Despite following the student debt debate closely and writing about it repeatedly, I find the whole conversation hard to track. This is mostly because very few advocates for the various plans are clear on why they want to cancel the debt they want to cancel. For example, a consensus seems in the process of being manufactured around a choice between $10,000 of debt relief per borrower (or for a subset), and $50,000 per borrower (or for a subset). Both NPR and Data for Progress have recently framed the issue this way, completely omitting the possibility of cancelling all student debt—despite proposals for full cancellation being in both the House and the Senate dating back to 2019. These and other discussions of debt cancellation tend to focus on the relative distribution effects of each plan (e.g., arguing about how different levels of cancellation will affect the racial wealth gap), but rarely address the different moral arguments for debt cancellation and how those factor in.

I think these moral arguments are important. The moral principles underlying student debt arguments won’t necessarily determine which plan is preferrable—there are various factors and constraints that could come into play for each position. But understanding these principles and how they interact helps clear up some of the muddiest arguments. Reflecting on our own principles can clarify our own thinking, and identifying opponents’ principles allows us to more effectively respond to their arguments. 

What follows are four reasons one might want to cancel student debt, and some thoughts on the consequences of each approach. I’ve done my best to give each goal its best possible argument, to carefully consider what appropriate relief might look like, and to give due consideration to what constraints might affect that relief. Compare them for yourself and see which is most appealing.

Student Debt Cancellation for Justice

1. The Argument

The most common formulation of this argument is: education is a public good and it is our collective social responsibility to provide it. When individual students must pay for their education, only those with money become educated, and we all suffer. For example, before public high school existed, the cost of education beyond elementary school was the responsibility of individual students and their families, and very few people went to high school (around 14 percent in 1910). The high school movement recognized that secondary education was a collective responsibility and shifted the financial burden of high school from individual students to the broader community. As a result, high school enrollment and graduation soared, immense economic benefits followed, and free public high school is now seen by most as an integral part of American society.

Similarly, in the early 1900s, paying for college was the responsibility of each individual student. As a result, very few people went to college and they were almost exclusively wealthy. But instead of following the example of the elementary and high school movements, reformers in the 1960s believed they could increase the number of people in college without shifting the financial burden off of individual students. The answer they devised was government-backed student debt. By allowing students to pay their college costs over time, they could provide less wealthy students with higher education but avoid the difficult political task of socializing the cost of it.

This was a mistake. The student debt system treated a symptom (lack of education for less wealthy students) but not the cause (the financial burden of education on students). As a result, college enrollment among less wealthy students increased, but less dramatically than did high school enrollment when it was made free. Generations of graduates found themselves tied down with large amounts of debt and predatory interest rates while their wealthy classmates graduated free and clear. As with any extension of credit, financing higher education made it less expensive for non-wealthy students on the front end, but more expensive for those students on the back end.

The argument that debt cancellation is a form of justice is something of a reparations argument—an attempt to, as best as we can, right a past wrong. This is my position, as I have written. The entire student debt apparatus is an unjust system. As Raúl Carrillo, Director of Public Money Action and Associate Research Scholar at Yale Law School explained to me recently, student debt functions as a tax on people who want to go to college but can’t afford it, simultaneously hampering those people’s financial lives for decades and disincentivizing others from going to college lest they suffer the same consequences. This is an unfair and untenable situation. Higher education, this argument concludes, should be free. It should always have been free. Student debt is and has always been unjust debt.

2. The Relief

The logical conclusion of this argument is somewhat extreme in terms of relief. If the financial burden of education should not fall on individual students, then every student who paid for college deserves a refund. It always should have been free, or at least there should have been a free option, so everyone who ever paid anything was wronged. When you fully socialize the cost of education, you socialize it for everyone, not just the less wealthy students. This is how elementary and high school work already—many of you, like me, probably went to a public high school with some very wealthy students. And although you might believe it is fair to have some non-free private colleges (I’m not so sure), you cannot rightfully charge people for going to those colleges when they should have had a free option but didn’t. Complete justice would be hitting a big reset button, providing everyone who has ever gone to college with a retroactive refund for all payments made, and doing things right going forward.

