Medical Bankruptcy: Still Common Despite the Affordable Care Act

Myriad anecdotes—of a Nobel laureate who sold his medal to pay medical bills,¹ or the more than 250 000 GoFundMe medical campaigns last year²—attest to the financial toll of illness on American families. National surveys confirm that medical bills frequently cause financial hardship,³ and the US Consumer Financial Protection Bureau reported that they were by far the most common cause of unpaid bills sent to collection agencies in 2014, accounting for more than half of all such debts.⁴

Less evidence is available on the medical causes of bankruptcy, a public and stigmatizing confession of impoverishment. In surveys conducted by researchers with the Consumer Bankruptcy Project in 2001⁵ and 2007,⁶ a majority of recently bankrupt debtors implicated medical bills or illness-related work loss as causes of their bankruptcy, findings that President Obama used to argue for passage of the Affordable Care Act (ACA). The ACA both expanded and upgraded health insurance coverage, banning preexisting illness exclusions, imposing a cap on out-of-pocket spending, and mandating coverage for essential benefits. Although these reforms might attenuate the risk of medical bankruptcy, increasing medical costs and stagnant incomes could have the opposite effect.

We sought to assess the incidence of medical bankruptcy in the current era using methods similar to those employed by the Consumer Bankruptcy Project in its 2001 and 2007 surveys. From court records of all US bankruptcy filers from 2013 to 2016, we randomly sampled 200 each quarter, abstracted their court record data, and (with institutional review board approval) mailed them a questionnaire closely modeled on the questionnaires used in those earlier studies.^5,6

Of the 3200 surveys we mailed, the postal service returned 108 as undeliverable and 910 debtors responded, for a response rate of 29.4%. Court records indicated that nonrespondents’ financial characteristics mostly resembled those of respondents; their median net worth was similar (–$32 947 vs –$30 409; P = .17), as were their assets, debts, and ongoing medical expenses (P > .05 for all comparisons), although nonrespondents had slightly higher monthly incomes ($2750 vs $2489; P < .001).

Table 1 displays debtors’ responses regarding the (often multiple) contributors to their bankruptcy. The majority (58.5%) “very much” or “somewhat” agreed that medical expenses contributed, and 44.3% cited illness-related work loss; 66.5% cited at least one of these two medical contributors—equivalent to about 530 000 medical bankruptcies annually.

TABLE 1—

Share of Debtors Citing Specific Contributors to Their Bankruptcy: United States, 2013–2016

Reasons Cited as Contributors to Bankruptcy	(A) Very Much Agree, %	(B) Somewhat Agree, %	(A) + (B) Very Much or Somewhat Agree, %
Medical-related reasons
Medical expenses	37.0	21.5	58.5
Medical problems causing work loss	27.9	16.5	44.3
Either of above	44.2	22.3	66.5
Change in family size such as birth or death	13.7	7.9	21.6
Any of above	50.1	21.4	71.5
Other reasons
Income loss (including persons with medically related work loss)	61.5	16.3	77.8
Unaffordable mortgage or foreclosure	29.2	15.8	45.0
Spending/living beyond means	17.2	27.2	44.4
Student loans	14.3	11.1	25.4
Divorce/separation	18.5	5.9	24.4
Tried to help friends/relatives	12.7	15.7	28.4

Note. The sample size of the survey was n = 910.

The share of debtors reporting a medical contributor before the ACA’s January 1, 2014 implementation (65.5%) and after implementation (67.5%) was similar (P = .37). Both of these figures are close to the 62.1% estimate from the 2007 survey, and in a difference-in-differences analysis we found no evidence that trends differed between states that did versus did not accept the ACA’s Medicaid expansion (P = .76). The responses regarding individual items in the current survey are also similar to those in 2007, when 57.1% of debtors cited medical bills as contributors to their bankruptcy and 40.3% cited income loss due to illness.⁶

Among those we surveyed from 2013 to 2016, medical debtors were more likely than other respondents to live with a spouse or partner but were similar in age, gender, and likelihood of being uninsured. Medical debtors more frequently self-reported fair or poor health (odds ratio [OR] = 2.88; P < .001), major disability (OR = 2.52; P < .001), foregoing needed medical attention in the two years prior to the bankruptcy filing (OR = 1.77; P < .001), and foregoing needed medications (OR = 2.65; P < .001).

