what they borrowed plus interest, or, in some designs, until they reach loan forgiveness.

The first income-driven plan in the United States became available in the early 1990s, and the Obama administration has since implemented additional programs for repaying loans as a percentage of income.2 Over time, the plans have become increasingly generous, making more borrowers eligible and asking less of them. But each new option makes the system more complicated, and bureaucratic barriers still make it difficult for many borrowers to enroll and remain in an income-driven plan. Although participation in these plans is increasing, most stay in the standard 10-year repayment plan, the default plan for students entering repayment, even when its required monthly payments take up a large share of their incomes. A number of recent proposals support making income-driven repayment automatic for all student loan borrowers.

Despite the growing consensus, the proliferation of proposals with differing details makes it a challenge for policymakers to incorporate the recommendations into legislation. To address this problem, the Urban Institute, with support from the Bill & Melinda Gates Foundation’s Reimagining Aid Design and Delivery (RADD) project, conducted an analysis of a large set of proposals relating to income-driven repayment. Detailed comparisons across proposals and specific policy designs are available in the full report (Baum and Johnson 2016). This brief summarizes the key findings.

Though there are many promising ideas in the proposals that have been put forward to improve income-driven repayment, there are also some pitfalls. In particular, many of the proposed provisions would be unnecessarily complicated, which is counter to the broad consensus that complexity is a serious problem in the current system. In addition, in an effort to protect borrowers from unaffordable payments, some of the proposals—and some of the policies that have already been implemented—err too far in the other direction, asking taxpayers to take too much responsibility for covering the debts of borrowers who could reasonably be expected to meet their obligations in a well-designed system.



The strongest consensus on reforming federal student aid centers on the need for simplification. There are too many different programs with too many different provisions. Multiple types of federal student loans as well as multiple repayment plans add layers of complexity to student loan repayment.3

The array of federal student loan programs includes the following:

  • Stafford subsidized loans for undergraduate students with documented financial need, in which the government pays the interest while borrowers are in school;
  • Stafford unsubsidized loans for all undergraduate and graduate students;
  • Parent PLUS loans for parents of undergraduate students who meet minimal standards for creditworthiness;
  • Grad PLUS loans for all graduate students; and
  • Perkins loans for undergraduate students whose institutions participate in the campus-based program.

In addition to the multiple federal loan programs, several repayment plans each have different expectations about monthly payments. The automatic option, into which students are placed if they do not actively make another choice, amortizes the loan with equal monthly payments over 10 years. Students can, instead, choose a graduated plan, under which payments start at a lower level and increase every two years, or if they hold over $30,000 in debt, an extended plan, which spreads the payments out for up to 25 years. There are also five different plans under which expected payments are a function of the borrower’s income during the repayment period.4

Whatever the strengths of all of these options, it is hard to argue that it is optimal to ask borrowers to choose among them (Boatman, Evans, and Soliz 2014). Instead, the government should design an IDR plan that is as equitable and efficient as possible. This could be the only plan available, or it could be accompanied by a fixed payment plan that borrowers could select as an alternative.


The Benefits of Income-Driven Repayment

Beyond the benefits of simplification, universal and automatic IDR would act as insurance on post-college labor market outcomes, supporting repayment when earnings are high and preventing defaults when earnings are low. Most borrowers have manageable payments under the standard 10-year repayment plan, but a growing number find their incomes too low to support the expected monthly payments while maintaining an acceptable standard of living. For some of these borrowers, the magnitude of debt is disproportionate to long-term income. But for many, the problem is that they are being asked to repay too much at the beginning of their careers, when their incomes are lowest (Dynarski and Kreisman 2013). Allowing borrowers to postpone higher payments until their incomes can reasonably support them will alleviate pressures on borrowers while diminishing the burdens of delinquency and default for both the government and borrowers.5