This big relief—full refunds for every student—is too big to be practical. There is probably no way to figure out how big it would be given the lack of centralized records, but it is fair to guess that it would be a greater expenditure than the federal government can comfortably make all at once. Given that there are constraints on enacting the full relief that the argument implies, what is the correct level of relief given these constraints? The answer depends on what you understand the constraints to be. 

3. The Constraints

The correct relief to remedy our five and a half decades of unjust education debt imposition is full cancellation of outstanding debt, full cancellation of future debt at public institutions, and refunds for borrowers who had their taxes levied and public benefits garnished for failing to pay their student debt. This is my position because I believe the constraints on full debt cancellation are relatively small, as I explained here, and that people who lost public benefits and tax refunds because of student debt were likely those who needed that money the most, and probably still need it.

Some thoughtful people disagree, and there is logical room to do so. You might think that student debt is unjust, but that the cost of full cancellation, or of cancelling $50,000 per borrower, is too high. (Again, you’d be wrong, see the link above.) If you believe the cost is too high, you might have any number of preferences for how to target the possible relief either to be the most fair or to alleviate the most harm. But, if you believe that student debt is unjust, the only reasonable constraint on how much relief you would give is the cost. (I use the term “cost” broadly. Debt cancellation doesn’t require that the government spend any money, only that it give up some fairly uncertain future revenue. And there are open questions about how much revenue the government needs and how much student debt will ever be paid back.) 

There are no debtors who don’t deserve cancellation. Their income, school choice, outstanding balance, and any other feature you can think of are irrelevant to what they deserve. If you hold this view, everyone deserves full cancellation and more, and the only question up for debate is what is possible. The answer, in my view, is full cancellation at the very least.

2. Student Debt Cancellation for Harm Reduction

1. The Argument

Tens of millions of people have student debt, and a substantial portion of them cannot pay their balances. These people were promised that higher education would open up financial possibilities for them and the debt they took on would be a worthy investment that would pay returns. But given that they are unable to pay the debt, that was clearly incorrect. It’s not that student debt as a concept is unjust, or that everyone has been harmed—the fact that some people are able to pay back their debt demonstrates that they likely got the promised return on their investment in higher education. However, it is clearly not working for everyone. People are suffering, being driven to isolation, depression, and suicidal ideation, and we have a collective responsibility to relieve that suffering to the greatest extent we can.

The version of this argument above is completely focused on suffering. This means there is no question of who deserves cancellation involved (or, conversely, who deserves to be in debt). But this form of reasoning often gets mixed up with various questions of deservingness. For example, sometimes you will hear that we need some cancellation but not for lawyers or dentists, or not for Ivy League graduates. These are not so much arguments that people aren’t suffering, but that they shouldn’t be suffering—maybe because their diploma is an unconditional golden ticket to riches, or just because we don’t find them sympathetic. If your only concern was relieving suffering, it wouldn’t make any difference if someone was a lawyer or a rideshare driver, a community college or a harvard (I would rather not capitalize it, thanks) alum. Dentists from harvard can be unemployed, can work low-paying jobs, can suffer under student debt. But some people see this as a moral failing on their part, and therefore exempt them from student debt relief even when the arguments seem to be otherwise directed toward harm reduction.

A less-discussed wrinkle in the harm reduction argument is how to account for income-driven repayment. Income-driven repayment, or IDR, is the name for a bevy of programs that link debtors’ required monthly payments to their “discretionary” income according to a number of formulas. Thanks to these programs, many debtors—possibly millions—have required monthly payments of $0. These people are still in debt, their debt balances grow each month due to interest, and those debt balances still may prevent them from doing things like buying a home, buying a car on reasonable terms, or even renting an apartment or getting insurance. But there is room to argue that they are not actually suffering because their monthly obligation is “affordable.” My view is that the psychological weight of a debt burden around your neck, especially when combined with the massive financial consequences you suffer if you forget to meet the strict bureaucratic means-testing requirements like correctly recertifying your income every year, is a harm to be taken seriously and one that warrants cancellation. But this could be a point of debate.

