Medical debt and collections in the United States

Abstract

Introduction

Medical debt and collections are common and large, but estimates differ widely.

Methods

We used 2 nationally representative surveys in 2024 that are associated with respondents’ credit records to measure medical debt and medical collections from survey participants directly and compare their responses with the medical collections reported to the credit bureau.

Results

In 2024, 36% of US households had medical debt, 21% had a past-due medical bill, and 23% were paying a medical bill over time to a provider. Medical and dental providers are thus one of the most common sources of credit to households. At the same time, 15% of people were contacted by someone other than their medical provider to collect a medical debt while only 12% had any medical collection on their credit record and 5% had a new collection. Active medical collections only partially overlap with medical collections on credit reports and neither represents all medical debt. We bound the mean total debt in active medical collection for people with such collections as between $2456 and $7931.

Conclusion

These debts are not interchangeable, so future policy and research should distinguish carefully between them.

Medical debt and collections are common and large. In 2024, 36% of US households had medical debt, 21% had a past due medical bill, and 23% were paying a medical bill over time to a provider. Medical and dental providers are thus one of the most common sources of credit to households. Some of this debt ends up in collections; 15% of people were contacted by someone other than their medical provider to collect a medical debt, but only a fraction of collections or debts are reported to credit bureaus. We estimate that medical debt worth $194 billion was in active collection.

Introduction

When households have difficulty paying bills or expenses, half point to medical bills as the primary cause.¹ Recent research suggests that medical debt is common.^2,3 When a household does not pay a medical debt, that debt may be sent to third-party collectors who may report the collection to a credit bureau. Until recently, medical collections were the most common kind of collection on credit reports.^4,5 Previous research has used medical collections reported to credit bureaus to understand health insurance expansions^6-9 and medical debt more broadly,⁴ but it is unclear how medical collections reported to credit bureaus are related to overall medical collections and to medical debt. In addition, many medical collections on credit reports have recently been removed,^10,11 so information on medical debt and collections will need to come from other sources going forward.

This study uses 2 surveys conducted starting in January 2024 (n = 4486) that are associated with the respondents’ credit records to estimate the incidence of medical debt, the amount and incidence of medical debt in collections, the relationship between medical debt in collection and collections reported to credit bureaus, and the demographic distribution of these different debts. We thus add to a vibrant recent literature^2-4,12 by providing comparative estimates of medical debt, medical collections, and medical collections reported to credit bureaus for the same people. Medical debt is common. Overall, 36% (SE, 0.9) of US households had medical debt, broadly defined. This includes 21% that had a past-due medical bill, 23% that were paying off a bill to a provider over time, and 17% that had taken out a loan to pay a medical bill, such as by paying with a credit card and revolving the balance. Medical and dental providers are thus one of the most common sources of credit to US households.

Medical collections are also common and large. In the year up to January 2024, 14.6% (SE, 0.6) of households reported being contacted by someone other than their medical provider to collect a medical debt in the previous year. These “active” medical collections do not correspond well to collections on credit bureau records. In the year up to January 2024, 9.9% of all people had only active collections, 4.7% had active collections and medical collections on their credit record at some point during the year, and an additional 6.8% had a collection on their credit record but no active collections. The major credit bureaus removed certain medical collections in 2023, but we included time before these collections were removed in our analysis. Moreover, we found a similar lack of correspondence between collections reported to a credit bureau and active collections using a similar survey in 2023. Because debt collectors hold only a portion of medical debt and do not report all the debt they do hold to credit bureaus, research using medical collections reported to credit bureaus may not be representative of what happens to medical debt or medical collections, although it may still be informative about financial well-being.⁹ As we show, different kinds of medical debt have different demographic distributions and so may have different effects. Future research will need to more carefully distinguish between debts.

With some assumptions, we calculate bounds on the amount of medical debt in active collection. For people with a collection, the mean most recent collection was $2456. People with health insurance have lower recent collections. Multiplying by the number of bills that people report under collection gives an upper bound on the average total amount under collection of $7931, assuming the most recent collection is representative of other collections, and a lower bound of $2456, assuming collections other than the most recent collection were for $0. Under intermediate assumptions about collections other than the most recent, we estimate mean total active medical collections to be $5069, implying that medical debt worth $194 billion was in active collection. Medical debt and medical collections are common even for people with health insurance or relatively high incomes, although the incidence of medical debt, medical collections, and the amount in collections are all lower among those with full household coverage or high incomes. One reason that medical debt is common for middle-income households is that many are paying off medical debt over time, either to the provider or through a loan from a financial institution. Another reason is that some medical debt represents disputed billing. We show that two-thirds of people who receive bills asking them to pay out-of-pocket believe their insurance should have paid some or all of it.