On average, postsecondary education pays off very well in the labor market, and its returns increase as workers progress in their careers (Hershbein, Harris, and Kearney 2014). As table 1 indicates, in 2014, median annual earnings of bachelor’s degree recipients ages 25 to 34 were $19,700 (96 percent) higher than median earnings of high school graduates. Among 35 to 44 year-olds, the gap was $25,700 (103 percent). The earnings premium for the typical associate degree holder age 25 to 34 was considerably smaller at $6,500 (32 percent), but associate degree recipients also tend to have less debt to repay (Baum et al. 2015).6 For individuals ages 25 to 34 with some college but no degree, including those who completed short-term certificates as well as those who left school with no credential, the earnings premium was only 3 percent. Among the older age group, however, the gap between those with some college but no degree and high school graduates was $5,100 (20 percent). These median returns to college suggest that most graduates could repay typical loans under an income-driven plan.

Table 1. Median Earnings of Individuals with Some College or More Relative to High Schoool Graduates

In addition to the different earnings premiums for different credentials, there is considerable variation among people with similar levels of education (Baum 2014). For example, 32 percent of 25- to 34-year-olds with bachelor’s degrees earned at least twice as much as the typical high school graduate in 2014, but 20 percent earned less than the median for high school graduates.7 Earnings vary by occupational choice, but also because of factors entirely out of workers’ control. For example, when the economy is weak—as it notably was during the Great Recession—even college graduates struggle to find good jobs. This variation in post-college outcomes suggests a need for repayment insurance.

A system that automatically links payments to one’s ability to pay will solve the liquidity problems facing many borrowers because of unemployment or low earnings. However, solving the problems of those who have already incurred debt may encourage future students to borrow amounts they are unlikely to be able to repay. A program that assures people they cannot get in over their heads is likely to encourage overborrowing and increase in cost over time.

A properly designed IDR system will protect borrowers against unforeseen difficulties. It will accommodate the variation in incomes associated with the range of returns to postsecondary education and the vicissitudes of the economy. At the same time, it will discourage students from borrowing excessively and will not support institutions charging tuition unreflective of program quality. It will target its subsidies to low- and moderate-income borrowers, rather than to those with relatively high incomes and unusually high levels of debt. Ultimately, IDR should support successful repayment and prevent avoidable defaults.


Factors to Consider in Designing Income-Driven Repayment

The concept that borrowers will repay their loans as a percentage of their incomes over time sounds straightforward. But, as most proposals we reviewed recognize, actually designing a universal IDR plan involves decisions about many details and tradeoffs among the goals of simplicity, relief for struggling borrowers, and equity for both taxpayers and borrowers. These details also have implications for the incentives facing both future borrowers and postsecondary institutions. The most generous provisions are not necessarily best for borrowers in the long run. Our recent report reviews and analyzes the tradeoffs inherent in several detailed proposals (Baum and Johnson 2016). Here, we discuss a few important aspects of program design, focusing on how compatible different options are with principles of sound public policy design.


Required Payments as a Percentage of Income

The Income-Based Repayment (IBR) plan, the major IDR program introduced in 2009 that has been a model for later modifications, only requires payments after borrowers’ incomes exceed 150 percent of the federal poverty level (FPL). It requires payments equal to 15 percent of the income above that level. The Pay As Your Earn (PAYE) program, implemented in 2012, reduced that assessment rate to 10 percent, and many recent proposals for improving IDR maintain this lower rate.8 Though there is not one optimal rate, lowering the assessment rate on income above an exclusion does more to help higher-income than lower-income borrowers, raising both efficiency and equity concerns.


Incorporating and Income Exclusion

The concept of an income exclusion is consistent with the basic motivation for IDR, which is to prevent people from being required to pay more than they can afford. With an exclusion in place, borrowers who have no discretionary income will not be forced to pay their debts instead of putting that money toward basic needs like food and housing. But what is reasonable beyond the income exclusion is a more difficult judgment call.