Like all surveys, ours relies on respondents’ candor. Moreover, the modest response rate—17.1% lower than the response rate in the 2007 study—mandates cautious interpretation of our current findings. However, the similarities between respondents and nonrespondents is reassuring. Even if the medical bankruptcy rate among nonrespondents were half that of respondents, the overall rate would exceed 40%.

Our findings contrast with those of a recent study analyzing the financial sequelae of hospitalization in California from 2003 to 2007.⁷ That study found that hospitalization increased medical debts and decreased employment and income, but it suggested that medical bankruptcies were uncommon. However, its econometric approach rests on four assumptions likely to underestimate the medical bankruptcy rate. First, its cohort excluded most persons with frequent hospitalizations, a group at high risk of medical bankruptcy. Second, it assumed that only hospitalized patients can suffer a medical bankruptcy, although patients hospitalized in the course of a year account for only 18.2% of out-of-pocket costs paid by US households (Himmelstein and Woolhandler, unpublished analysis of data from the 2015 Medical Expenditure Panel Survey). Third, it assumed that a child’s, elderly parent’s, or other relative’s illness never causes a bankruptcy. Finally, the study’s assumption that every bankrupting illness starts at the moment of an initial hospitalization is contradicted by its cohort’s upsloping rate of bankruptcy filings in the baseline period prior to hospitalization. Because bankruptcy rates do not rise with age, this suggests that financial distress from illness frequently predated hospitalization. Because the study estimated medical bankruptcies from changes in filing trends before versus after hospitalization, failure to account for the upsloping baseline probably introduced a substantial downward bias.

The California study’s authors argued that survey-based ascertainment of the causes of bankruptcy is unreliable, because debtors cannot know the true cause of their financial predicament—just as heart attack patients cannot know what caused their illness.⁷ Yet in our (D. U. H. and S. W.) clinical experience, most such patients can accurately identify the smoking, dietary habits, and family history that put them at risk. Moreover, debtors are peculiarly well positioned to identify the contributors to the bankruptcy. As part of their bankruptcy proceedings, all of our respondents had recently prepared detailed documentation of their assets, debts, and current finances, and had sworn to its accuracy.

Medical bankruptcy has garnered public attention because it resonates with the abuse that Americans—including many middle-class Americans—suffer at the hands of our health care finance system. Despite gains in coverage and access to care from the ACA, our findings suggest that it did not change the proportion of bankruptcies with medical causes. That’s not surprising because the chronically poor—the group most affected by the ACA’s coverage expansion—have reduced access to credit, have few assets (such as a home) to protect, and face particular difficulty in securing the legal help needed to navigate formal bankruptcy proceedings.

Moreover, medical costs continue to outpace incomes, 29 million remain uninsured, and many of those with health insurance face unpredictable and unaffordable out-of-pocket costs as copayments and deductibles ratchet up. And few Americans have adequate disability coverage, leaving them vulnerable to illness-related income loss that amplifies the financial distress caused by medical bills.

Rather than acting to make health care more affordable, the current administration seems intent on further hollowing out coverage: encouraging a migration to bare-bones, short-term insurance policies that leave enrollees largely unprotected; allowing states to impose Medicaid work requirements that threaten to swell the ranks of the uninsured; and joining a suit that would end enforcement of the ACA’s preexisting condition coverage mandate.

The results of the midterm election—in which health care was the most prominent issue—stand as a rebuke to these retrograde steps. Instead, policymakers should move forward from the ACA and implement programs that guarantee coverage that is not just universal but also comprehensive, as well as sick leave and disability coverage that replaces income during illness.

Although death is inevitable, good public policy can ensure that financial suffering from illness is not.