2. The Relief

The appropriate relief in the minds of harm reducers will depend both on cost constraints and on what is considered to be harm. Remember, student debt as a system for providing higher education is, in this view, not bad in itself, so people who argue for harm reduction generally won’t support full cancellation. Some borrowers are able to comfortably pay their debt, after all, and that debt is not unjust in this view, so those people are not being harmed. Cancellation of their debt, therefore, would be a waste, no matter how small the cost. (There may be an argument that student debt as a concept is just, but that every dollar of debt currently outstanding is causing unnecessary harm, and that full cancellation is appropriate on harm reduction grounds. I haven’t seen this case made explicitly, but it’s possible.)

The pure harm reduction arguments tend to focus on income for structuring relief. The idea is that people with lower income are more likely to be suffering under their debt, and people with higher income are more likely to be able to pay their debt. 

This is interesting and potentially inconsistent for a few reasons. First, it’s not always true that higher income people are not suffering, and the income cutoff is often linked closely to questions of deservingness. For example, a legal services attorney might have an above-average salary but such a large debt load that repayment is impossible. If income is the only relevant measure, this person might not get relief despite their suffering. This might be an accepted trade-off of needing to find some line to draw and not having a better option than income—there are always edge cases and no means-testing is perfect—but it is often a moral judgment that an attorney with that much debt should be making more money. 

(A quick note on incentives: it might be a social good for attorneys and doctors and other professionals to take jobs that pay less, because those jobs tend to be more broadly beneficial than the jobs that pay more. For example, you might consider an attorney doing legal services and making $60,000 per year to be doing more social good than an attorney making $250,000 per year structuring corporate mergers or suing one investment bank on behalf of another investment bank, or even defending a large employer against a race discrimination lawsuit brought by the first attorney. Similarly you might want doctors to choose lower-paying jobs in less served areas instead of all flocking to the fanciest hospitals and cities. If you do think this, then you don’t want a student debt system that morally judges the lower-paid attorneys and doctors for not making enough money, and deprives them of cancellation to which they might otherwise be entitled. It often seems that people who are mad about highly-educated professionals not making enough to pay their debt maybe haven’t thought all the way through their positions.)

Second, as mentioned above, one might believe that the right measure of suffering is not income on its own, but the ratio of income to required payments. The idea would be that regardless of a debtor’s overall debt burden, the relevant question is just how their income and expenses shake out each month. In this view, a debtor who makes $4,000 per month with a required debt payment of $1,500 might be suffering more as a result of their debt than someone who makes $2,000 per month but only pays $300. And someone whose required payment is $0 per month thanks to income-driven repayment might not be suffering at all, no matter how low their income is. Along the same lines, the most direct line to reduce suffering might not be debt cancellation at all. If you view the monthly payment as the important factor, it might be more effective to expand income-driven repayment than to cancel any debt.

Again, you might not find any of that compelling. I don’t. Having student debt places both financial constraints and administrative burdens on people regardless of their required payment. And as their balances grow while their payment is low or nothing, the stakes for meeting their administrative requirements get ever higher. In year one of $0 payments, failing to recertify your income to a loan servicer’s often arbitrary satisfaction might put you on the hook for monthly payments calculated on a $40,000 balance. After 10 years making $0 payments, failing to recertify your income could put you on the hook for payments calculated on a $68,000 balance (assuming 7 percent interest). Not to mention that whether you certify or not, that $68,000 will be sitting on your credit file, causing issues with everyone from lenders to landlords to employers. And you will know that you now owe $68,000, which might make you somewhat afraid to increase your income and risk having to make payments of any size.

Though it’s imperfect, income is at least a somewhat reasonable measure for relief in this view. But picking an amount to cancel, like $10,000 or $50,000, seems especially arbitrary. Some people might have no income but $200,000 debt burdens—a small bit of cancellation will not significantly affect the harm they suffer from their student debt. Someone else might have high enough income to just barely qualify for cancellation, but have a relatively low and perfectly manageable debt balance. They might not be suffering on a monthly budget basis despite falling into the cancellation category based on income. What would make more sense is cancelling the amount of debt for each person that is causing their suffering.