In a 2006 study, Baum and Schwartz helped to lay the groundwork for IBR and examined multiple approaches to defining an appropriate assessment rate. The study proposed an income exclusion at 150 percent of FPL and argued that the percentage of discretionary income required for all education debt payments—including private and federal loans—should rise with income, but never exceed 18 to 20 percent (Baum and Schwartz 2006). The 15 percent rate incorporated into IBR is consistent with this recommendation. The Obama administration lowered the rate to 10 percent in an effort to further ease the burden on borrowers.9


Alternative Progressive Designs

One assessment rate applied to income above an exclusion yields a progressive structure, with higher-income borrowers paying a higher percentage of their incomes than those with lower incomes. But the effective rates would increase more rapidly if people paid, for example, 10 percent of income between 150 percent and 200 percent of FPL, 12 percent of income between 200 percent and 250 percent of FPL, and 15 percent of income above 250 percent of FPL. Some proposals suggest this type of graduated rate structure, which is similar to the federal income tax system.

An approach that would be harder to reconcile with principles of sound policy design would be using different assessment rates for people with different incomes. Some proposals suggest, for example, 10 percent for households below 300 percent of FPL, and 15 percent for those with higher incomes (Burd et al. 2013; HCM Strategists 2013). This approach would create a cliff effect, a common hazard in policy design causing an arbitrarily sharp change in subsidy at the income cutoff. For a single person under this proposal, an increase in annual income from $35,305 to $35,315 would increase the required monthly payment from $147 to $221.10 Avoiding cliff effects is a basic component of sound public policy design.


Lowering Assessment Rates

Provided that borrowers have the option of making higher payments, lowering the assessment rate can only benefit them. Because of the income exclusion, however, lowering the assessment rate on discretionary income is most helpful to borrowers with higher incomes and higher debts. The change in assessment rates from IBR (15 percent) to PAYE (10 percent) does not help lower-income borrowers because they make little or no payment under either system. Former graduate students, in contrast, can benefit enormously from the lower assessment rate. Since graduate borrowing is limited only by cost of attendance, borrowers with incomes well above $100,000 can make much lower monthly payments under the 10 percent rate than the original 15 percent rate and receive significant loan forgiveness.

Table 2 illustrates how a single individual’s annual income translates into required IDR payments under 15 and 10 percent assessment rates. The last two columns demonstrate that the benefits of reducing the assessment rate from 15 to 10 percent are larger in both dollar terms and as a percentage of discretionary income for individuals with higher incomes. For single individuals with incomes below 150 percent of FPL ($17,655 or less in 2015) the income exclusion reduces discretionary income to zero, so lowering the assessment rate from 15 to 10 percent has no effect. This holds for a significant percentage of borrowers. In 2014, 30 percent of individuals ages 25 to 34 with at least some college had incomes below $17,500, including those with no income.11 Among adults age 18 and older, 39 percent had incomes this low.12 Lowering the repayment rate, therefore, benefits those who are significantly above the bottom of the income distribution.

Using a low assessment rate on income above an exclusion is not a good strategy for helping struggling borrowers. It raises questions of equity and targeting, since the ostensible goal is to help borrowers whose incomes do not reflect typical payoffs to their education. It also raises efficiency problems. Providing significant benefits to borrowers with large amounts of debt may encourage students to borrow more, knowing that the taxpayers will bear a significant part of the risk.

Table 2. Effect of Lowering the Assessment Rate on Payment as a Percentage of Monthly Income

Selecting the appropriate percentage of discretionary income required for repayment is intrinsically tied to the question of an income exclusion and the question of whether and after how many years remaining balances will be forgiven. A low assessment rate combined with a short period before forgiveness will lead to an expensive program that heavily subsidizes borrowers. Policymakers can avoid this by choosing assessment rates high enough and income exclusions low enough that most borrowers will fully repay their loans.


Forgiving Unpaid Loan Balances

Existing IDR programs in the United States, unlike those in most other countries, forgive remaining debt after borrowers have made payments for a specified number of years.13 Under IBR, the required time is 25 years; PAYE reduced that time to 20 years. The latest IDR program, REPAYE (Revised Pay as You Earn), will forgive unpaid debt after 20 years for individuals with only undergraduate debt, and after 25 years for those with any graduate debt.