Data and methods

Data sources

Our primary data sources are 2 surveys mailed in January 2024 and their association with respondents’ credit bureau records. The surveys are part of the annual Making Ends Meet survey series from the Consumer Financial Protection Bureau (CFPB). The surveys sample from the CFPB's Consumer Credit Information Panel (CCIP), a nationally representative 1-in-50 sample of credit records maintained by 1 of the 3 nationwide consumer reporting agencies, which contains approximately 6 million consumer records each quarter. The surveys oversampled people who had medical collections on their credit record and who were likely to be experiencing financial difficulty.

The surveys consist of a main survey and follow-up survey to respondents from the previous year, both of which contained the same questions on medical bills and collections, so we combine the 2 simultaneous surveys to create our main sample (n = 4486). The combined data were weighted to be representative of the CCIP and so are representative of consumers with a credit record and all analysis uses survey weights. The annual report for the surveys describes survey methods.¹³

Medical debt definitions

The survey, following a similar strategy in previous research,³ asks 3 broad questions about medical debt, including whether “you and your household” had (1) “any medical or dental bills that are past due or that you are unable to pay,” (2) “any medical or dental bills that you are paying off over time directly to the provider,” and (3) owed “money because you took out a loan or used a credit card to pay medical or dental bills.” The questions as they appeared on the paper survey are shown in Figure S1. These questions may overlap and a household may have more than 1 kind of debt, so we defined having medical debt as answering “yes” to any of these questions. Other surveys focus on specific kinds of medical debt; we review this research in Appendix A1.

Collection definitions

We measure medical debt in collection in 3 ways. We define an “active collection” as one in which the survey respondent reports that, in the past year, they were “contacted by someone other than your medical or dental provider to collect a past-due medical or dental bill” (see Figure S2 for the full question wording).

Medical collections on credit reports are collections that a debt collector has decided to furnish to a credit bureau and indicated was for medical debt. Credit bureau reporting rules require a furnisher to update information periodically and many debt collectors do not update regularly, so collections of all kinds often fall off credit reports.¹⁴ Moreover, when a debt is sold to a new collector, it generally appears as a new collection if it is furnished. Thus medical collections appear and disappear frequently from credit reports and the same debt may show up as different collections over time.

We define “new credit report collections” (new CR collections) as any medical collections that first appeared on a credit report over the surveys’ 12-month lookback period (February 2023 through January 2024) to match the survey questions. Finally, we define “credit report collections” (CR collections) as having any medical collection on the credit report in the 12 months before the survey, no matter when it first appeared. We discuss comparisons between our measures of CR collections and other measures in the literature in Appendix A1. During this period, the 3 major credit reporting agencies changed their reporting policies. In April 2023, the credit reporting agencies removed all medical collection under $500 from consumer credit records. Also in 2023, Colorado and New York banned the reporting of medical collections. Our 12-month period includes credit reports before these changes, but may miss some medical collections that would have been on credit reports in earlier years, and includes some collections that are no longer on reports. We show results from a survey in 2023 in Table S1, looking back over 2022 before medical collections were removed, that shows more CR collections but a similar relationship between CR collections and active collections.

Demographic definitions and exposures

We define health insurance coverage by whether the survey respondent answered “yes” to having health insurance at the time of the survey for themselves and everyone in the household. Non-homeowners include renters and people who neither rent nor own. We follow the demographic definitions for other variables in the survey annual report.¹⁵

Methods

All results are weighted to be nationally representative of adults with a credit record. Standard errors take into account survey weights using standard statistical packages (Stata version 18).

Limitations

This study has several limitations. First, the sample size limits our ability to distinguish statistically between small demographic groups. Second, the medical debt questions asked about the household while the medical collections questions asked about the individual respondent to compare directly to credit records. Respondents may not be aware of all debts held by household members, so may have understated household medical debt, and may have included collections experienced by other household members, overstating individual medical collections. Third, our estimates of the sum of all medical debt in active collection provide bounds under different assumptions rather than precise point estimates.

