Daniel A. Austin

A. Consumer Bankruptcy: Overview and Causes
1. Overview of Consumer Bankruptcy
2. Causes of Consumer Bankruptcy
B. Medical Bankruptcy
C. Medical Debt as a Cause of Bankruptcy
A. Bankruptcy Schedules
B. Surveys
A. Medical Debt is the Single Largest Cause of Consumer Bankruptcy. B. Twenty-Six Percent
of Debtors “Agree” or “Strongly Agree” That
They Filed Because of Medical Bills.
C. Sixty-One Percent of Debtors Report Medical Debt and Average
Medical Debt is $9,374.09
D. Eighteen Percent of Debtors Have Revised Medical Debt Greater
Than Half of Their Annual Income or Total Unsecured Debt

[Vol. 67:1


Daniel A. Austin*


In his 2009 State of the Union Address, President Barack Obama pleaded with Americans to support
healthcare reform, stating, “This is a cost that now causes a bankruptcy in America every thirty
seconds.”1 That jaw-dropping statistic was based on a study co-authored by Senator Elizabeth
Warren (D. Mass.) (then a professor at Harvard Law School), which concluded that 62.1% of consumer
bankruptcies are medical bankruptcies.2 The figure has been widely cited by lawmakers,
academics, and the media in support of expanded government healthcare.3 Recently, Senator Warren
co-sponsored legislation to create a new category of those filing for bankruptcy: the “medically
distressed debtor,” who would be exempt from stringent bankruptcy filing requirements.4 On the
other side, commentators and lawmakers who oppose greater government involvement in healthcare have
disputed the study’s findings.5 The issue of medical bankruptcies continues to be a focal point
in the healthcare debate.6
Several other studies have examined medical debt in bankruptcy. Using a variety of methods, these
studies have alternatively sought to support the Warren study, refute it, or replace it as the
authoritative source on medical bankruptcies. The studies have produced a wide range of estimates
for medical debt, feeding opposite positions in the debate over healthcare policy.
This study seeks to close that gap by drawing upon medical debt and other data from consumer
bankruptcy cases in 2013, and responses to a nationwide survey of recent bankruptcy filers. The
data adduced in this study shows that medical bills are the single largest causal factor in
consumer bankruptcy—but not to the degree found in the study cited by President Obama. My study
concludes that medical debt is the predominant causal factor in 18% to 26% of all consumer
* Associate Professor, Northeastern University School of Law. I am grateful to
Professors Rashmi Dyal-Chand, Kristin Madison, and Daniel Medwed for their comments. Thanks also
to Elliott Hibbler, NUSL Senior Law Librarian, for invaluable assistance. I am indebted to
attorneys Susan Grossberg, Carl Davis, Ron Satija, Carl Bekofske, Marge Burks, and many other
professionals who graciously shared their perspectives. Rachel Tritter (NUSL 2014) organized data
and provided advice. Thanks to the National Association of Consumer Bankruptcy Attorneys, the
American Bankruptcy Institute, and the Boston Bar Association for publicizing the survey. Thanks
also to NUSL student research assistants Evan Segal, Daniel DeBlander, Bradford Melson, Maja
Jachimowicz, Lawrence Fleming, Hillary Knight, Patrick Mahoney, Garth Davis, Chrisiant Bracken,
Max Ferullo, Daniela Mayer, and Molly Gachignard.
1. President Barack Obama, State of the Union Address (Feb. 24, 2009),
2. See infra notes 42–43 and accompanying text.
3. See infra note 56.
4. See infra notes 97–98 and accompanying text.
5. See infra notes 57–61 and accompanying text.
6. See, e.g., Erika Gonzalez, Bankruptcies up as medical costs soar, COLORADO PUBLIC NEWS, (May 22,
2012, 12:27 PM), soar/.


It is important to stress that in most cases, no single element can be cited as the “cause” of the
bankruptcy. The decision to file bankruptcy is typically the product of factors such as long-time
financial patterns, family and lifestyle decisions, job loss and sudden adverse events, advice from
others, and, ultimately, the individual debtor’s perception of the value and utility of filing
bankruptcy. Accordingly, the purpose of this article is not to establish a bright line definition
of “medical bankruptcies,” but rather to look at medical debt as a predominant factor of
This article will proceed as follows: Part II gives an overview of consumer bankruptcy and its
causes, and introduces the issue of medical bankruptcy; Part II also reviews previous studies on
medical bankruptcy, and shows why no studies to date offer a sound analysis of the role of medical
debt in bankruptcy; Part II concludes by introducing a coherent definition of the term “medical
bankruptcy”; Part III discusses the sources of data and methodology used in this study, while also
explaining caveats and types of data excluded from the study; Part IV presents the findings and
analysis of the medical debt data.


The term “medical bankruptcy” is sometimes used in bankruptcy and healthcare policy dialogue to
connote a consumer bankruptcy filing in which medical debts were the dispositive causal factor.
The term does not exist in the Bankruptcy Code, and the alleged impact that medical debt had on a
debtor’s decision to file may depend in part on the perspective of the individual researcher. This
Part gives an overview of consumer bankruptcy in general, and then focuses on medical bankruptcy.
I conclude Part II with my own formulation of how to determine if medical debt was the primary
causal factor in a bankruptcy.

A. Consumer Bankruptcy: Overview and Causes

1. Overview of Consumer Bankruptcy

The purpose of consumer bankruptcy is to provide “the honest but unfortunate debtor” with a
financial “fresh start” by discharging debt that the debtor has no reasonable prospect of paying.7
A personal bankruptcy is commenced by filing a bankruptcy petition,8 which includes schedules of
assets, liabilities, income, expenses, and other forms.9 Once the bankruptcy is filed, actions
to enforce obligations against the debtor are automatically stayed.10
There are primarily two types of consumer bankruptcy: Chapter 7 bankruptcy11
and Chapter 13.12 In a Chapter 7 case, the debtor surrenders his or her non-exempt
7. Grogan v. Garner, 498 U.S. 279, 286-87 (1991) (citing Local Loan Co. v. Hunt, 292 U.S. 234,
244 (1934).
8. 11 U.S.C. § 301 (2012).
9. Id. §§ 521(a)(1)–(2).
10. Id. § 362(a).
11. Id. § 707(b)(2).
12. Id. § 1322. Individuals may also file for bankruptcy under Chapter 11, but individual Chapter
11 cases are rare.

[Vol. 67:1

assets to a trustee, who sells the assets to pay unsecured creditors pro rata.13 The remaining
debts are discharged and the case is closed, usually within a few months.14 The Code exempts
certain assets from the reach of the trustee, and these are sufficiently generous to allow most
debtors to keep all of their property.15 In contrast, in a Chapter 13 case, the debtor must pay
her monthly “projected disposable income” to a Chapter 13 trustee under a plan of reorganization
that can last from three to five years.16 The trustee in turn pays unsecured creditors a pro
rata portion of their unsecured claims.17 If the debtor completes all payments under the plan, the
remaining debts are discharged.18
Except for secured debts such as home mortgages and car loans, most consumer debt is
nonpriority unsecured debt19 and typically includes claims such as credit card debt, medical bills,
utility bills, personal loans, legal claims, and student loans. These types of debt are generally
dischargeable in consumer bankruptcy, with certain exceptions.20 Nonpriority unsecured debt is
reported by debtors on bankruptcy Schedule F.21
Chapter 7 debtors must pass a “means test” to determine if the debtor qualifies for Chapter 7
relief. Simply put, if the debtor’s gross income is above the state median income, the debtor will
be presumed to have abused the bankruptcy process if he files a Chapter 7 bankruptcy.22 The
debtor must rebut the presumption,23 move to dismiss the Chapter 7,24 or convert the case to
Chapter 13 case.25 A similar test is used in Chapter 13 bankruptcy to determine the amount of
13. Id. §§ 726(a)–(b).
14. Id. § 524(a).
15. Id. §§ 522(b)(1)–(3).
16. Id. §§ 1322(a)(1)–(4), 1325(b)(1)(B), 1325(b)(1)(4)(a).
17. Id. §§ 1302(b)(5), 1325(a)(4) (debtor pays unsecured creditor an amount equal to or greater
than what the creditor would have received in a hypothetical liquidation, which is usually a
fraction of a dollar).
18. Id. § 1328(a) (with certain exceptions, the court shall grant the debtor a discharge of all
provided for by the plan or disallowed under section 502).
19. In consumer bankruptcy there are two types of unsecured debt: priority and nonpriority.
Priority debts are described in sections 507(a)(1)–(10) and include domestic support obligations
and certain taxes and other allowed unsecured obligations. As provided under section 523(a),
priority debts are generally not discharged in bankruptcy and are paid before nonpriority unsecured
debts. Id. § 726(a)(1). All other unsecured debts are nonpriority unsecured debts and are paid
after priority debts. Id. §
20. Id. § 523(a). For example, student loan debt is excepted from discharge unless the debtor
proves that paying the loans would result in “undue hardship.” Id. § 523(a)(8); see generally
Daniel A. Austin, The Indentured Generation: Bankruptcy and Student Loan Debt, 53 SANTA CLARA L.
REV. 329,
372-98 (2013) (discussing strict case law standards for discharge of student loan debt).
21. See 11 U.S.C. § 521(a)(1)(B)(i) (providing that the debtor must file a schedule of assets and
liabilities); Fed. R. Bankr. P. 1007(b)(1)(A) (specifying that debtors must use “the appropriate
Official Forms”). Schedule F is the official form for nonpriority unsecured debts. U.S. Courts,
Bankruptcy Forms, U.S. COURTS.GOV,
(last visited Sept. 12, 2014).
22. 11 U.S.C. § 707(b)(2)(A). Debtors with primarily business debts are not subject to means
testing. See id. § 707(b)(1).
23. Id. § 707(b)(2)(B).
24. A debtor may motion the court to dismiss the case, however, dismissal must be granted by the
court. Id. § 707(a).
25. Id. § 707(b)(1).