Debt forgiveness is a popular idea. Images of borrowers struggling under the burden of student debt while they are trying to help their children go to college or even entering retirement are not appealing. And it seems unlikely that many people who have been unable to repay their debts after 20 or 25 years will manage to do so later.


Costs of Generous Loan Forgiveness

Most IDR proposals include provisions for debt forgiveness, but some do not. Loan forgiveness is not cheap. The US Department of Education estimates that one-quarter of borrowers in IBR will receive loan forgiveness, worth an average of $41,000 on original loan balances averaging $39,500.14 The president’s fiscal year 2016 budget estimates the subsidy rate on an unsubsidized Stafford loan issued in 2016 and repaid through IDR at 14 percent, while a loan repaid through the standard repayment plan has a subsidy rate of −24 percent (Office of Management and Budget 2015, 378). One estimate suggests that that forgiving unpaid balances would account for about half of the cost to the government of a universal IDR plan (Akers and Chingos 2014). The authors argue that the cost of loan forgiveness is unjustified, because the goal of the plan—making monthly payments affordable—does not depend on this provision. Moreover, loan forgiveness provides incentives for students to borrow more and institutions to charge more.

Because forgiveness depends on the combination of debt and income, it is currently available to those with low debts if they have low enough incomes, and it is available to those with high incomes if they have high enough debts. Table 3 provides examples of high-, medium-, and low-income borrowers reaching loan forgiveness. The lower-income borrower reaches 20-year forgiveness with just $10,000 of debt, and the highest-income borrower reaches forgiveness by entering repayment with $150,000 of debt. In this example, the cost of forgiving the higher-income borrower’s remaining debt ($12,639) is more than twice the cost of forgiving the lower-income borrower’s remaining debt ($4,974).

Table 3. Debt Forgiveness across Incomes


Reducing the Generosity of Forgiveness

The REPAYE plan is meant to address the high cost and harmful incentives of forgiveness by increasing the time to forgiveness for those holding graduate debt. But dividing borrowers into separate categories for different treatment is problematic, creating a cliff effect analogous to the one described above for assessment rates. The REPAYE structure will create inequities for borrowers who have significant undergraduate debt but borrow small amounts for graduate school. Though this may not be a common situation, a student who starts graduate school and borrows a few thousand dollars on top of considerable undergraduate debt could increase her total payment amount by much more than the amount she borrowed for graduate school.

Setting a debt threshold beyond which a longer repayment period would apply would create a cliff that would likely affect even more borrowers. An extra dollar of undergraduate or graduate debt could extend the repayment period five years, so a borrower could end up repaying thousands of dollars more than someone with a similar income path and just a few dollars less in debt.

A better way to reduce the costs and unintended consequences of generous loan forgiveness would be increasing the time to forgiveness in small steps for increasing debt levels. This would reduce the incentive for students to borrow more, because additional debt would always increase the total amount they would have to repay.

It would also be reasonable to place a limit on the amount of debt that can be forgiven. Borrowers with larger debts would not lose the benefit of loan forgiveness, but there would be a clear limit on how much anyone could benefit from this provision. This approach would best be accompanied by imposing limits on borrowing under the Grad PLUS program, which allows graduate and professional students to borrow up to the full cost of attendance (including living expenses) less other financial aid received.15

Another approach is to cap interest accumulation, forgiving at least part of the interest due that is not covered by required payments. The new REPAYE plan forgives half of any interest exceeding the required monthly payment as it becomes due. This approach addresses the problem of debt levels that balloon while borrowers’ incomes are low, but could end up paying interest for well-off borrowers whose incomes are only low for a short time.

There are strong arguments on both sides of the loan forgiveness debate. But cautions about the potential expense of this approach and the targeting of the benefits are critical, since the incentives for overborrowing and for raising tuition levels in the face of unlimited forgiveness could be strong. Moreover, forgiving debt after minimal payments creates a situation where many people with different debt loads but similar incomes repay the same amounts.