Results

Medical debt incidence

Table 1 shows medical debt overall and its 3 components in 2024. Overall, 36.3% (SE, 0.9) of households reported having some sort of medical debt in 2024. Medical debt was more than twice as common for the lowest income households as for the highest income. Looking at the components of medical debt, 21.4% had a past-due medical bill, 22.8% were paying off a bill over time, and 16.8% borrowed and still owed money for a loan to pay a medical bill. Medical and dental providers were thus one of the most common sources of credit to US households. For comparison, the only loans that are more prevalent are credit card debt, mortgages, and auto loans.¹⁶

Table 1.

Medical debt incidence in 2024.

	Any medical debt (n = 4452)		Past-due bill (n = 4433)		Paying off over time (n = 4429)		Used credit for bill (n = 4432)
	%	SE	%	SE	%	SE	%	SE
Overall	36.3	0.9	21.4	0.7	22.8	0.7	16.8	0.7
By race or ethnicity
White	34.0	1.1	18.2	0.9	21.9	1.0	15.2	0.8
Black	47.8	2.2	34.9	2.1	29.0	2.0	20.6	1.8
Hispanic	40.3	2.2	24.3	1.9	23.2	1.9	21.5	1.9
Other	27.5	2.7	16.8	2.3	18.5	2.3	13.5	2.0
By health insurance
Some in HH are uninsured	39.1	1.8	28.2	1.7	20.8	1.5	17.4	1.4
All in HH have health insurance	35.8	1.0	19.4	0.8	23.6	0.9	16.9	0.8
By homeownership
Homeowner	31.7	1.0	15.4	0.8	22.0	0.9	15.4	0.8
Non-homeowner	43.5	1.5	30.6	1.4	24.2	1.3	19.1	1.2
By annual household income
$35 000 or less	40.7	1.7	28.7	1.6	21.1	1.4	16.3	1.3
$35 001–$65 000	39.4	1.8	26.6	1.6	24.9	1.6	19.7	1.5
$65 001–$100 000	40.4	1.9	20.0	1.6	27.5	1.8	22.7	1.7
$100 001–$175 000	34.4	2.0	15.6	1.6	23.8	1.8	16.1	1.6
$175 001 or more	17.6	2.0	6.0	1.2	13.1	1.8	4.4	1.1
By age (years)
18–34	37.6	1.9	23.5	1.7	21.4	1.7	18.0	1.5
35–44	41.9	2.1	26.3	1.8	25.3	1.8	18.7	1.6
45–54	42.0	2.1	24.7	1.8	28.9	2.0	20.2	1.7
55–64	39.8	2.0	23.8	1.8	26.6	1.8	18.5	1.6
65+	26.0	1.5	12.4	1.1	16.7	1.3	11.3	1.1
By urban/rural location
Metro	36.4	1.0	21.0	0.8	23.2	0.9	16.4	0.8
Some urban	38.8	2.7	26.3	2.5	23.9	2.4	17.5	2.0
Rural	40.6	4.0	25.0	3.7	24.0	3.1	19.9	3.1
By gender
Female	39.7	1.2	23.8	1.0	25.2	1.0	18.5	0.9
Male	33.0	1.3	18.8	1.1	20.5	1.1	15.1	1.0

This table estimates the proportion with medical debt using survey weights to be nationally representative. Observations are for the overall estimates. Source: 2024 Making Ends Meet survey.

Abbreviation: HH, household.

Active and credit record collection incidence

When someone cannot pay, or does not think they owe, a medical debt, it may go into collections. Table 2 shows how medical collections measured 3 ways compare across demographic groups. Overall, 14.6% (SE, 0.6) of consumers reported that they were contacted to collect a medical debt by someone other than their provider in the year before the survey. Medical debt collections vary substantially by demographic group: 17.2% of respondents whose household income was $35 000 or less were contacted, while only 4.8% of households whose income was more than $175 000 were contacted.

Table 2.

Medical collections in 2024.

