The Biden administration recently announced that it would forgive approximately $500 billion in student debt. Up to $20,000 of debt will be forgiven for Pell Grant recipients and $10,000 for other borrowers. Borrowers whose income was less than $125,000 ($250,000 if married) in 2020 or 2021 are eligible.
In its brochure, the White House explains its goals. He says he wants to target the relief to low- and moderate-income borrowers, especially black borrowers, borrowers who haven’t earned a degree and those who have defaulted on their student loans. There is no doubt that these borrowers struggle the most with their loans and represent the most vulnerable group of borrowers.
Therefore, it is important to understand how well the debt forgiveness policy is aimed at providing relief to these groups, both to assess whether it is achieving stated goals and to inform the design of future changes to grant and loan programs that are intended to help low- and middle-income groups pay for college. What elements of debt forgiveness policies have benefited low-income borrowers, black borrowers, non-degree borrowers, and delinquent borrowers? How much did the relief cost (in terms of the federal budget) for Pell and non-Pell borrowers? How much of the total aid budget was spent on each group?
To answer these questions, I rely on publicly available data from the Ministry of Education (2004/2009 Post-Secondary Enrollment Survey), which asks a representative sample of students who started school in 2004 about their demographic and economic background, their educational attainment, and use of financial aid (including federal loans) over the next 12 years. It is probably the best available survey of student loan borrowers relevant to understanding recent debt relief policies, although it does not include information on borrowers’ post-college earnings.
Because the BPS data include information on receipt of Pell Grants and federal student loans, they offer insight into one of the most important elements of debt relief policy design: distinguishing between Pell Grant recipients and non-Pell Grant borrowers and providing more relief to the former than to the latter.
Pell is a means-tested program. Eligibility is determined by a strict income and means test or participation in means-tested programs while students are undergraduates. This means that almost all Pell recipients are from families that make less than $60,000 a year, and most make much less. Black students are twice as likely to qualify for Pell grants as white students. Therefore, Pell eligibility is a very effective way to target financial aid to disadvantaged groups. Indeed, technocrats in Washington (myself included) have argued that concentrating student loan debt forgiveness on Pell Grant recipients would be an efficient and coherent way to provide aid to those who need it most.
Unlike Pell Grants, federal student loans are not means-tested. Students of any income can qualify. Students must complete the Free Application for Federal Student Aid (FAFSA). If their income or assets are too high (or if they are a graduate student), they are not eligible for a Pell Grant, but are still eligible for student loans. In other words, by definition, borrowers who have federal student loans but never received a Pell Grant were either too wealthy to qualify for Pell Grants as students or chose not to apply for financial aid.
The table below uses BPS data and information from a White House fact sheet to compare the amount of benefits, my cost estimates, and the benefits and characteristics of borrowers in each of the two groups.
In total, the policy relieves about $500 billion in debt. Of that, 72% (or $360 billion) was owed by Pell Grant recipients. While Pell Grant recipients could receive up to $20,000 in aid, the actual average amount of relief they received was about $13,300 because many of them owe less than $20,000 (and many less than $10,000).
Moreover, the $360 billion in forgiveness for Pell recipients didn’t cost the government $360 billion because most of those loans will never be paid in full (in present value terms, which is how loans are calculated for federal budget purposes). More than a third of Pell grant recipients, for example, defaulted on their loans within 12 years of enrolling in school; they represent about 90 percent of all defaulting borrowers. Using borrower payment data, Catherine and Yannelis estimate that the lowest-earning 40% of borrowers will repay (in net present value) only about $0.45 of every $1 of debt. In the table, I more conservatively assume that Pell Grant borrowers will repay $0.70 per dollar. In that case, the budget cost would be around 250 billion dollars. On a per-borrower basis, that means the policy costs about $9,260 on average, more than the 27 million Pell recipients expected to benefit.
As a stand-alone policy, the relief provided to Pell recipients involved modest budgetary costs, and its effect on inflation is likely to be small. It is cost-effective in the sense that it enables significant debt reduction per dollar of budget expenditure.