The appropriate design will protect borrowers with lifetime incomes too low to support their debts in the long run without directing significant subsidies toward those who are relatively well off but have accumulated very large debts. Loan forgiveness should not become a widespread grant program covering a significant portion of college costs.16


Key Provisions of an Income-Driven Repayment Plan

Appropriate assessment rates and a sound approach to forgiving unpaid balances are critical components of an income-driven repayment plan. A number of other elements could also either improve the equity and efficiency of the program or generate problems in terms of how the benefits are targeted and unintended consequences. Those consequences are likely to include incentives for overborrowing and significant transfers from taxpayers to borrowers whose circumstances the program is not designed to address.

Our analysis of a wide array of proposals, with a focus on the targeting of benefits and of unintended consequences leading to inequities and to undesirable incentives, leads to the following conclusions:

  1. Enrollment in IDR should be universal and automatic, whether or not there is an alternative payment plan available. Making IDR the automatic or the only repayment option would simplify the process for borrowers and ensure access to all who may benefit. Borrowers should receive clear information on the implications of speeding up or slowing down repayment, and those who prefer to pay their debts off more quickly should have that option.
  2. Automatic payroll withholding could simplify the system and improve efficiency. In an IDR system, the federal government must be able to quickly and accurately adjust payments when incomes change. Current procedures for income verification are burdensome for borrowers and many fall out of the program as a result. Collecting income-driven payments using employer-withholding procedures already in place for income, Social Security, and Medicare taxes would also address delinquency among those who neglect their student loans despite having the income to pay. The system should be designed to minimize the burden on the Internal Revenue Service, which is operating under significant resource constraints.
  3. There are pros and cons to forgiving outstanding balances after a specified period of repayment. If forgiveness is incorporated into a universal IDR plan, policymakers should choose assessment rates high enough and income exclusions low enough that most borrowers would be expected to fully repay their loans. Forgiveness should safeguard against unexpected difficulties and not provide incentive for overborrowing. In general, students who borrow more should repay more than those with similar incomes who borrowed less. Designs that incorporate sharp cutoffs between groups are not a fair or effective way to limit the cost of forgiveness. A better design might instead limit the amount of debt forgiven. An alternative to general loan forgiveness is to forgive unpaid interest beyond some limit, preventing balances from growing too much while borrowers’ incomes are insufficient to support repayment.
  4. Interest rates, income exclusions, required payment levels, and time to forgiveness affect borrowers with different income patterns over time differently. The extent to which borrowers benefit from IDR depends on their ratios of debt to income, but both the level and the timing of income are significant. Income exclusions are particularly important to lower-income borrowers; lower interest rates have the most impact on borrowers with larger debts; the treatment of unpaid interest most affects borrowers who struggle early in repayment but fully repay later.
  5. Making payments affordable is not the same as reducing total dollars paid. A concern with making IDR universal is that lengthening the repayment period increases the total dollar amount repaid. However, this occurs only for borrowers for whom repaying sooner would require unaffordable payments. Moreover, the total amount paid must be discounted to account for both inflation and the value of holding money in the present. Depending on interest rates, borrowers could actually save or make money by extending their repayment period despite paying more dollars in total interest.
  6. The plan must work for older borrowers as well as young adults. About 40 percent of postsecondary students are age 25 or older.17 IDR reduces repayment difficulties for those whose incomes do not support their debt, but since older students have less time to repay before retirement, forgiveness may be a particularly important safeguard for these borrowers. Income exclusions based on poverty guidelines, which take family size into account, may also be more significant for older borrowers, who are more likely to support dependents.
  7. Hold institutions accountable. Many proposals call for greater institutional accountability for improving student outcomes. Imposing financial consequences for their students’ inability to make loan payments has the potential to induce institutions to behave more responsibly.



The broad consensus that the federal student loan system should automatically place borrowers in an income-driven repayment plan is consistent with a simpler student aid system. This approach would create a stronger insurance plan for borrowers whose unforeseen circumstances lead to severe difficulties with debt repayment.