	Active Collection (n = 4426)		New CR Collection (n = 4485)		Any CR Collection (n = 4485)
	%	SE	%	SE	%	SE
Overall	14.6	0.6	5.4	0.3	11.6	0.5
By race or ethnicity
White	12.6	0.7	4.0	0.4	10.0	0.6
Black	23.3	1.9	10.8	1.2	20.5	1.7
Hispanic	16.9	1.7	7.4	1.1	12.4	1.4
Other	10.9	1.8	3.4	0.9	6.3	1.2
By health insurance
Some in HH are uninsured	17.2	1.4	9.6	1.0	17.8	1.4
All in HH have health insurance	14.0	0.7	4.3	0.3	9.7	0.5
By homeownership
Homeowner	12.1	0.7	3.3	0.3	8.1	0.5
Non-homeowner	18.5	1.1	8.9	0.7	17.2	1.0
By annual household income
$35 000 or less	17.2	1.3	9.4	0.9	19.2	1.3
$35 001–$65 000	18.6	1.4	6.5	0.7	12.4	1.1
$65 001–$100 000	15.3	1.4	4.0	0.6	9.6	1.0
$100 001–$175 000	12.0	1.4	2.8	0.6	7.1	1.1
$175 001 or more	4.8	1.1	1.1	0.5	2.9	0.8
By age (years)
18–34	14.8	1.4	6.1	0.7	11.9	1.1
35–44	17.0	1.5	7.8	0.9	14.6	1.3
45–54	16.6	1.6	6.6	1.1	14.3	1.5
55–64	18.6	1.6	6.1	0.9	14.1	1.3
65+	9.2	1.0	2.1	0.4	6.3	0.8
By urban/rural location
Metro	14.3	0.7	5.1	0.4	11.3	0.6
Some urban	18.3	2.2	7.1	1.3	15.0	1.8
Rural	16.5	3.2	11.7	2.9	16.6	3.1
By gender
Female	16.5	0.9	6.7	0.5	14.2	0.8
Male	12.7	0.9	4.2	0.5	8.8	0.7

This table estimates the proportion with medical collections using survey weights to be nationally representative. Active collections are contacts in the year up to January 2024 based on the survey. New CR collection is any new medical collection on the credit record in the year up to January 2024. Any CR collection is having any medical collection on the credit record in the year up to January 2024. Observations are for the overall estimates. Sources: 2024 Making Ends Meet surveys and the Consumer Credit Information Panel.

Abbreviation: HH, household.

The percentage with a new CR collection in the past year (5.8%) or any CR collection (11.6%) was lower than the percentage with active collections. There are important demographic differences between CR and active collections, suggesting that CR collections are not just a random selection of active collections. In particular, CR collections and new CR collections declined much more rapidly with income than active collections. These demographic differences imply that CR collections are not a good proxy for medical collections in general.

Relatively few consumers had both an active collection and a CR collection, while more had one or the other. In the year up to January 2024, 9.6% of all people only had an active collection, 4.7% had both an active collection and a medical collection furnished to the credit bureau, and an additional 6.9% had a collection on their credit record but were not contacted. These results suggest that less than one-third of active collections correspond to a CR collection and that credit reporting frequently occurs even in the absence of active collection efforts. While certain medical collections were removed in 2023 during the survey's lookback period,^10,11 we find very similar results for the same survey in 2023, as shown in Table S1. Table S2 shows the demographic distributions of these different collection types.

Compared with people with CR collections, people with active collections but not CR collections were older, had higher incomes, were more likely to be homeowners, and were more likely to be insured. However, the sample sizes of groups with collections are relatively small, so the point estimates have large standard errors. Our main conclusion is that people with active collections appear to be overall in a better financial condition than people with CR collections.

Number of active medical collections

Table 3 shows how many bills people were contacted about in the previous year. We report the 3 options presented on the surveys: 1 bill, 2–4 bills, and 5 or more bills. For people who were contacted, 34% were contacted for 1 bill, 52% for 2–4 bills, and 14% for 5 or more bills. Demographic groups that were less likely to be contacted were also contacted about fewer bills. For example, the lowest income groups are more likely to be contacted, as shown in Table 2, and are contacted about more bills if they are contacted, as shown in Table 3.

Table 3.

Number of bills contacted about in 2024, if contacted.