Conversely, borrowers who did not receive Pell Grants rarely default, have higher loan repayment rates, and pay more interest because they owe unsubsidized undergraduate and graduate loans that have the highest interest rates (and are therefore expected to repay more than 1 USD for every $1 of debt they owe). For these reasons, the cost of forgiving $140 billion of their loans is likely to be much closer to $140 billion. On a per-borrower basis, that means the cost of providing $10,000 in assistance to the 16 million borrowers in this group was about $8,750.
In other words, the administration spent about the same amount per borrower on Pell Grant recipients as it did on other borrowers, even though the Pell Grant recipients are from much more disadvantaged backgrounds.
How disadvantaged? The table describes the characteristics of the two groups. Pell Grant recipients are more likely to be poor, have poorer educational outcomes, are more likely to be black, and come from less educated families compared to other college and graduate students and the broader US population. For example, Pell Grant recipients were 23% black and 17% Hispanic. Only 23% obtained a diploma by 2009 (five years after first enrollment); 55% dropped out without obtaining any degree or credential.
They are disproportionately from poor families: 42% of dependent parents of Pell Grant recipients were in the bottom 25% of the income distribution when they entered college, and almost none are in the top quintile. Few of their parents had either a bachelor’s degree (28%) or a bachelor’s degree (11%). They are a group in a significantly disadvantaged position.
By contrast, borrowers who have never received a Pell are relatively affluent compared to Pell recipients and the average American household. Only 6% are black and 7% Hispanic; 83% are Caucasian or Asian. Their parents are well-educated and affluent: 55% have a bachelor’s degree and 29% have a bachelor’s degree (making them twice as likely to have a bachelor’s degree than the average American). 38% grew in the top 25% of the income distribution, and only 3% at the bottom. And they are much more likely to have received a BA. Nearly 60% completed a BA degree after five years of enrollment (which is above the average for US college students). Overall, 70% have earned some type of credential, including an associate’s degree or certificate.
One takeaway from the table is that the total dollar amount for debt relief may be a misleading indicator of how much is being spent per borrower. The $20,000 in relief given to Pell recipients costs about the same amount, per borrower, as the $10,000 in relief given to the non-Pell group because of differences in how much Pell students owe and their expected ability to repay.
Another key finding is how effective the means test is in identifying and targeting disadvantaged groups. In fact, 89% of all black borrowers and 84% of Hispanic borrowers received a Pell Grant. Pell borrowers represent 90% of borrowers in default. And 79% of all school leavers were Pell Grant recipients. If the goal was to help these specific groups, why not spend all the money on them, instead of spending the same amount on non-Pell loan recipients as Pell recipients?
Looking ahead, Biden proposed a significant increase to the Pell Grant, which would increase the number of middle-class borrowers who qualify (because of the way the grant is calculated) and lower Pell recipients’ out-of-pocket costs for college (and reduce or eliminate their need for a loan). As the table above suggests, that aid is well-targeted at students who don’t have the family resources to afford college upfront or the ability to pay it off later.
At the same time, Biden also proposed expanding the generosity of income-based repayment plans for future borrowers so that many of them will pay back less than they borrowed. Many of the benefits and much of the cost of that plan will go to borrowers who did not receive a Pell Grant. Like untargeted debt relief policies, this relief will not be progressive, will cost more, and will benefit advantaged groups than other major spending programs.
 The table reflects the number of borrowers in each bracket the White House expects to benefit, including the effect of the $125,000/$250,000 income limit. However, because the BPS data do not include borrowers’ current income, I did not adjust for borrower characteristics to account for the potential effect of income restrictions. Because the income limit is high and few borrowers are affected, it is not clear that this would significantly change the demographics within each group.
The Brookings Institution is funded through the support of a diverse array of foundations, corporations, governments, individuals, and foundations. A list of donors can be found in our annual reports published here. The findings, interpretations and conclusions in this report are solely those of the authors and are not influenced by any donations.