The design of the program involves many judgment calls. The policy should be simple and easy to understand. Reviewing the detailed proposals for reform makes it clear how challenging this goal is. Even reformers committed to simplifying the system end up advocating multiple provisions that are difficult to understand and would create a complex array of rules and regulations that vary with the particular circumstances of the borrower, the loans, and the institution.

As our review of these proposals reveals, alternative program designs would lead to different outcomes for borrowers and for taxpayers. There is not one right answer to the question of which design would be most equitable and efficient, but several important guidelines do emerge from our analysis.

The program should aim to prevent borrowers from facing unmanageable payments, allow flexibility in the timing of payments to accommodate varying income paths, and protect borrowers from unanticipated circumstances out of their control. The goal should not be to minimize payments for as many borrowers as possible. The program design should ensure that most borrowers eventually repay their debts and that incentives for institutions to overcharge and for students to overborrow are minimized.

For further discussion and analysis on the policy proposals discussed here, please see the full report (Baum and Johnson 2016). A reference table comparing proposed designs for reforming education tax credit and deduction is available here.



  1. Federal Student Aid Portfolio Summary,” US Department of Education, Federal Student Aid, Federal Student Loan Portfolio, accessed January 29, 2016.
  2. "Income-Driven Plans," US Department of Education, Federal Student Aid, accessed January 29, 2016.
  3. Before July 2010, students could borrow federally guaranteed loans through the Federal Family Education Loan program in addition to federal direct loans. “Loans,” Federal Student Aid, US Department of Education, accessed December 21, 2015.
  4. Income-driven plans vary in their eligibility requirements. For example, Income-Based Repayment (IBR) and Pay As You Earn (PAYE) are only available to borrowers with sufficiently high debt-to-income ratios. “Repayment Plans,” Federal Student Aid, US Department of Education, accessed December 15, 2015.
  5. Consequences of delinquency and default on student loans include ineligibility for further aid, damage to credit ratings, tax offsets, wage garnishment, and Social Security garnishment. See Cunningham and Kienzl (2011) or “Understanding Default,” Federal Student Aid, US Department of Education, accessed December 15, 2015.
  6. Associate degree recipients who graduate from for-profit institutions are much more likely to take on debt than those who graduate from public two-year institutions (88 versus 41 percent).
  7. US Census Bureau (2015), Current Population Survey, Table PINC-03.
  8. US Department of Education, “Education Department Launches ‘Pay As You Earn’ Student Loan Repayment Plan,” news release, December 21, 2012.
  9. Congress passed this change to IDR as part of the Affordable Care Act in 2010, along with a shorter 20-year period to forgiveness. That legislation set implementation for 2014, but the Obama administration used its authority to speed implementation, making PAYE plans available to certain borrowers at the end of 2012.
  10. With income just below 300 percent of FPL ($35,310), the borrower pays 10 percent of her $17,650 income above 150 percent of FPL. With income just above 300 percent of FPL, the borrower pays 15 percent her $17,660 income above 150 percent of FPL.
  11. US Census Bureau (2015), Current Population Survey, Table PINC-03.
  12. US Census Bureau (2015), Current Population Survey, Table PINC-02.
  13. The UK forgives debt after 20 years. New Zealand and Australia do not include provisions for loan forgiveness in their IDR plans (Lochner and Monge-Naranjo 2015).
  14. Federal Perkins Loan Program, Federal Family Education Loan Program, and William D. Ford Federal Direct Loan Program; Notice of Proposed Rulemaking,” 77 Fed. Reg. 137 (July17, 2012), p. 42122.
  15. The Direct PLUS program is for graduate and professional students (Grad PLUS) as well as parents of dependent students (Parent PLUS). The institution determines the cost of attendance. See “PLUS Loans,” Federal Student Aid, US Department of Education, accessed December 16, 2015.
  16. See James and Kelly (2015) for further discussion on the moral hazard of loan forgiveness and Delisle and Holt (2012) for more detail on potential costs of current forgiveness provisions for high-income, high-debt borrowers.
  17. NCES Education Digest 2014, Table 202.50.



Akers, Beth, and Matthew M. Chingos. 2014. Student Loan Safety Nets: Estimating the Costs and Benefits of Income-Based Repayment. Washington, DC: Brookings Institution.