	1 bill (n = 720)		2–4 bills (n = 720)		5 or more bills (n = 720)
	%	SE	%	SE	%	SE
Overall	34.3	2.2	52.1	2.3	13.6	1.6
By race or ethnicity
White	36.9	3.1	50.2	3.2	13.0	2.1
Black	29.1	4.1	58.7	4.4	12.2	2.7
Hispanic	29.2	4.8	51.1	5.6	19.7	5.2
Other	45.0	8.6	47.5	8.6	7.4	6.3
By health insurance
Some in HH are uninsured	25.0	3.6	56.5	4.3	18.5	3.7
All in HH have health insurance	37.5	2.6	50.6	2.7	11.8	1.8
By homeownership
Homeowner	36.3	3.1	51.5	3.2	12.2	2.2
Non-homeowner	32.3	3.0	52.9	3.3	14.9	2.4
By annual household income
$35 000 or less	24.4	3.3	57.8	4.0	17.8	3.4
$35 001–$65 000	38.7	4.3	48.8	4.2	12.5	2.2
$65 001–$100 000	39.5	4.9	45.5	5.0	14.9	4.1
$100 001–$175 000	34.1	5.5	58.5	6.0	7.4	3.7
$175 001 or more	53.1	12.8	46.9	12.8	—	—
By age (years)
18–34	30.9	4.7	56.7	5.1	12.4	3.8
35–44	31.1	4.3	54.0	4.7	14.9	3.2
45–54	31.8	4.7	48.1	5.1	20.1	4.5
55–64	28.6	4.3	58.8	4.8	12.7	3.4
65+	56.0	5.6	36.6	5.4	7.4	2.5
By urban/rural location
Metro	35.3	2.6	51.7	2.7	13.0	1.9
Some urban	28.0	5.7	58.5	6.6	13.5	4.6
Rural	28.2	10.0	61.3	10.4	10.4	5.1
By gender
Female	33.0	2.8	50.5	2.9	16.5	2.3
Male	35.7	3.7	53.4	3.9	10.9	2.5

This table estimates the proportion with different medical collections frequencies, conditional on having at least 1 active medical collection, using survey weights to be nationally representative. Rows consist of 100% of the sample, although may not add up exactly because of rounding. Observations are for the overall estimates. Sources: 2024 Making Ends Meet surveys.

Abbreviation: HH, household.

Active medical collection size

Table 4 shows that the bills being collected were often large and the total amount in collections was potentially very large. The surveys asked people who had been contacted, “The most recent time you were contacted, about how much was the bill?” The first column in Table 4 shows the mean amount by demographic group. Because amounts vary widely, the mean is sensitive to a few large reported values. We winsorize the top 99% of recent collections shown in Table 4, removing these very high collection amounts.

Table 4.

Mean collection amounts reported (in US dollars), conditional on being contacted.

	Recent collection amount (n = 686)	SE	Estimated total collection amount (n = 686)	SE
Overall	2456	259	7931	1156
By race or ethnicity
White	2497	434	8664	2033
Black	2215	301	6986	1163
Hispanic	2679	433	7793	1102
Other	2334	684	5209	1901
By health insurance
Some in HH are uninsured	3490	673	12 343	3702
All in HH have health insurance	2068	255	6385	849
By homeownership
Homeowner	2351	357	7989	1961
Non-homeowner	2583	382	7931	1220
By annual household income
$35 000 or less	3427	603	10 423	1770
$35 001–$65 000	2022	226	6184	1046
$65 001–$100 000	2468	783	10 103	4716
$100 001–$175 000	1621	317	4517	1029
$175 001 or more	1420	767	3760	2299
By age (years)
18–34	1903	299	5968	1141
35–44	2323	330	7598	1172
45–54	2915	901	12 311	5403
55–64	2194	308	5894	694
65+	3356	1071	9294	3261
By urban/rural location
Metro	2469	295	7871	1440
Some urban	1660	301	5608	1328
Rural	5764	3145	15 445	9466
By gender
Female	2770	413	9473	1990
Male	2047	299	6059	921

This table estimates the mean most recent collection and mean total collections. We estimate the total collection amount by multiplying the most recent collection by the number of bills a respondent was contacted about (see Table 3) assigning the distribution: “1 bill” = 1, “2–4 bills” = 3, and “5 or more bills” = 6. Responses to the most recent collection are winsorized at the 99th percentile prior to all computations. Observations are for the overall estimates. Sources: 2024 Making Ends Meet surveys.

Abbreviation: HH, household.

The mean most recent collection was $2456. The mean collection falls as income rises; not only were high-income households less likely to be contacted (Table 2) and contacted about fewer bills (Table 3) but they were contacted about smaller bills when they were contacted, as shown in Table 4. People with health insurance had a mean most recent bill half as large as those without.

The second column in Table 4 estimates the mean total amount in collection by combining the amount of the recent bill with the number of bills contacted about shown in Table 3. If the most recent bill contact is randomly selected from bills actively in collection, then this method estimates mean total active collections. However, if debt collectors contact people more frequently about larger bills, then the most recent bill will tend to be larger than the typical bill and the estimated total will be too large. We therefore report several variations that bound the total amount.

Assuming the most recent bill is representative, of the 14.6% of people contacted about at least 1 bill, the average amount under collection for all bills was $7931. If we instead assume that all other bills other than the most recent were for $0, then the $2456 in Table 4 column 1 is also the total active collection in the previous year.