This may sound more difficult than it is. We already have income-driven repayment formulas that calculate how much debtors should be able to afford each month. And we have decided that the standard term for repaying student debt should be 10 years. So we could use tax data and repayment formulas to determine, based on everyone’s current income, how much they should be able to pay back in the next 10 years, or before they hit 10 years since graduation. If someone’s calculated payment is $0 per month, they can’t afford anything, and their whole balance would be cancelled. If someone’s calculated payment is $300 per month and they graduated this year, their balance would be adjusted down to $36,000 ($300 x 12 months x 10 years) no matter where it started. If you wanted to limit it to 10 years since graduation, a similar person who graduated six years ago would have their balance adjusted to $14,400 ($300 x 12 months x 4 years left to get to 10). Again, this would not be perfect, and would open the floor for arguments about the adequacy of the “discretionary income” formulas. But it would be most consistent with a harm reduction approach.

(I feel the need to reiterate that I do not believe this is an adequate solution. I believe it would leave people burdened by unjust debt and that it is wrong to means-test justice.)

3. The Constraints

As in our first scenario, the main constraint on harm reduction cancellation would be a broad notion of cost. But, as above, this cost is very difficult to determine because we don’t know how much student debt people will pay back. Even limiting the amount of cancellation doesn’t help here: cancelling one person’s $10,000 debt might “cost” a lot if that person would accrue a decade of interest and then get a good job and pay it off, and cancelling another person’s $50,000 might “cost” very little if they would otherwise make low or $0 payments and then qualify for one of the existing cancellation programs (like the death discharge—you qualify by dying).

It might be easiest to calculate the “cost” of cancellation using my proposal above to use the formulas we already have, figure out what people can “afford” based on their current income over 10 years, and cancel the rest. This is because the calculation assumes, with some reason, that every dollar cancelled will not be paid back. If someone is paying $0 we assume they will keep paying $0, and if they’re paying $300 we assume they’ll keep paying $300. Again, this isn’t perfect: people’s incomes change and it will overcount some and undercount others. But it at least tries to approximate how much revenue the government is giving up by cancelling debt. In fact, it cancels only the debt that will likely result in no government revenue. Hence, the “cost” is nothing! (Why hasn’t this been put forward?)

3. Student Debt Cancellation for Redistribution

1. The Argument

Federal student debt is affecting the incomes and wealth of tens of millions of people. Since the federal government owns this debt, it has a lot of flexibility to adjust the debt to achieve its goals. One primary goal for the government is to create a fair distribution of wealth and income. With this in mind, many people focus on the possibility of various forms of redistribution when it comes to student debt cancellation.

The redistribution arguments tend to be hard to tease out from the harm reduction arguments. They will often come hand-in-hand, and it’s hard to sort out just what is important to the person making the argument. The argument might sound something like: 

People in default on their student loans are disproportionately women and people of color, and most people in default have balances under $50,000. Meanwhile, very high debt balances tend to be carried by disproportionately white debtors with graduate degrees. Therefore it is a matter of racial/gender justice to limit cancellation to $50,000 per borrower. 

Another variant is to talk about the various wealth gaps (gender, racial, overall). Many advocates will point out that very high debt balances are held disproportionately (but not exclusively!) by white debtors, and lower balances are held disproportionately by non-white debtors (even though, as a whole, students of color are more likely to have student debt and more likely to have more student debt than their white classmates). As a result, they say, full cancellation would increase the overall racial wealth gap while $10,000 or $50,000 cancellation might lower it. We want to achieve a more just wealth and income balance, including repairing harms that were done on the basis of race and/or gender, and debt cancellation is a tool we can use to do that. Economist Marshall Steinbaum has offered a thoughtful response to this argument on its own terms, but I think there are additional interesting dimensions to discuss. 

(This distribution argument could just as easily be made in terms of redistributing to poorer people in general regardless of race/ethnicity/gender, but I think the race- and gender-related arguments are more common and more interesting.)