disposable income” that must be paid each month to fund the plan.26
The number of bankruptcy cases fluctuates from year to year. A record 2.01 million consumer
bankruptcy petitions were filed in 2005, as debtors rushed to file before the passage of the
Bankruptcy Abuse Prevention and Consumer Protection Act amendments took effect. Bankruptcy filings
plunged to just over 600,000 in
2006, as shown in Figure 1. They steadily rose through 2010, until gradually declining again each
year thereafter.

Figure 1: Number of Consumer Bankruptcy Cases Annually from 2005 to 2013


Number of Bankruptcy Cases

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

Source: Administrative Office of United States Courts.27

It is important to remember that even though studies analyze bankruptcy and medical debt in terms
of percentages, the actual number of bankruptcies is not constant and varies from year to year.

2. Causes of Consumer Bankruptcy

Numerous studies have looked at the causes of consumer bankruptcy, but overall there are three main
theories: sudden financial shocks, chronic poor use of resources, and deliberate action on the part
of debtors.
Researchers who see sudden financial shocks as the cause of bankruptcy point to specific events
such as income loss, marital or family problems, and medical

26. Id. §§ 1325(b)(2)–(3).
27. The Administrative Office of United States Courts maintains bankruptcy filing statistics,
including reports for the years 2007 through 2013, as listed here. U.S. Courts, Bankruptcy
Statistics, U.S. COURTS.GOV, (last
visited Sept. 12,
2014). The number of bankruptcy cases filed for the years 2004 through 2006 can be found here:
Thinking of Filing for Bankruptcy?–You’re Definitely Not Alone, THE BANKRUPTCY REPORT (Oct. 18,

[Vol. 67:1

expenses.28 These theorists reject overuse of credit as the main culprit.29 In contrast,
other commentators assert that long-term consumption patterns contribute more to personal
bankruptcy than sudden adverse events.30 Related to the overuse of credit, some authors attribute
collection pressures, not debt levels per se, as a primary element in bankruptcy filing.31
Several authors assert that a combination of high amounts of consumer credit, coupled with an
“unexpected insolvency event” causes most bankruptcies.32 For example, a 2008 survey of
bankruptcy debtors in Utah listed (in order) employment problems (loss of job, cut in pay, etc.),
overuse of credit cards, poor money management, illness and injury, aggressive debt collection, and
divorce or family problems as the main factors for filing bankruptcy.33 On the other end of the
debate, an oft-cited article concludes that consumer bankruptcy is a rationally strategic decision,
finding that the rate of filing is directly proportional to the financial benefit the debtor will
gain through bankruptcy.34
Consumer practitioners that I have polled point to interruption of income stream (including
unemployment, underemployment, and small business failure) as among the leading causes of
bankruptcy, alongside major personal and family changes, such as divorce, death of spouse, new
child, or loss of support from family
29. Darryl E. Getter, Contributing to the Delinquency of Borrowers, 37 J. CONSUMER AFF. 86, 98-
99 (2003).
30. Ning Zhu, Household Consumption and Personal Bankruptcy, 40 J. LEGAL STUD. 1, 17-18 (2011)
(concluding that medical conditions and unemployment rates are only slightly more adverse for
bankrupt households than non-bankrupt ones); contra Amy Traub, The Debt Disparity: What Drives
Credit Card Debt in America, DĒMOS 22 (May 1, 2014), (noting that unemployment
and lack of health care correspond with higher credit card debt).
Mann & Katherine Porter, Saving Up for Bankruptcy, 98 GEO. L.J. 289, 321 (2010) (rejecting the
model of sudden financial shocks as a trigger for bankruptcy and postulating instead that as debt
pressures accumulate over time, debtors “save up” emotional and financial wherewithal to file
32. Thomas A. Garrett, The Rise in Personal Bankruptcy, FEDERAL RESERVE BANK OF ST. LOUIS 7 (Oct.
2006), (citing sources
and listing divorce, death of spouse, loss of job, and medical expenses as insolvency events); see
also Personal Bankruptcy: A Literature Review, CONGR. BUDGET OFFICE (Sept. 2000), (reviewing
existing studies, which indicate that an individual is more likely to file when confronted with
adverse economic and personal issues, such as sudden medical debt, following accident or illness,
divorce, loss of income, and poor debt management).
33. Jean M. Lown, Consumer Bankruptcy in Utah (USA): Who Files and Why?, 32 INT’L. J.
CONSUMER STUD. 233, 237 (2008).
34. Scott Fay, Erik Hurst, & Michelle J. White, The Household Bankruptcy Decision, 92 AM. ECON.
REV. 706, 706 (2002) (predicting that an increase of $1,000 in household financial benefit would
result in a 7 percent increase in the number of bankruptcy filings). Another way that bankruptcy
can be said to be a rational calculation is when the debtor files as a means to avoid home
foreclosure. For example, in Chapter 13 bankruptcy, a debtor who is behind on her home mortgage
may cure the default in incremental payments over time. 11 U.S.C. § 1322(b)(5). However, this is
a tool used in bankruptcy, and does not relate to why the debtor is financially distressed.


members.35 Divorce or separation can be a problem when debtor(s) go from one household to two,36
or when an adult child returns to live with a parent.37 Other causes cited by bankruptcy
attorneys include medical debt, heavy use of credit cards and other consumer credit (particularly
by elderly or unsophisticated borrowers), high mortgage debt (decreasing as a factor in bankruptcy)
and high student loan debt (increasing as a factor in bankruptcy).38 Practitioners note that
some debtors use credit cards just to get by.39 In Massachusetts, where medical claims are much
lower for debtors, loss of income (including loss due to a catastrophic medical event) is viewed as
a leading cause of consumer bankruptcy.40