In Table S3, we provide an intermediate calculation. Rather than assuming other collections are as large as the most recent or are all zero, we use the relationship across respondents with different numbers of collections to estimate how the size of the most recent collection, on average, is related to the other collections (see Table S3 footnotes). This calculation gives us an intermediate value of $5069. With an adult population of 262 million,¹⁷ these calculations imply that, on the low end, there was $93.9 billion in medical debt actively in collection, at the high end $303.4 billion, with the intermediate variation suggesting $193.9 billion.

Discussion

Our medical debt estimates compare closely with other survey research with comparable definitions,³ but are higher than surveys or research that use more restrictive definitions of medical debt.^12,18,19 We review other survey definitions and results in more detail in Appendix A1.

One reason that survey definitions matter is that a substantial portion of medical debt and medical debt in collection may represent disputed debt that some definitions exclude. It seems that many people have medical debt not because they cannot pay but because complex third-party billing²⁰ or denials²¹ mean that they believe their insurance should have paid. A 2023 Making Ends Meet survey question asked people with an out-of-pocket medical bill not covered by their insurance, “Did you expect your health insurance (if you had any) would pay for any of this amount?”, and 64% (SE, 1.8) of those with insurance believed it should pay some or all of the bill, which likely contributes to nonpayment and to bills sent to collections.

Our estimates of the total bills under collection are higher than estimates of medical collections furnished to a single credit bureau⁴ because, as Table 2 shows, only a fraction of active medical collections are furnished. We measure CR collections over 1 year to match the survey, and so find a larger share of people with a CR collection than point-in-time estimates,^11,22 but find comparable estimates when we restrict to a single month (see Appendix A1).

Most active collections are not furnished to a credit bureau, even before recent removals (see 2023 values in Table S1), which may explain why programs that remove medical collections from credit reports by buying, then forgiving, medical collections have limited impact on financial distress.²³ Our results suggest that medical collections furnished to credit bureaus do not generally appear to be the debts that collectors are most actively engaged in collecting. The recent removal of medical collections under $500 from all credit records, and the adoption of state policies banning medical collection reporting altogether, may expand the discrepancy between active collections and CR collections and increase the degree to which measurement of medical debt using credit records understates total medical debt.

A substantial literature has documented that more health insurance reduces medical collections reported to a credit bureau, generally using Medicaid expansions.^6-9 Reporting raw differences, households where everyone is covered by health insurance have lower CR collections (Table 3) and substantially lower mean active collection amounts. However, we find smaller raw differences by health insurance for the incidence of active collections, or for medical debt more generally. Our results do not have a causal interpretation, and we do not control for other attributes. Indeed, because we measure health insurance at the time of the survey, people with expensive medical care and debt may sign up for insurance afterwards, so medical debt may cause some people to have insurance rather than the reverse. Yet, our findings do suggest that results about medical collections on credit reports, which our evidence suggests are often not being actively collected, do not necessarily apply to medical collections or to medical debt overall. For example, because health insurance does not cover all health care, if people with health insurance use more health care services, they may generate more medical debt, even if most of their health care is covered by insurance. As we discuss above, some medical debt, including medical collections, may represent disputed debt, which is more likely for people with insurance.

Conclusion

We estimate the incidence of medical debt and the incidence, frequency, and amounts of medical debt in collection by third parties. People living in low-income households were particularly likely to have medical debt and active collections. Medical debt and medical debt in collection were common, but most active collections were not furnished to a credit bureau. Medical collections reported to a credit bureau are not the same as medical collections that debt collectors are actively attempting to collect, and neither represents all medical debt. There are important demographic differences between these different kinds of medical debt, so future research and policy should carefully distinguish between them.

This article focuses on measuring different debts, but different debts may have different implications for households and medical providers. For example, 1 experiment suggests that buying medical collections reported to credit bureaus has limited impact on peoples’ financial well-being,²³ but it is possible that buying active collections may have larger effects. Disputed medical debt—which perhaps someone could pay but believes their insurance should have paid or has not been appropriately billed—may have different implications for financial well-being from debt being paid off over time to a provider or financial institution, which, in turn, may be different from debt someone cannot pay.

Supplementary material

Supplementary material is available at Health Affairs Scholar online.

Funding

None.

Data availability

Public use files for the surveys are available at: https://www.consumerfinance.gov/data-research/making-ends-meet-survey-data/ (accessed August 14, 2025).