Baum, Sandy. 2014. Higher Education Earnings Premium: Value, Variation, and Trends. Washington, DC: Urban Institute.

Baum, Sandy, and Martha Johnson. 2016. Strengthening Federal Student Aid: An Assessment of Proposals for Reforming Federal Student Loan Repayments and Federal Education Tax Benefits. Washington, DC: Urban Institute.

Baum, Sandy, Jennifer Ma, Matea Pender, and D’Wayne Bell. 2015. Trends in Student Aid 2015. New York: College Board.

Baum, Sandy, and Saul Schwartz. 2006. “How Much Debt is Too Much? Defining Benchmarks for Manageable Student Debt.” New York: College Board.

Boatman, Angela, Brent Evans, and Adela Soliz. 2014. Applying the Lessons of Behavioral Economics to Improve the Federal Student Loan Programs: Six Policy Recommendations. Indianapolis, IN: Lumina Foundation.

Burd, Stephen, Kevin Carey, Jason Delisle, Rachel Fishman, Alex Holt, Amy Laitinen, and Clare McCann. 2013. Rebalancing Resources and Incentives in Federal Student Aid. Washington, DC: New America Foundation.

Cunningham, Alisa F., and Gregory S. Kienzl. 2011. Delinquency: The Untold Story of Student Loan Borrowing. Washington, DC: Institute for Higher Education Policy.

Delisle, Jason, and Alexander Holt. 2012. Safety Net or Windfall: Examining Changes to Income-Based Repayment for Federal Student Loans. Washington, DC: New America Foundation.

Dynarski, Susan, and Daniel Kreisman. 2013. Loans for Educational Opportunity: Making Borrowing Work for Today’s Students. Washington, DC: Hamilton Project, Brookings Institution.

HCM Strategists. 2013. Doing Better for More Students: Putting Student Outcomes at the Center of Federal Financial Aid. Washington, DC: HCM Strategists.

Hershbein, Brad, Benjamin Harris, and Melissa Kearney. 2014. “Major Decisions: Graduates’ Earnings Growth and Debt Repayment,” Washington, DC: Hamilton Project, Brookings Institution.

James, Kevin J., and Andrew P. Kelly. 2015. “Balancing Risk and Responsibility: Reforming Student Loan Repayment.” Washington, DC: American Enterprise Institute Center on Education Reform.

Lochner, Lance, and Alexander Monge-Naranjo. 2015. “Student Loans and Repayment: Theory, Evidence, and Policy.” NBER Working Paper 20849. Washington, DC: National Bureau of Economic Research.

Office of Management and Budget. 2015. The Appendix, Budget of the United States Government, Fiscal Year 2016. Washington, DC: Office of Management and Budget.


About the Authors

Sandy Baum is a senior fellow in the Income and Benefits Policy Center at the Urban Institute and Professor Emerita of Economics at Skidmore College. She has written and spoken extensively on college access, college pricing, student aid policy, student debt, affordability, and other aspects of higher education finance. She coauthors the College Board’s annual reports, Trends in Student Aid and Trends in College Pricing.

Martha Johnson is a research associate in the Income and Benefits Policy Center at the Urban Institute, where she supports policy analyses and program evaluations relating to higher education and the social safety net.



This brief was funded by the Bill & Melinda Gates Foundation as a part of its Reimagining Aid Design and Delivery project. We are grateful to them and to all our funders, who make it possible for Urban to advance its mission.

The views expressed are those of the authors and should not be attributed to the Urban Institute, its trustees, or its funders. Funders do not determine our research findings or the insights and recommendations of our experts. Further information on the Urban Institute’s funding principles is available at

About the Urban Institute

Cross-Center Initiative

Cross-Center Initiative: 
To reuse content from Urban Institute, visit, search for the publications, choose from a list of licenses, and complete the transaction.

LATEST IN Education and Training

To reuse content from Urban Institute, visit, search for the publications, choose from a list of licenses, and complete the transaction.