B. Medical Bankruptcy

A handful of studies have looked at medical debt as a causal factor in consumer
bankruptcy. One such study in the American Journal of Medicine has received the lion’s share of
attention41: David Himmelstein et al., Medical Bankruptcy in the United States, 2007: Results
of a National Study (2009).42
Second to that is a study in the journal Health Affairs: David Himmelstein et al., Illness and
Injury as Contributors to Bankruptcy (2005).43 Elizabeth Warren was a co-author of both articles.
In Himmelstein’s 2009 study (hereinafter Himmelstein 2009), researchers mailed surveys to 4,976
randomly selected debtors who filed for bankruptcy between January and April in 2007.44 The
seven-page questionnaire contained twenty-four primary questions, with numerous sub-parts.45 The
questions ranged
35. Email from Richard Pearson, Attorney, Prescott & Pearson, P.A., to author (Oct. 30, 2013,
18:21 EST) (on file with author) (pointing out that although student loans are
generally not dischargeable in bankruptcy, the effect of massive student loans on household
budgets causes many debtors to file bankruptcy in order to get rid of dischargeable claims so that
they can afford to pay their student loans); email from Robin Deleo, Attorney, Employer, to author
(Oct. 1, 2013, 13:37 EST), (on file with author) (citing as causes: home mortgage arrears;
unemployment and underemployment; lack of financial sophistication; and divorce).
36. Email from Susan Grossberg, Attorney, Grossberg Law Offices, to author (Feb. 5, 2014, 13:39
EST) (on file with author) (identifying pressures on single-income households resulting from
37. Email from Margaret A. Burks, Chapter 13 Trustee, Cincinnati, Ohio, to author (Feb. 10, 2014
08:32 EST), (on file with author) (citing increased financial pressures when adult children return
38. Id.; Grossberg, supra note 36. A poll of consumer bankruptcy lawyers conducted in 2012 found
student loan debt has “significantly increased” as a source of consumer debt. Student Loan Debt
39. Email from Dai Rosenblum, Attorney, Dai Rosenblum Esq., to author (Feb. 6, 2014, 03:19 PM EST)
(on file with author).
40. Email from Joanna Allison, Attorney, Presley Law, PLLC, to author (Feb. 5, 2014 12:02 PM EST)
(on file with author).
41. See infra note 56.
42. David U. Himmelstein, Deborah Thorne, Elizabeth Warren & Steffie Woolhandler, Medical
Bankruptcy in the United States, 2007: Results of a National Study, 122 AM. J. MED. 8 (Aug. 2009), [hereinafter Himmelstein 2009].
43. David U. Himmelstein, Elizabeth Warren, Deborah Thorne & Steffie Woolhandler, Illness and
Injury as
Contributors to Bankruptcy, HEALTH AFFAIRS (Feb. 5, 2005),
content/early/2005/02/02/hlthaff.w5.63.full.pdf [hereinafter Himmelstein 2005].
44. Himmelstein 2009, supra note, 42 at 741.
45. Himmelstein 2009 questionnaire (on file with author).

[Vol. 67:1

from basic demographics to detailed work-history information.46 One full page was devoted to the
debtor’s and dependents’ health and health insurance.47 Survey recipients were asked to
participate in follow-up phone interviews.48 In conducting telephone interviews, the questions
posed by the research assistants were guided by a computer program with thousands of potential
questions on a wide variety of subjects.49 Respondents received financial compensation to
participate.50 The researchers received 2,314 completed questionnaires and conducted 1,032 phone
interviews.51 The answers were used to determine whether a given case was a medical bankruptcy.
The authors broadly defined medical bankruptcies to include situations when debtors cited illness
or medical bills for filing bankruptcy, when debtors mortgaged a home to pay medical bills, or if
the debtor had uncovered medical bills greater than $1,000 or lost two weeks or more of work due to
illness or injury in the two years prior to filing.52
Himmelstein 2009 concluded that “62.1% of all bankruptcies have a medical cause.”53 The authors
compared this percentage to the results of Himmelstein’s
2005 study (hereinafter Himmelstein 2005), a study of debtors who filed bankruptcy in 2001, in
which they asserted that 46.2% of debtors had filed because of medical costs.54 Thus,
Himmelstein 2009 postulated that the odds that a bankruptcy had a medical cause was 2.38 times
higher in 2007 than in 2001.55 The Himmelstein studies have been widely cited in Congress and
46. Himmelstein 2009, supra note, 42 at 742.
47. Id.
48. Id.
49. Email from David Himmelstein, Professor, City Univ. of New York Sch. of Pub. Health at
Hunter Coll., to author (Jan. 15, 2014 9:16 AM EST) (on file with author).
50. In Himmelstein 2009, recipients were first mailed a letter introducing the survey. The
questionnaire and $2 were mailed a few days later. Nonrespondents received another questionnaire
and an additional $2. Finally, anyone who still had not responded was offered $50 to complete the
questionnaire. Himmelstein 2009, supra note 42, at 742.
51. Id. The study authors did not disclose the number of debtors who responded only
after additional mailings and payment. Id.
52. Id.
53. Id.
54. Id. at 744. Whereas the 2001 study surveyed medical bankruptcies among all debtors, the 2007
study results were tailored to reflect recent homeownership. See id. at 743.
55. Id. at 744.
56. See, e.g., Medical Bankruptcy Fairness Act: Hearing on H.R. 901 Before the Subcomm. on
Commercial and Admin. Law of the H. Comm. on the Judiciary, 111th Cong. 2 (2010) (statement of J.
Cecelia G. Morris, Bankr. S.D.N.Y.) (“It is well documented that around half of all bankruptcies
are the result of a serious medical problem.”); Sen. Edward M. Kennedy, Health Care as a Basic
Human Right: Moving from Lip Service to Reality, 22 HARV. HUM. RTS. J. 165, 166 (2009) (“Every 30
seconds in the United States, a family is forced into bankruptcy because of unexpected medical
expenses.”); Press Release, Opening Statement of Sen. Max Baucus at Sept. 22, 2009 Mark-Up of the
America’s Healthy Future Act, Chairman, U.S. Senate Comm. on Finance (Sept. 22, 2009),
6c888d1c71f1 (“Every 30 seconds, another American files for bankruptcy after a serious health
problem.”) (last visited Sept. 21, 2014); Christina LaMontagne, NerdWallet Health finds Medical
Bankruptcy accounts for a majority of personal bankruptcies, NERDWALLET HEALTH (Mar. 26, 2014),
nerdwallet-health-study-estimates-56-million-americans-65- struggle-medical-bills-2013/ (“We relied
on a widely cited Harvard study published in 2009.”); Teresa Tampkins, Medical bills prompt more
than 60 percent of U.S. bankruptcies, CNNHEALTH.COM (June 5, 2009,
9:33 AM),; Anne Underwood, Insured,


Released on April 12, 2009, Himmelstein 2009 was viewed by some as having been purposefully
designed to support enactment of the Affordable Care Act,57 and both studies have received their
share of criticism.58 One of the main complaints is that the definition of medical bankruptcy used
in the studies is overly broad. For example, David Dranove and Michael L. Millenson fault
Himmelstein’s inclusion of debtors with more than $1,000 in medical debt, noting that non-debtor
households with average income comparable to those in the Himmelstein studies typically spend more
than twice that amount on medical care per year.59 At most, according to Dranove and Millenson,
Himmelstein 2005 supports a finding of no more than 17% as medical bankruptcies.60 A 2011
article by Tal Gross and Matthew Notowidigbo likewise found the Himmelstein numbers too high,
concluding that lack of health insurance to cover out-of-pocket medical costs accounted for
approximately 26% of bankruptcies, although their study looked only at low income debtors.61
The ambitious scope of the Himmelstein studies is commendable. As a practical matter, it is not
possible to record the income and expenditure details of bankruptcy debtors before they file.
Among other reasons, no one, not even the future bankruptcy debtors themselves, knows who will file
bankruptcy in the years before they actually do so. Himmelstein and his co-authors designed their
surveys as an alternative method to obtain that information, which could only become available once
their subjects actually became bankruptcy debtors. As such, the design of the study is well
matched to the goals of the study. However, the Himmelstein studies’ detailed financial questions
were answered years after the fact by memory during phone interviews, without the benefit of
contemporaneous records or memoranda. The study participants were people undergoing severe
financial distress who likely never expected to be asked such questions when they incurred the
expenses in the years prior to bankruptcy. This raises concerns about how accurately respondents
could recall that information.62 This is particularly so
but Bankrupted Anyway, N.Y. TIMES (Sept. 7, 2009, 11:30 AM),
57. Todd Zywicki, In Elizabeth Warren We Trust?, WALL ST. J. (Sept. 30, 2010, 12:01 AM), (asserting that the
release of the Himmelstein study was “timed to promote President Obama’s health-care reform law”).
58. See, e.g., id.; see also Megan McArdle, Why Warren’s New Bankruptcy Study is So Bad, THE
ATLANTIC (June 5, 2009, 8:15 AM),
warrens-new-bankruptcy-study-is-so-bad/18834/ (asserting that the absolute number of medical
bankruptcies has decreased).
59. See David Dranove & Michael L. Millenson, Medical Bankruptcy: Myth Versus Fact, HEALTH AFFAIRS
W77-W78 (Feb. 28, 2006), (citing
U.S. Census Bureau finding that in two years, a middle-class household will spend an average of
$4,500 on health care). This paper was in response to Himmelstein 2005, but Himmelstein 2009 uses
the same
$1,000 threshold as its definition of a medical debtor. See Himmelstein 2009, supra note 42, at
60. Dranove & Millenson, supra note 59, at W79.
61. Tal Gross & Matthew J. Notowidigdo, Health Insurance and the Consumer Bankruptcy
Decision: Evidence from Expansions of Medicaid, 95 J. PUB. ECON. 767, 776 (2011). The authors also
assert that increasing Medicaid eligibility by 10% would decrease bankruptcies by about 8%. Id.
62. Memory of specific details drops exponentially with the passage of time, if the event was not
particular significance, and is subject to replacement by subsequent events and
information, self-