People who make this argument also tend to use income cut-offs and debt amounts as tools to accomplish the redistribution they seek. That redistribution, though, is most often described in terms of race, ethnicity, and/or gender. It is a reparations argument dressed as a means-tested program. I suspect there is a practical reason for this—advocates of this view might worry that outright cancelling the debt held only by women and people of color would be illegal somehow, given that our legal system is basically on the level of Prager U when it comes to believing “reverse discrimination” is a meaningful concept. 

Is this worry well-founded? I don’t know. The federal government has a lot of leeway when it comes to cancellation. The normal way one would challenge a discriminatory policy like this would be to sue, but a white borrower would have to explain how they are harmed by cancellation of non-white debt. That might be a hard thing to explain, given that they wouldn’t be harmed by it. In any case, I suspect people avoid the full reparations argument because they have a sense that it would never fly. So instead they talk about income and debt amounts, which can get the reparations motivation twisted up with harm reduction, but there is sufficient overlap to make it all comprehensible at least.

Like the harm reduction argument, the reparations argument in its pure form assumes that the student debt system (at least as a concept) is a fair one. That is, our default view of unjust debt is that it should be cancelled regardless of its distribution effects. For example, when the government goes after a credit card company for charging illegal fees, it doesn’t seek removal of those fees only if doing so will close the racial wealth gap. If something was taken unjustly, it is returned without question. But if the student debt system itself isn’t an injustice, then it’s perfectly reasonable to want to use it to close wealth gaps. In this view, no one who is left in debt has been wronged by that debt—cancellation is a bonus and not a necessity—so we can use it to accomplish whatever social goals we want, and redistribution/reparations is a worthy social goal.

Again, there is space for an argument here that I haven’t quite seen. One could reasonably disagree with my point above about justice. This argument would go: 

Actually we have never just corrected immediate injustices without regard to distribution effects. In fact, all of U.S. history has consisted of failing to correct the theft of entire lives worth of wealth and labor of non-white people and non-men, whether that be through slavery, redlining, uncompensated domestic work, wage gaps, employment discrimination, etc. 

As Dr. Martin Luther King, Jr. put it, America has defaulted on its promissory note. There’s nothing unusual about some people being left holding the bag for others, it’s only that this has only ever gone in one direction. One could plausibly argue, therefore, that even if student debt is unjust, and even if we could cancel it all, it might nonetheless be better to focus cancellation on groups who have suffered historical wrongs, to give them the good check for once and let others get the bounce. 

(Of course, King also wrote that “injustice anywhere is a threat to justice everywhere,” so I’m not sure he would want to leave some people in unjust debt just to make a point.)

2. The Relief

There is not much more to say about the relief than what is above. The arguments here tend to focus on finding the right measures for determining the distribution effects of the various cancellation plans, and there is a lot to argue about there. As with harm reduction, the discourse has coalesced around $10,000 vs. $50,000. But the alternatives we have already discussed, like using the IDR formulas to determine the cancellation amount, may have the same or better distribution effects. 

In fact, you could come up with all kinds of ways to structure cancellation to maximize distribution effects much more precisely than dollar cutoffs. You could cancel the debt of everyone in default, or of everyone who went to a for-profit college—since these both apply disproportionately to women and people of color, these plans would presumably have the desired distribution characteristics. The fact that these arguments are uncommon and/or non-existent suggests again that the redistribution advocates probably have some combination of redistribution and harm reduction in mind, and maybe are just marshaling whatever arguments they can to advance their preferred plan, rather than structuring their preferences around their overall goals.

It is worth noting that as a result of no one making the full-throated reparations argument, none of the mainstream relief plans would actually be full reparations along any lines. That is, because there are debtors of all races, ethnicities, genders, and social classes who have very large debt burdens, none of the plans on the table will, for example, cancel all the debt and only the debt held by Black people. (As it happens, one plan would cancel all the debt held by Black people: full cancellation. But the distribution-oriented people seem to be the most adamantly opposed to full cancellation, often because it will help white people too. More on this below.)

