Medical debt is the largest US debt in collections. Erasing it changed little.

Medical debt is the single largest category of debt Americans owe to collection agencies — an estimated $140 billion as of mid-2020, more than all other kinds of debt in collections combined [s1]. That figure has fueled a wave of programs that buy up and cancel medical debt. But when researchers ran the rigorous test — randomly relieving some people's medical debt and not others' — they found something uncomfortable: on average, erasing the debt did not measurably improve people's finances or their mental health [s2]. The scale of the problem and the disappointing result of the obvious fix are both true, and both worth knowing.

How big the problem is

The $140 billion figure comes from a study published in JAMA in 2021, which analyzed a large, nationally representative sample of consumer credit reports from the credit bureau TransUnion covering January 2009 through June 2020 [s1]. It found that medical debt had become the largest source of debt in collections in the United States, and that roughly 18% of Americans held medical debt in collections [s1]. Because the data predates the pandemic, it reflects ordinary US health care, not a crisis surge.

The study also found the burden was not evenly spread. Medical debt was concentrated in lower-income communities and, notably, was substantially higher in states that had not expanded Medicaid under the Affordable Care Act — a pattern consistent with the idea that broader insurance coverage reduces the medical bills that end up in collections [s1]. That geographic split is one of the clearest signals in the data that medical debt is downstream of how people are, or are not, insured.

Association is not causation

It is tempting to move directly from "people with medical debt are worse off" to "medical debt makes people worse off." Plenty of studies document that people carrying medical debt report more forgone care, more financial stress, and worse mental health. But those are associations, and they run into an obvious confound: the people who accumulate large medical debts are, on average, sicker, poorer, and more precarious to begin with. Their worse outcomes could be caused by the underlying illness and hardship rather than by the debt itself. Distinguishing the two requires an experiment — actually removing the debt from some people and comparing them with otherwise similar people whose debt was left in place.

The experiment, and its surprising result

That is what a team of economists did, partnering with the charity RIP Medical Debt to run two randomized experiments. They relieved medical debt with a face value of $169 million for 83,401 people between 2018 and 2020, chosen at random from a larger pool, and tracked outcomes using credit reports, collections data, and a detailed survey [s2].

The results, reported in a 2024 National Bureau of Economic Research working paper, were close to null. Debt relief had no measurable effect on credit access, credit utilization, or financial distress on average [s2]. It produced a moderate, statistically significant reduction in how much of their existing medical bills people paid — that is, relieved people paid down somewhat less of their remaining medical debt — and it had no effect on mental health on average, with the pre-registered analysis even finding detrimental effects for some subgroups [s2]. In plain terms: canceling the debt did not lift the financial or psychological weight the way almost everyone, including the researchers, expected it to.

This is exactly the kind of finding that gets buried, because it is unsatisfying and cuts against a popular remedy. It deserves to be reported plainly. It does not mean medical debt is harmless or that relief is pointless for the individuals freed of a specific obligation. What it means is narrower and more precise: buying up already-defaulted medical debt, after the fact, does not appear to be an effective lever for improving the average recipient's financial security or health — likely because by the time a bill is in collections, much of the damage (to credit, to care decisions, to stress) has already been done, and canceling the paper obligation does not reverse it.

What the evidence supports

Two things hold at once. Medical debt is genuinely enormous in aggregate — the largest debt in US collections, touching nearly one in five Americans, and concentrated where insurance coverage is thinnest [s1]. And the most rigorous test of relieving it found little average benefit, which points away from downstream debt forgiveness and toward preventing the bills — through insurance coverage and limits on how aggressively they are pursued — as the more promising place to intervene [s2]. The uncomfortable synthesis is that the size of the medical-debt problem and the ineffectiveness of the most intuitive solution are both well supported by evidence, and any honest account has to carry both.