because the economic stress resulting from medical debt would have been a highly negative
experience.63 The fact that Himmelstein 2005 and 2009 rely so heavily on this type of data may, in
part, account for why so many other researchers dispute their findings.
Following publication of Himmelstein 2005, Iowa Republican Senator Charles Grassley requested that
the U.S. Department of Justice, through the Executive Office of the United States Trustee,
assess the Himmelstein findings.64 In response, the Justice Department reported that the
U.S. Trustee Program reviewed
5,203 bankruptcy cases filed during the same period as those in the Himmelstein study and reported
that 54% of debtors listed no medical debt.65 Of those with medical debt, 90.1% listed medical
debt under $5,000, and medical debt accounted for only 5.5% of total unsecured debt.66 Thus, the
letter asserts, “The conclusion that 50 percent of consumer bankruptcies are ‘medical related’ . .
. is not substantiated by the official documents filed by debtors.”67
Even though the Justice Department’s letter did not expressly refute the Himmelstein findings, the
letter has been viewed as such by some commentators. An article published by the American
Enterprise Institute (AEI) faults Himmelstein
2009 for failing to explain or reconcile the much lower rates of medical debts reported in the
Justice Department’s letter.68 In addition, the AEI article asserts that Himmelstein 2005 and
2009 did not compare household characteristics of non- filers (which have similar levels
of medical debt), overlooks alternative explanations for bankruptcy filings, and uses an
overly broad definition of “medical bankruptcies.”69
Melissa B. Jacoby and Mirya Holman conducted an analysis of debtors who participated in a 2007
study known as the Consumer Bankruptcy Project (CBP).70
The purpose of their study was to show that bankruptcy schedules can underreport medical debt, and
therefore bankruptcy schedule alone are not sufficient for determining the percentage of medical
bankruptcies.71 The study compared the medical debt reported by CBP participants on their
bankruptcy schedules with

distortions, unreal memories, confidence of the witness, malleability of memory, inability to
retrieve memories, and other factors. See generally Elizabeth F. Loftus, James M. Doyle & Jennifer
E. Dysart, EYEWITNESS TESTIMONY: CIVIL AND CRIMINAL 50-69 (5th ed. 2013). In addition, the
method of questioning, the wording of the questions and the witness’s perceptions of the
person asking the questions can dramatically shape how an individual “remembers” events. Id. at
63. Id. at 27-32; 54-55 (noting that memory is usually worse for stressful events).
64. Letter from William E. Moschella, U.S. Assistant Att’y Gen., to Charles E. Grassley, U.S.
Senator, Iowa (Feb. 10, 2005) available at
65. Id.
66. Id.
67. Id.
68. Arpana Mathur & Thomas P. Miller, Clarifying the Research on Medical Bankruptcy, AM. ENTER.
INST. (Apr. 27, 2011),
69. Id.
70. Melissa B. Jacoby & Mirya Holman, Managing Medical Bills on the Brink of Bankruptcy, 10
YALE J. HEALTH POL’Y L. & ETHICS 239, 242 (2010).
71. Id. at 243. Jacoby & Holman did not attempt to calculate the percentage of medical bankruptcy.


answers they provided as part of the CBP.72 Among other things, Jacoby and Holman report that
although 50% of debtors listed no medical bills on Schedule F, only 22% of CBP respondents claimed
to have incurred no medical expenses in the two years prior to filing bankruptcy.73 This
disparity could result because some debtors paid medical bills before filing, paid them using
credit cards, or omitted them from their bankruptcy schedules altogether.74 Nevertheless, these
medical expenses could have contributed to the financial distress of the debtors, but would not be
disclosed on bankruptcy schedules.75 The study suggests that the Justice Department’s letter may
have undercounted the number of medical bankruptcies.
A report by The Commonwealth Fund purported to find a high correlation between medical debt and
household financial difficulties, including bankruptcy.76
To collect data, the authors of the study had Princeton Survey Research Associates
International conduct telephone surveys of adults between April and August
2012.77 Interviewers asked about the respondents’ medical costs, health insurance, and related
issues during the preceding two-year period.78 Using a survey sample designed to generalize to
the U.S. adult population, the authors determined that
41% of respondents, representing 75 million U.S. adults, had medical debt problems.79 Of these, 6%
(representing 4 million adults) told Commonwealth Fund interviewers that they filed bankruptcy
because of medical debt.80 The problem with this number is that it significantly exceeds the
total number of bankruptcies filed during this period for all reasons combined. According to
official Department of Justice statistics, there were 1.5 million bankruptcies filed in 2010,81 1.3
million in 2011,82 and 1.1 million in 2012.83 This means that the number of bankruptcies filed
during any two-year period covered by the Commonwealth Fund Report could be as high as 3 million,
but is more likely closer to 2.6 million. Thus, it is not possible that 6% of all U.S. adults
(i.e., 4 million people) could have filed bankruptcy because of medical reasons, as the
Commonwealth Fund Report claims.

72. The authors sought to combine the bankruptcy court record method used in the U.S. Trustee
Program’s analysis with the questionnaire or survey method used in the Himmelstein studies. Id. at
73. Id. at 265, 268.
74. Id. at 271-272.
75. Id. at 273.
76. Sara R. Collins, Ruth Robertson, Tracy Garber & Michelle M. Doty, Insuring the Future: Current
Trends in Health Coverage and the Effects of Implementing the Affordable Care Act, THE COMMONWEALTH
FUND 9 (Apr. 2013),
1681_collins_insuring_future_biennial_survey_2012_final.pdf [hereinafter Commonwealth Fund Report].
77. Id at 22.
78. Id. at 1-2, 6, 9, 12.
79. Id. at 6.
80. Id. at 9.
81. Office of Judges Programs, 2010 Report of Statistics Required by the Bankruptcy Abuse
Prevention and Consumer Protection Act of 2005, ADMIN. OFFICE OF THE U.S. COURTS 5 (2011),
82. Office of Judges Programs, 2011 Report of Statistics Required by the Bankruptcy
Prevention and Consumer Protection Act of 2005, ADMIN. OFFICE OF THE U.S. COURTS 5 (2012),
83. Office of Judges Programs, 2012 Report of Statistics Required by the Bankruptcy
Prevention and Consumer Protection Act of 2005, ADMIN. OFFICE OF THE U.S. COURTS 5 (2013),