3. The Constraints

The strongest distribution arguments, in my opinion, only hold weight if our capacity to cancel is constrained somehow. That is, even though I think the whole system is unjust, I could potentially support structuring limited cancellation to serve as some form of reparations if that amount of cancellation is all we could do. If we can only achieve limited justice, then distributing that justice to atone for some past harms is a reasonable thing to do.

But often redistribution/reparations advocates make a stronger claim. They will argue that even if we can do full cancellation, we shouldn’t do it, because it would disproportionately help white debtors. I addressed this argument in some detail in my article linked above about the path to free college through debt cancellation. Here it is enough to say that I find this reasoning backwards and baffling. The important feature to understand about student debt is that it does not help anyone for other people to be in student debt. Imagine you and I are both in $100,000 of student debt. There is no material difference for me between the situation where my debt is cancelled and the situation where both of our debts are cancelled. I have no interest in you remaining in debt. Similarly, Black student debtors have no interest in white student debtors remaining in debt. 

The arguments suggest that it helps Black people if white people are in debt because the median wealth gap might be smaller. My view is that this is an empty benefit. If a Black and a white borrower both have negative $50,000 of wealth (i.e. $50,000 of debt), and we cancel the debt of the Black borrower and not the white borrower, then we have improved the overall racial wealth gap. But we haven’t helped the Black borrower more by virtue of that improvement than if we had cancelled everyone’s debt. 

An objection: what if the white debt does benefit Black people? Joe Biden recently said he wanted to limit debt cancellation and instead use that money on early childhood education. Isn’t debt owed to the government a public asset? So isn’t your federal student debt my money?

This is a point at which people get very confused, so I want to be extra clear. Federal student debt is potential future revenue going to the federal government, yes, but it is not some definite amount of money that can be allocated however the government likes. Just as full debt cancellation doesn’t “cost” $1.7 trillion, there’s no way to go and spend uncancelled debt money on another program. Again (and again and again, forever), much or most student debt will never be repaid. That is millions of lifetimes of burden, struggle, and misery for $0 of federal revenue. Lots of debt, no early childhood education, no programs for struggling neighborhoods or minority communities, nothing. Just suffering. 

So while in an abstract sense it may be arguably true that payments to the federal government  are a joint project if not a joint asset (again, see my interview with Raúl Carillo on this and how to think about it from a modern monetary theory perspective), these particular future payments are an empty promise.

4. Student Debt Cancellation for Economic Growth

1. The Argument

There is little question that full student debt cancellation would result in overall economic growth, more prosperity for everyone, and even increased government revenues. Full cancellation is therefore a no-brainer on the macroeconomics alone. 

I haven’t actually seen anyone make this argument directly and on its own, but I don’t see why not. Stimulating growth is the only argument that is ever given for tax breaks, and that argument is often much more spurious. We structure such a tremendous amount of federal policy around the idea of stimulating economic growth—why is it just a side point here?

2. The Relief

The only serious study on this point (linked above) shows maximum economic benefit from full cancellation.

3. The Constraints

Since the whole point of the analysis is that federal revenue would increase as a result of cancellation, not cancelling all the debt is the constraint.

The student debt discourse would be an excellent case study to examine how people choose their preferred policy by some method or another (what sounds reasonable? what sounds radical? what plan their favorite people/politicians like?) and then jam their arguments into whatever shape that policy gives them. But it is nonetheless very helpful to break down all the different possible principles we could be working from. Even if people rarely work directly from those principles, we all still have a sense that these are important considerations, and we all either consciously or unconsciously assign them varying weights. So it may be helpful to recognize when someone else is focusing on harm reduction to narrow in on the points they will find most persuasive. Or, conversely, it may be helpful to highlight all the important principles a harm reduction approach misses.
For me, the big answer here could not be clearer. There is one approach that achieves educational justice, eliminates all the harm caused by student debt, increases the net worth of all people of color, women, and working class people who have student debt, and creates the most economic growth. What’s more, it is within the president’s power to accomplish, and it would likely increase federal revenue in addition to bringing economic benefits for the entire society. That approach is full cancellation of all outstanding student debt.