On March 22, 2018, The New England Journal of Medicine (NEJM) published a Perspective article by Carlos Dobkin, Amy Finkelstein, Raymond Kluender, and Matthew J Notowidigdo titled “Myth and measurement — the case of medical bankruptcies” concluding, “our findings suggest that medical factors play a much smaller role in causing U.S. bankruptcies than has previously been claimed.”

Along with other background material, this was covered in the Quote of the Day for March 22, 2018:


Today, June 7, 2018, the NEJM has published a response by David Himmelstein, Steffie Woolhandler and Elizabeth Warren:


Re: Myth and Measurement — The Case of Medical Bankruptcies

The New England Journal of Medicine, Correspondence, June 7, 2018


Dobkin et al. have made an important contribution in clarifying the relationship between health shocks and economic risk; like us, they have shown that health crises have major economic consequences for families and that even the insured are not adequately protected. However, in their Perspective article (March 22 issue), they mischaracterize our studies implicating medical problems as contributors to approximately 60% of personal bankruptcies, and their claim that medical bankruptcies are uncommon rests on methodologic choices that do not capture all medical causes of bankruptcy.

Contrary to their claim that our inferences about the causal relationship between medical bills and bankruptcy did not align with our respondents’ experiences, almost everyone we labeled “medically bankrupt” explicitly told us that medical problems caused their bankruptcy. For example, 41.8% of debtors interviewed specifically cited illness as a cause of their bankruptcy, 37.8% cited illness-related income loss, and 54.9% cited medical costs.

Dobkin et al. estimate the share of bankruptcies attributable to hospitalization from the change in slope of bankruptcy-filing trends after an index hospitalization. Yet, their data show that the rate began rising before hospitalization. Since the rate of bankruptcy does not increase with age, the increasing rate of bankruptcy before hospitalization could well be due to previous medical costs. Estimated bankruptcy rates based only on a change in filings after an index hospitalization are probably underestimates. In addition, they excluded anyone hospitalized in the 3 years before the study period, thus omitting many people with frequent hospitalizations — a group likely to be at high risk for medical bankruptcy.

The authors assume that hospitalization is the sole indicator of a medical problem that could lead to financial distress. But families can drown in medical debts without a hospitalization — they may spend hours in an emergency department after an accident, followed by months of physical therapy, or have chronic conditions requiring drugs costing tens of thousands of dollars. The authors explain that most people who incur high total medical expenditures have been hospitalized. Yet out-of-pocket — not total — expenditures are most salient to bankruptcy risk; an analysis of 2015 Medical Expenditure Panel Survey data reveals that only 18.2% of out-of-pocket spending was incurred by people hospitalized during the year. In effect, the authors excluded the people who incurred 81.8% of out-of-pocket costs.

Finally, the analysis by Dobkin et al. is not designed to measure bankruptcy associated with a child’s or spouse’s illness. However, a child’s terminal illness or a spouse’s long-term care can bankrupt a family.

Although they acknowledge the limitations of their analysis, the authors assert that their results “suggest that medical factors play a much smaller role in causing U.S. bankruptcies than has previously been claimed.” Yet medical bills account for a majority of unpaid debts sent to collection, and many other studies confirm that illness often inflicts financial suffering.

Debtors’ self-reports do have limitations. But hospitalization is only part of the story, and understanding medical bankruptcy requires multiple forms of empirical investigation, including asking debtors about their histories. Characterizing debtors’ self-reports as “myth” is demeaning to people struggling with health care costs, and artificially narrowing the definition of medical bankruptcy does not improve understanding of its causes.

David U. Himmelstein, M.D.
Steffie Woolhandler, M.D., M.P.H.
City University of New York at Hunter College, New York, NY

Elizabeth Warren, J.D.
U.S. Senate, Washington, DC


The authors reply: Himmelstein et al. argue that if bankruptcy filers are asked what caused their bankruptcy, a large share will say medical expenses. But their approach is not a credible way to estimate the causes of bankruptcy. It is akin to asking patients with cardiac disease what caused their heart attack; they probably do not know whether it was poor genes, poor diet, stress, or other factors. A related problem is social desirability bias, which makes it hard to take at face value explanations reported by the bankruptcy filers.

Causal estimates require isolating a potential cause and its effect on the outcome of interest. This is why we examined the effect of hospitalizations on bankruptcy and why other, similar studies have examined the effect of automobile accidents or cancer diagnoses on bankruptcies. These studies corroborate our conclusion that medical bankruptcies are a very small share of personal bankruptcies.

Continuing to focus on medical bankruptcies distracts from the actual considerable economic costs of illness and injury. Our research highlights that for Americans — even those with health insurance — hospitalizations substantially decrease employment and income; by contrast, in Denmark, people are heavily insured against reduced earnings due to illness and injury. It is here that research and policy need to focus.

Amy Finkelstein, Ph.D.
Raymond Kluender, B.S.
Massachusetts Institute of Technology, Cambridge, MA

Matthew J. Notowidigdo, Ph.D.
Northwestern University, Evanston, IL


It is difficult to understand why Amy Finkelstein and her colleagues are fixated on discrediting the landmark study on medical bankruptcy when they concede the “considerable economic costs of illness and injury.” There are a plethora of studies confirming the magnitude of medical debt and that our current health care financing system leaves far too many exposed to financial hardship.

Finkelstein et al make the very valid point that we we need to be insured against reduced earnings due to illness and injury – a point also made in the classic medical bankruptcy study by Himmelstein et al: “improved programs are needed to replace breadwinners’ incomes when they are disabled or must care for a loved one.” In focusing narrowly on income loss, it would be wrong to neglect the profound consequences of debt due to medical bills, even for those who are insured.

Suggesting that medical bankruptcy is a myth risks reducing the political support needed to revise our health care financing system so that it does not add financial hardship as an additional consequence of illness or injury. We emphatically do need a well designed, single payer national health program – an improved Medicare for all.

Stay informed! Visit www.pnhp.org/qotd to sign up for daily email updates.