Accordingly, the Commonwealth Fund Report does not provide a credible picture of medical
Several studies have looked broadly at the correlation between illness or injury and bankruptcy.
In a 1991 survey, 19.3% of debtors claimed they filed bankruptcy because of a medical problem,
including loss of job due to injury or illness to themselves or to a family member.84 However,
only 5.7% claimed to have filed specifically because of medical bills.85 In another study,
economists using data from bankruptcies filed between 1984 and 1995 concluded there was no direct
connection between sudden health shocks and bankruptcy.86 Both studies are dated and both look at
data from only one year.
A recent study by Edward R. Morrison et al. examined adverse health shocks as a trigger of
bankruptcy filings by looking at Utah drivers involved in auto accidents (including drivers who
were determined to be not at fault for their accidents).87 This group of drivers was treated as a
proxy for consumers in general who experience “health shocks.”88 The study reported that the
bankruptcy rate for drivers who were admitted to an emergency room after a crash (“EDA Admit”) was
45% higher than the rate among drivers who did not seek medical care (“Not EDA Admit”).89
Although this may show a correlation, the authors found that it did not establish causation because
bankruptcy rates were “thirty to fifty percent” higher for EDA Admit drivers in every pre-crash
year.90 Thus, the authors conclude that households whose financial characteristics make them more
likely to file for bankruptcy are also more likely to have severe auto accidents, but that health
shocks (as represented by auto accidents) are not a causal factor on bankruptcy filing rates.91
The Morrison study is interesting, but may be affected by certain variables that were not accounted
for in the study. First, Utah has one of the highest rates of personal bankruptcy; therefore it
is more probable that an auto accident victim may also have filed bankruptcy in Utah than in other
states.92 Second, auto accidents are not necessarily representative of adverse health shocks in
general, particularly illness or on-the-job injury. As Morrison admits, financially distressed or
fragile individuals might correlate with a propensity to suffer severe auto accidents and risky
behavior, and not the other way around.93 Third, while the authors note that

MIDDLE CLASS: AMERICANS IN DEBT 145 (Yale Univ. Press 2000).
85. Id.
86. Scott Fay, Erik Hurst & Michelle J. White, The Household Bankruptcy Decision, 92 AM. ECON. REV.
706, 712-714 (2002), available at (concluding that adverse
events, with the exception of divorces, are not statistically significant predictors of the
nonstrategic bankruptcy filing model).
87. Edward R. Morrison, Arpit Gupta, Lenora M. Olson, Lawrence J. Cook & Heather Keenan,
Health and Financial Fragility: Evidence from Car Crashes and Consumer Bankruptcy 1 (Apr. 6,
88. Id.
89. Id. at 10.
90. Id. at 11.
91. Id. at 21.
92. Lown, supra note 33, at 233.
93. Morrison, supra note 87, at 10-11.


Utah has “distinctive socioeconomic characteristics,”94 they do not explore the fact that 62.2% of
the residents of Utah are members of the Church of Jesus Christ of Latter-day Saints (Mormons),
which prohibits the use of alcohol.95 Since active members of the church do not drink alcohol
and 31% of auto accident deaths are alcohol-related,96 the author’s data could also suggest that
less-religious people are more likely to file bankruptcy. Although the Morrison study suggests an
intersection between persistent risky behavior and bankruptcy, it does not disprove the theory that
health shocks are a significant cause of bankruptcy.
In June 2014, Senators Sheldon Whitehouse (D. R.I.) and Elizabeth Warren introduced the Medical
Bankruptcy Fairness Act, which would exempt “medically distressed debtors” from the onerous means
testing requirement now required of most consumer Chapter 7 debtors.97 The bill defines a
medically distressed debtor as, inter alia, one who, in the three years prior to filing, incurred
or paid for out-of- pocket medical expenses for themselves, family members or dependents, that were
not paid by a third-party payor and were greater than the lesser of 10% of adjusted gross income in
each year or $10,000.98 Although the Act does not overtly equate a “medically distressed debtor”
with a “medical bankruptcy,” it achieves the functional equivalent, because debtors who qualify
under the Act would not have to also satisfy the strict “means testing” criteria as currently
required for all other debtors.
Many debtors would likely qualify as a medically distressed debtor under the proposed legislation,
even if they did not file bankruptcy specifically because of medical debt. As I show below, 61% of
debtors report medical debt, and the average debtor has incurred $9,374 in out-of-pocket medical
expenses as reflected on Schedule F.99 Average annual income for debtors in this study is
$40,920,100 so most debtors would easily meet the 10% threshold, if not also the $10,000
94. Id. at 22.
95. Matt Canham, Census: Share of Utah’s Mormon Residents Holds Steady, THE SALT LAKE TRIBUNE (Apr.
17, 2012, 2:02 PM),
96. Alcohol-Impaired Driving, NAT’L HIGHWAY TRAFFIC SAFETY ADMIN. 1 (Dec. 2013), (this figure is from 2012).
97. Medical Bankruptcy Fairness Act, S. 2471, 113th Cong. (2014). Parallel legislation was
introduced in the House as H.R. 4917, 113th Cong. (2014). A similar bill was also introduced in
2008 as H.R. 5138, 110th Cong. (2008), but was not reported out of committee.
98. Medical Bankruptcy Fairness Act, H.R. 4917, 113th Cong. § 2(a)(1)(A)(i) (2014).
definition would also include persons who did not receive domestic support obligation payments in
excess of $10,000 because of medical problems experienced by a payor, or who experienced a change
in employment or work status that resulted in a decrease in wages due to a medical condition, as
well as the spouse of any such persons. Id. § 2(a)(1)(A)(ii).
99. See infra, p. 21.
100. Id.
101. The broad scope of the Medical Bankruptcy Fairness Act could also have far reaching
consequences unrelated to medical debt. For example, section 6 of the Medical Bankruptcy Fairness
Act would allow a debtor with student loans to receive a discharge of the loans without having to
prove “undue hardship” as otherwise required under 11 U.S.C. § 523(a)(8). Medical Bankruptcy
Fairness Act, H.R. 4917, 113th Cong. § 6 (2014). The “undue hardship” standard is a very strict
standard under current case law. See Austin, supra note 18. At present, approximately 22% of
debtors list student loan debt on their bankruptcy schedules, but fewer than one in 300 attempt to
discharge the debt. Daniel A.


C. Medical Debt as a Cause of Bankruptcy

Part of the reason that the previous studies do not provide a consistent picture of medical
bankruptcy is that there is no commonly accepted measure for when medical debt can be considered
the primary cause of the bankruptcy. Rather than use a bright line term such as medical
bankruptcy, it seems preferable to formulate how medical debt can be utilized as a causal metric of
consumer bankruptcy. It can also help in providing more precise terminology for communicating about
bankruptcy and healthcare policy. The formula used in this article is as follows:

Medical debt is the predominant causal factor of a bankruptcy if it constitutes more than 50% of
the debtor’s annual income or 50% of the debtor’s total unsecured debt, or, if the debtor herself
determines that medical debt was the primary reason for filing.

I use 50% of the debtor’s total unsecured debt or 50% of the debtor’s annual income in my
definition of predominant causal factor because it means that medical bills pose more of a demand
on the debtor’s spending or on the debtor’s income than any other single source. A
bankruptcy filing is assumed to have been primarily caused by medical debts if the debtor says
it was.102 As I have noted, before filing, debtors typically wrestle for a protracted period
about the debts they owe, how the debts will be repaid, options for reducing expenses or increasing
income, negative effects on future credit, debt negotiation or consolidation, and all other
concerns that argue for or against filing bankruptcy. By the time the deliberation process
reaches its conclusion, a debtor has almost certainly formed overall impressions as to why she is
filing bankruptcy, even if the debtor cannot recall the financial minutiae that caused her
circumstances. This may be a subjective conclusion, but it leads to an objective result—filing
bankruptcy. Accordingly, if a debtor states that she filed bankruptcy because of medical debt, then
the bankruptcy can be considered a medical bankruptcy.
My definition of medical bankruptcy is intended for use on a macro level. On a micro level, it may
certainly be possible to show that a given case is or is not a medical bankruptcy by a detailed
review of the debtor’s past financial and health history, family circumstances, external resources,
personal traits and attitudes, etc. But it is not feasible to do this type of analysis on a large
scale. It is possible, however, to access verifiable data and interpret the same based on logical
assumptions and reasoned analysis. The definition and methods I use in this study are intended to
accomplish this objective.
The reason for avoiding the term medical bankruptcy is that the decision to file bankruptcy is a
complicated one. It is typically based on a stressful combination of events, behaviors, and
personal values established over a period of years. Attempting to isolate or elevate one cause over
any other risks oversimplifying the

Austin, Student Loan Debt in Bankruptcy: An Empirical Assessment, SUFFOLK U. L. REV. (forthcoming)
(manuscript at 5, 7), available at If
the act became law, it would be much easier for student loan debtors to meet the criteria of a
medically distressed debtor than it would be to prove “undue hardship,” and it is possible that
many more student loans would be discharged than under present law.
102. This assumes the debtor is being truthful and is not unduly influenced by the questioner.


complicated individual calculus of bankruptcy. As shown in surveys103 and as noted by bankruptcy
practitioners,104 debtors file for a variety reasons. On the other hand, seeking to understand
medical debt as a causal factor, even a predominant one, is useful in order to help quantify the
different elements that cause consumer bankruptcy.


