Ask a group of higher-education policy wonks about the best way to address our nation’s growing student loan problem, and you’re likely to find one solution overlaps party or ideology: Enroll more borrowers in income-driven loan repayment.

IDR, as it’s commonly known, has existed in some form at the federal level since the 1990s and provides a simple but elegant solution for student borrowers who increasingly come out of school with high debts and low incomes. Rather than making lump-sum payments for 10 years, payments rise and fall with a borrower’s annual income, theoretically ensuring that student loans will never take up an unreasonable portion of anyone’s monthly budget. If college does not result in the economic payoff a student or family expects, loans are forgiven after 20 or 25 years.

Since the creation of IDR plans, and because of aggressive pushes by the Obama administration that resulted in the Pay As You Earn and Revised Pay As You Earn plans, most federal borrowers are eligible for monthly payments equaling 10 percent of their income above a poverty threshold.

IDR has long been a bipartisan darling — President George W. Bush signed a 2007 law creating the precursor to today’s plans, and one of the few higher ed-focused proposals pushed by Donald Trump on the campaign trail involved limiting payments to 12.5 percent of income and forgiving loans after 15 years.

But for all the bipartisan support and expansion of IDRs during the past 10 years, advocates and researchers remain befuddled by a series of questions. In a world in which all borrowers, in theory, can make payments aligned with their income, and in which borrowers below the poverty line could have payments of zero, why has the percentage of borrowers at least 90 days late on payments remained nearly unchanged for six years? Why are there still 8 million borrowers in default on their student loans? Why have we been unable to prevent debt from ruining the lives of a substantial portion of our students of color?

Some have chalked it up to complexity. The sheer number of plans, potential errors by loan servicers, and the need for borrowers to opt in and continually verify their income might explain why borrowers fall behind when they might be able to make much lower monthly payments.

For these reasons, you often hear commentators and experts say that rather than a student loan crisis, we have a “repayment crisis,” solvable through better consumer information or automatic withholding.

While there is almost certainly truth to these critiques, and some value in the solutions, there’s one overlooked flaw of income-driven loan repayment. It prevents IDR from being a silver bullet and could put our country’s borrowers of color even further behind.

It’s right there in the name.

Obviously, income-driven loan repayment considers only a borrower’s income when calculating a reasonable payment. It reflects what you make, not what you own.

In a country in which white families have $10 in wealth for every $1 held by black or Latino households, this can matter a great deal.

Before going further, it’s important to note that there are good reasons why we would not want student loan repayment to consider the wealth a borrower has in a home or retirement account. After all, we don’t want someone feeling compelled to sell their home or withdraw from a 401(k) to pay off education debt.

But as long as we tout policies that are income-sensitive but wealth-blind, we ignore the fact that households with seemingly similar incomes can have substantially different resources to pay off loans.

Which means that, for some middle-income, low-wealth borrowers, manageable monthly payments are anything but.

Looking at data from the 2016 Survey of Consumer Finances, we see that the median net worth for younger white households (those headed by someone age 44 and younger) with student loans is $15,400.

The net worth for young black households with student debt? Negative $2,180.

Even if we exclude housing wealth, disparities are stark. The typical white family making near the median household income (between $50,000 and $60,000) has more than four times the financial assets as black and Latino families earning the same amount.

According to the Department of Education’s Repayment Estimator, a borrower with an adjusted gross income of $50,000 with a $30,000 student loan balance (about the national average for a bachelor’s degree recipient) would owe $266 a month, or just over $3,100 a year, under the Revised Pay As You Earn plan.

In other words, the annual student loan obligation makes up less than 10 percent of the typical financial wealth of a white household but nearly 40 percent of the financial wealth of a typical middle-class black household. What seems like a reasonable monthly payment for some may be a source of anxiety for others.

This may explain why 2.6 percent of college-educated white households report being late on any debt payment, while 11 percent of college-educated black households report being behind on a debt.

In this way, completing a bachelor’s degree does not inoculate black families from struggling with debt — quite the opposite, in fact.

Again, this is a function of a deep-rooted racial bias in how Americans are able to build wealth, and it reflects how this bias endures across generations.

White millennials are more likely to report receiving financial support from family, while black and Latino parents are far more likely to need and expect financial support from their children. White families often transfer wealth to children, and families of color frequently must transfer wealth upward to help relatives.

The goal of streamlining and improving loan repayment options is noble and could be useful in situations in which a lower monthly payment is the difference between on-time payments and falling behind.

But if we care about making sure higher education reduces rather than reflects inequality, we can’t simply tackle the mechanism people use to pay it. We have to reduce the debt itself.

Mark Huelsman is a senior policy analyst at Demos, a nonpartisan public policy organization.