This study primarily draws from two sources of data: (1) debt and income amounts reported by
debtors on bankruptcy schedules; and (2) debtor responses to a nationwide survey. Bankruptcy data
is accessible and provides a relatively objective measure of debt. Because survey data is based
on the debtor’s perceptions, it represents a more subjective percentage of medical bankruptcy.

A. Bankruptcy Schedules

Financial information provided by debtors on bankruptcy schedules has a strong measure of
credibility. The information is relatively current, the debtor is legally obligated to be
truthful,105 and it is in the debtor’s own interest to provide complete and accurate information in
order to have as much debt as possible discharged in bankruptcy.106 In addition, bankruptcy
schedules are publicly available and accessible to researchers.107
To obtain bankruptcy data from debtor schedules, I used an online random selection tool to select a
“basket” of ten bankruptcy jurisdictions.108 According to the U.S. Census Bureau, in 2013, 86.6%
of U.S. residents were covered by some type of medical insurance.109 The average medical
insurance rate for the “basket” jurisdictions is 85.1%.110 This is only slightly less than the
national average,111 so the sample jurisdictions are representative of the U.S. as a whole. For
each of the

103. See infra, Table 1, and accompanying text.
104. See supra, notes 33 through 38 and accompanying text.
105. The Bankruptcy Code provides that the debtor shall not be granted a discharge if the debtor
knowingly makes a “false oath or account” with respect to the case. 11 U.S.C. § 727(a)(4). In
addition, the debtor can be denied discharge for any fraudulent conduct committed in connection
with the bankruptcy on or within one year prior to the petition date. Id. § 727(a)(7).
106. A debt of an individual debtor is excepted from discharge if it is “neither
listed nor scheduled . . . with the name, if known to the debtor, of the creditor to whom such
debt is owed . . . .” Id. § 523(a)(3).
107. Bankruptcy petitions and related materials are filed electronically, and are accessible with a
registered account through the Public Access to Electronic Court Records (PACER) at
108. The jurisdictions are Eastern District of Arkansas, District of Arizona, Southern District of
California, Middle District of Georgia, Southern District of Indiana, Northern District of New
York, District of Oregon, Northern District of Oklahoma, Western District of Pennsylvania, and
Eastern District of Wisconsin.
109. Health Insurance Coverage Status by State for All People: 2013, U.S. CENSUS BUREAU (last visited Nov.
110. Id.
111. Health Insurance Highlights: 2013, U.S. CENSUS BUREAU
www/hlthins/data/incpovhlth/2013/highlights.html (last visited Nov. 28, 2014).


“basket” jurisdictions, student research assistants randomly selected thirty Chapter
7 cases, and twenty Chapter 13 cases filed in 2013. We accessed individual bankruptcy schedules
on PACER, and recorded a variety of data from the schedules on Excel spreadsheets. The
data used for this study included debtors’ income, unsecured debt, credit card debt, and medical
Debtors report monthly income on Schedule I, and nonpriority unsecured debt on Schedule F.112
There are three columns on Schedule F for the debtor to enter information. The first column is
for the name and address of the creditor, the second is for date and consideration for the claim,
and the third column is for the amount. As long as the creditor name and address are correctly
identified, the debt can be discharged even if the consideration or amount is unclear or even
Credit card bills and medical bills are typically considered “nonpriority” unsecured debt. To
obtain the amount of medical debt, I or my student research assistants looked in Schedule F, column
1 for names that indicated a medical service provider or medical goods—such as a physician,
medical practice group, hospital, lab, ambulance, dentist, eyeglass center, etc. We also looked
for medical services lenders or medical credit card issuers113 and for collection agencies
representing medical claims. In addition, we reviewed the description of the consideration in
column 2. If the creditor name was clearly a medical provider, or if the description of the
consideration was clearly for a medical purpose, we included the claim as a medical debt. We used
a similar method for determining credit card debt. I capped both medical debt and unsecured debt
at a maximum of
$200,000 to eliminate statistical outliers.114
In many cases, the medical debt amounts listed on Schedule F do not represent the entire amount of
medical debt within the schedule. This is because the debtor may have charged medical bills to a
credit card and subsequently filed bankruptcy with the charge outstanding. The card issuer (not
the medical service provider) would be listed as the creditor in column 1 on Schedule F. With a
credit card claim, there is usually only a summary description in column 2, such as “consumer
purchases” or “credit card charge.” This is especially likely if the debtor’s attorney used
proprietary software to prepare the bankruptcy petition and schedules.115
Most software applications allow the attorney to electronically import a debtor’s
112. This is true in the vast majority of cases, but there can be exceptions. First, if a medical
bill was litigated to judgment and a lien attached, then the debt would potentially be a secured
debt. Second, if a medical debt was listed in a state family court order as an obligation of the
debtor owed to a spouse, then the debt would be priority debt. See supra note 19. These are
highly exceptional circumstances and are not relevant for purposes of this study.
113. Medical credit cards are growing in use, but some lenders are under fire for allegedly
practices. See, e.g., Ann Carrns, A Medical Credit Card Has Surprising Costs, N.Y. TIMES, Dec. 10,
114. Himmelstein et al. discarded cases with medical debt over $100,000. See supra note 42.
115. All federal bankruptcy courts require documents to be filed electronically, unless the debtor
is pro se. This is far easier to do with bankruptcy software, which submits the data
instantaneously in a single digital file. The alternative is to digitally convert each printed
page of a petition and schedules and then manually upload them on the court’s e-filing
site. Bankruptcy software automatically calculates exemptions and performs the laborious
“means testing” calculations. Although all responsible consumer bankruptcy attorneys will
individually review and modify petitions and schedules prior to filing, preparing and filing the
documents using a proprietary software package is highly efficient and nearly all attorneys who
regularly represent consumer debtors use them.


credit report and automatically populate Schedule F. However, with credit card bills, the
software does not list each individual credit card expenditure. In order for medical expenses
charged to a credit card to be listed separately, the debtor would have to inform his attorney of
the medical reason applicable to each charge. The attorney would then have to open the credit
card claim entry in the software program and manually enter a description of the expense.116 In
fact, some debtors do explain the type of charges they made, but without listing the specific
amounts. Because a precise description of the claim is not required for the debt to be discharged,
most debtors and their attorneys do not complete these extra steps.117
To account for medical debt that may be hidden in a credit card claim on Schedule F, it was
necessary to formulate a multiplier factor. To do this, I reviewed each Schedule F to
determine the amount of credit card debt listed by each debtor. Unfortunately, there is currently
no system that regularly maintains figures on the percentage of healthcare costs that patients
charge to credit cards. Nevertheless, Amy Traub and Catherine Ruetschiln conducted a national
survey on credit card debt of low- to middle-income households.118 In their survey of 997
households, Traub and Ruetschlin found that average credit card debt for this group totaled $7,145,
of which $1,678 was medical expenses.119 Using these numbers, approximately 23% of all credit
card charges are for medical expenses. Therefore, to account for medical debt charged to credit
cards in each case, I multiplied the amount of credit card debt by .23, and added the result to the
medical debt reported on Schedule F. This sum constitutes the Revised Medical Debt (“RMD”). For
debtors that did not report medical debt, I likewise listed 23% of credit card debt as RMD.
One of the possible concerns about using this method is that the credit card multiplier, 23%, comes
from the Traub and Ruetschlin study of lower- to middle- income debts. Would the same value
apply for higher income debtors? I performed a regression analysis comparing RMD and
annual income. The resulting correlation coefficient (R) was .01, which is extremely low,
meaning that there is essentially no linear relationship between income and medical debt. While
not conclusive, this suggests that for people who file bankruptcy, there is no distinction between
levels of medical debt based upon income. Accordingly, using
.23 as a credit card multiplier for all debtors is valid.
There is an element of medical debt that I do not include in calculating the debtor’s medical
debt. Schedule F is a snapshot of the debtor’s assets and liabilities as of the
petition date. It does not look backward to report the debtor’s pre-bankruptcy spending if such
spending is not a debt owed as of the date of
116. Some types of medical debt are seldom reported to credit agencies. For example, hospitals,
doctors, and medical providers rarely report payment information to credit bureaus unless they
become delinquent and go into collection. What is reported accounts for only about .07% of credit
agency data overall. Connie Prater, 15 Tips for Paying High Medical Bills, CREDITCARDS.COM (June
3, 2013),
117. To be clear, conscientious bankruptcy counsel will not simply rely on a credit report, but
the schedules with clients before filing. Competent counsel will ensure that their clients have
separately provided necessary debt information that is missing from the credit report.
118. Amy Traub & Catherine Ruetschlin, The Plastic Safety Net: 2012, DĒMOS 22 (May 22, 2012),
119. Id. at 1, 22.


filing. If a debtor had been spending a substantially disproportionate amount on certain types of
debts, and as a result was not paying other kinds of debt, a researcher looking only at the
bankruptcy schedules might conclude that the debts listed on the schedules caused the bankruptcy,
rather than the pre-bankruptcy spending.
Jacoby and Holman assert that medical debt in bankruptcy is underreported in this way. For
example, they argue that the medical profession is highly effective in getting patients to pay, so
debtors may be more likely to owe non-medical bills when they file.120 In addition,
cash-strapped patients may feel compelled to maintain a good relationship with their doctor and
leave other creditors unpaid.121
Plus, CBP participants report having paid more in pre-bankruptcy medical expenses than
they listed on Schedule F.122 Thus, pre-bankruptcy medical expenses would not appear on a schedule
of creditors but could be a contributing factor in the bankruptcy.123
Although Jacoby and Holman demonstrate how medical debt can be underreported on bankruptcy
schedules, their study does not try to determine the actual percentage of medical bankruptcies.
They do not compare medical industry billing practices to billing and collection practices for
other types of consumer debt, or contrast pre-bankruptcy spending on medical expenses with spending
on other types of expenses. Jacoby and Holman note that 80% of CBP participants spent
$5,000 or less in medical expenses during the two years prior to bankruptcy.124
However, average out-of-pocket medical spending for all U.S. households in 2007 (the year of the
CBP study) was $2,613.125 Multiplied over two years, medical
120. Jacoby & Holman, supra note 70, at 249-52. In addition, a CFPB report finds that consumers
with more medical than non-medical collections tend to have slightly higher credit scores than
those with primarily non-medical collections. This suggests that consumers may give their medical
debt somewhat higher payment priority than non-medical debt. Kenneth P. Brevoort & Michelle
Kambara, Data Point: Medical Debt and Credit Scores, CONSUMER FIN. PROT. BUREAU 5-6 (May 20, 2014),
On the other hand, a 2009 study asserted that medical businesses are not particularly efficient at
collecting out-of-pocket medical debts, with collection rates of 50 to 70% for “small dollar
liabilities” for insured patients, but only 5 to 10% for self-insured patients. The authors
further noted that bad debt in the medical industry was 13% in 2009, and rising rapidly. Patrick
Finn, Thomas Pellathy & Subham Singhai, US Healthcare Payments: Remedies for an Ailing System,
121. Jacoby & Holman, supra note 70, at 272, 279. Although Jacoby & Holman do not cite any
authority, this does happen. In a recent case that I reviewed, the debtors were an elderly couple
and the husband was being treated for cancer. At the time they filed bankruptcy, they owed $25,000
in credit card payments. About two-thirds of this amount was charges for food and other essentials
that had been shifted to their credit card so that they could pay their ongoing medical expenses.
The debtors feared being “cut off” by their medical providers if they fell into arrears. Notes
regarding this case are in my possession. But this is an anecdotal example and there is no data to
show that this is common. Just as patients might feel personally compelled to pay their doctor,
medical professionals may feel equally motivated to reduce or waive payments from patients who
cannot afford to pay.
122. Id. at 268.
123. Id. at 272.
124. Id. at 268.
125. Total out-of-pocket expenditures in 2007 were $293,647,000,000. There were 112,377,977
households, so the average is $2,613 per household. National Health Expenditure Data, CMS.GOV,


spending for average U.S. households may have been slightly more than for most of the CBP debtors,
yet most households do not file bankruptcy. In short, the extent to which pre-bankruptcy medical
expenses compel consumers to file bankruptcy is unclear. For this reason, I do not include it in
my bankruptcy schedule calculations. To the extent that pre-bankruptcy medical expenses figured in
the debtor’s decision to file bankruptcy, it is reflected in the survey results.
In addition, I do not take into account bankruptcies filed because of loss of wages from illness or
accident. To be sure, many academics and practitioners correlate long-term sickness and injuries
with bankruptcy.126 But debts that might lead a person to file bankruptcy after loss of income
may or may not be related to medical expenses. Additionally, a key purpose of this study is to
enable medical debt to be used more precisely in dialogue about healthcare policy. The Affordable
Care Act does not cover loss of income127 and there is no serious political effort at this time to
expand its coverage. Thus, loss of income as a result of a medical condition or an accident is
excluded from the scope of medical bankruptcy.
Some debtors take out home equity loans to pay medical bills. There is no source that regularly
tracks the use of home equity loans following disbursement, but the fact that lenders often
advertise home equity loans as a means to pay medical bills shows that the industry views medical
bills as a market driver.128
Jacoby and Holman report that 30% of debtors in the CBP survey cited medical debt as the reason
they filed bankruptcy.129 Of those, 11% claimed they obtained a home equity loan to pay
medical bills,130 which equals just over 3% of all debtors.131 But this number does not
reveal the amount of home loan debt incurred,
Reports/NationalHealthExpendData/NationalHealthAccountsHistorical.html (last updated Jan. 7, 2014).
126. See, e.g., Alena Allen, State-Mandated Disability Insurance as a Salve to the Consumer
Bankruptcy Imbroglio, 2011 BYU L. REV. 1327, 1338-1341 (2011) (noting disability as a cause of
consumer financial problems); Scott Ramsey et al., Washington State Cancer Patients Found to be at
Greater Risk for Bankruptcy Than People Without a Cancer Diagnosis, 32 HEALTH AFFAIRS 1143,
1143 (June 2013) (finding that cancer patients in the Western District of Washington during the
of 1995 to 2009 were 2.65 times more likely to file for bankruptcy than people without cancer);
Rosenblum, supra note 39 (asserting that “missing work or having to retire due to
illness” is a significant contributor to bankruptcy).
127. The ACA contains provisions on the type of benefits that an insurance plan must provide under
the Act
and loss of income due to injury or illness is not covered. Patient Protection and Affordable Care
Act, 42
U.S.C.A. § 18022 (West, Westlaw through P.L. 113-93). The HealthCare.Gov website, which describes
coverage under the ACA, also does not list these types of losses. See Why Health Coverage Protects
You From Health and Financial Risks, HEALTHCARE.GOV, coverage/ (last visited Oct. 12, 2014).
128. See, e.g., Broderick Perkins, Four Uses for Home-Equity Loans, REALTOR.COM, (“Home equity
loans are also a boon if you are hit with medical bills or some other emergency”) (last visited
12, 2014). Home equity loans are also used for credit consolidation, consumer purchases,
education, vacations, second homes, cars, boats, recreational vehicles, and the other sundry “big
ticket items you’ve always wanted.” Id.
129. Jacoby & Holman, supra note 70 at 273.
130. Id. at 274.
131. Jacoby & Holman correlate the incidence of home equity loans to higher out-of-pocket medical
expenses in the two years preceding the bankruptcy, but do not establish a specific dollar amount
for expense paid or debt incurred. Id.


medical expenses paid, or any other use of the loans. Since we cannot objectively measure this
variable, I exclude it from my calculation of RMD. To the extent that debtors perceive medical
expenses to be the cause of bankruptcy, this will be reflected in the survey responses, including
those debtors who took out home equity loans to pay for medical expenses.

B. Surveys

In order to gauge a debtor’s subjective determination as to why she filed bankruptcy, I distributed
a survey to bankruptcy debtors that asks the debtor a simple question: what caused you to file
bankruptcy?132 The survey consisted of a single page, with a statement at the top: “I filed
bankruptcy because,” followed by six potential reasons. The first three reasons were “loss of
job,” “medical bills,” and “divorce or marital problems.”133 The remaining three were
“accident or illness to me or a family member,” “too much spending,” and “reason not listed here.”
I used the Likert Scale for debtor responses. The survey also included a box to write additional
comments. Approximately 10% provided written comments. The survey was anonymous and, except
for the optional comments, would ordinarily take less than two minutes to complete. Because
the survey was anonymous, responses did not correlate to specific bankruptcy cases.
I distributed the survey in several ways. First, I mailed a copy of the survey along with a
stamped return envelope to 480 randomly selected debtors from “basket” jurisdictions. Of these, 57
were returned as undeliverable. There were 56 completed responses for a response rate of 13% from
valid addresses.134 I also emailed a copy of the survey to bar associations and bankruptcy
professionals throughout the U.S., asking practitioners to invite clients to complete the survey.
From that, I received 345 completed surveys. Finally, I created an online version and publicized
the link.135 I received 45 responses online. Thus, from all sources I received a total of 446
survey responses.
132. The survey was directed to 2013 filers, although a small number of the survey respondents
filed in 2012.
133. These elements have been repeatedly identified by academics and practitioners as some of the
major causes of bankruptcy. See, e.g., Jay L. Zagorsky & Lois R. Lupica, A Study of Consumers’
Post- Discharge Finances: Struggle, Stasis, or Fresh Start?, 16 AM. BANKR. INST. L. REV. 283, 295
(2008) (inferring that financial distress and debt is typically associated with “divorce,
sickness-related expenses [and] job loss”); A. Mechele Dickerson, Consumer Over-Indebtedness: A
U.S. Perspective, 43 TEX. INT’L. L.J. 135, 146 (2008) (finding that the primary causes of
bankruptcy in the U.S. are “medical debts, a divorce, or a job interruption”); Allison, supra note
40 (“most of our folks [are people] whose income has crashed”).
134. This is a higher response rate than some surveys, but lower than others. See, e.g.,
Christopher Tarver Robertson, Richard Egelhof & Michael Hoke, Get Sick, Get Out: The Medical Causes
of Home Mortgage Foreclosures, 18 HEALTH MATRIX 65, 81 (2008) (indicating a 7% response
rate); Himmelstein, supra note 42, at 1 (noting a 46.5% response rate). In Himmelstein 2009,
recipients were first mailed a letter introducing the survey; the questionnaire and $2 were mailed
a few days later. All those who did not respond received another questionnaire and another $2.
Finally, anyone who had still not responded was offered $50 to complete the questionnaire. Id. at
2. It appears that the persistence and money offered in Himmelstein 2009 positively affected the
response rate.
135. The link to the online survey is: Voluntary Online
Bankruptcy Survey, azw/viewform (last
visited Oct. 12, 2014).



This section will discuss the findings of this study.

A. Medical Debt is the Single Largest Cause of Consumer Bankruptcy.

Medical debt is the single largest cause of consumer bankruptcy as shown by debtor responses to the
survey. Debtors chose medical bills as the reason they filed more than any other specific reason.
Second was loss of job, followed by excessive spending. The category “Other” covered all other
reasons not listed; individual debtor comments described these as taxes, helping family members,
gambling, drug use, bad investments, business debts, judgments, unpaid loans, etc. Table 1
presents the results.

Table 1: Survey Answers

Strongly agree Somewhat agree &
strongly agree Lost job 102 (19%) 175
(20%) Medical bills 109 (21%) 221 (26%)
Divorce or marital problems
55 (10%) 83 (10%)
Illness or injury 75 (14%) 125 (15%) Too much
spending 66 (13%) 82 (10%) Other
117 (22%) 163 (19%) Totals
524 849
B. Twenty-Six Percent of Debtors “Agree” or “Strongly Agree” That They Filed
Because of Medical Bills.

Table 2 condenses Table 1 into just two survey answers: medical bills and all other reasons

Table 2: Medical Bills Compared to all Other Reasons

Strongly agree Somewhat agree &
strongly agree Medical bills 109 (21%) 221(26%)
All other reasons 415 (79%) 628 (74%)

As the table shows, 26% of the survey responses were “somewhat and strongly agree” for medical
bills as their reason for filing bankruptcy. A more conservative approach would be to use only
“strongly agree” as the reason for filing, which would generate the result that 21% of bankruptcies
are medical bankruptcies. Debtors were not limited to just one response, meaning they could choose
“somewhat agree” or “strongly agree” to any of the other five reasons. Therefore,
26% represents the broadest possible interpretation of the survey data and hence is the highest
possible percentage of medical bankruptcies.


C. Sixty-One Percent of Debtors Report Medical Debt and
Average Medical Debt is $9,374.09

Sixty-one percent of all debtors in the study reported medical debt on Schedule
F. The average medical debt directly reported on bankruptcy schedules was
$5,970.80. However, in order to account for medical debt charged to credit cards, I
take the amount of medical debt reported by the debtor on Schedule F, and add
23% of the total credit card debt, which is also listed on Schedule F. This provides the Revised
Medical Debt. The average credit card debt was $14,796, of which
23% is $3,403.08. Adding this amount to the medical debt actually reported by debtors gives a sum
of $9,374.09, which is the average Revised Medical Debt. In other words, in 2013, a typical
Schedule F filed in a consumer bankruptcy reflected an average of $9,374 in medical debt.

D. Eighteen Percent of Debtors Have Revised Medical Debt Greater
Than Half of Their Annual Income or Total Unsecured Debt

The average unsecured debt for debtors in this study was $55,967.78, while the average annual
income was $40,920. Of these debtors, 9% had Revised Medical Debt greater than half of their total
unsecured debt, and 15% had Revised Medical Debt greater than annual income. Combined, 18% of all
debtors had Revised Medical Debt greater than half their total unsecured debt or annual income.
This indicates that medical debt was the predominant factor in 18% of consumer bankruptcies.


A debtor’s decision to file bankruptcy typically involves many factors, both internal and external.
Attempting to ascribe a specific type of debt as the cause of any bankruptcy poses the risk of
oversimplifying a complicated set of inputs. However, there is no doubt that many Americans
struggle with medical bills.136
The data presented in this study shows that medical bills play a significant role in the debtor’s
bankruptcy calculus.
This article has discussed the practical difficulties of trying to reconstruct a debtor’s past
financial history in order to determine whether a bankruptcy is a medical bankruptcy. In the
alternative, analyzing what debtors report on their bankruptcy schedules, while accounting for
potential underreporting of medical debt, will help us understand the causal role of medical debt
in bankruptcy. In a case where medical bills make up over 50% of the debtor’s total unsecured debt
50% of the debtor’s annual income, it can be projected that medical debt was the predominant factor
in the debtor’s decision to file. Based on data from bankruptcy
136. Pamela Yip, Millions Struggling with Medical Debt, DALLAS NEWS (May 13, 2013, 6:16 PM)
consumers.ece (stating that 41% of adults had difficulty paying medical bills or are paying them
over time); Robin A. Cohen, et al., Problems Paying Medical Bills: Early Release of Estimates From
the National Health Interview Survey, January 2011- June 2012, CTR. FOR DISEASE CONTROL 1 (June
june_2012.pdf (indicating that for the first half of 2011, one in five persons under age 65 lived
in households that struggled to pay medical bills within the last 12 months).


schedules, medical debt was the predominant factor in approximately 18% of all bankruptcies.
Because bankruptcy schedules may not reflect elements such as the debtor’s pre-bankruptcy
expenditures, individual or family health background, alternative resources, or personal
decision-making process, a second method can be useful in gauging the debtor’s subjective calculus.
As a general rule, we can expect debtors to have weighed considerations and formed conclusions
about their financial circumstances and the reasons for filing bankruptcy, even if they cannot
recall costs and expenditures with specificity. Accordingly, debtor surveys as to why they filed
capture this subjective element. Based on survey data, approximately 26% of debtors filed
bankruptcy because of medical bills, more than any other single reason.
My two results on when medical debt can be considered the predominant cause of bankruptcy may be
unsatisfying, especially if one is focused on a bright line test. Again, the decision to file
bankruptcy involves many elements, both objective and subjective. Accordingly, this article
attempts to employ both objective and subjective data. Academics, lawmakers, and media will
continue to differ on the role of medical debt as a causal factor in consumer bankruptcy, but this
study suggests that it is the predominant element within a range of 18 to 26% of cases.