On the Evidence of These Numbers Why Consumers File for Bankruptcy

<p>At the heart of current debates about bankruptcy reform are different images, or stereotypes, of typical
consumer filers. On the one hand is the image of the poor but honest debtor who blamelessly falls so far
behind in his or her obligations that a fresh start under Title 11 is an appropriate remedy. On the other hand
is the image of the crafty cheater or moral weakling who abuses the good intentions of the statute by
walking away from debts that he or she could, in fact, pay. Much of the discussion surrounding the reform
debates relies frankly or implicitly on these stereotypes.

</p><p>It is our position that no stereotype can fairly characterize the great range of persons and circumstances
found among the millions of consumer debtors. What all studies agree on is that almost all consumer
debtors, particularly in chapter 7, are on the lower rungs of the American income ladder; this is hardly a
stereotype. Beyond this simple characterization, what one encounters among such debtors is great <i>variety.</i>
The mass of consumer debtors cannot be characterized by any description so overly simple as the
stereotypes frequently used in editorials and sound bites.

</p><p>It is doubly unfortunate when statistical and economic inference are misused to argue for the truth
of a distorted stereotype. An example of this recently appeared in an op-ed article in a major newspaper.<small>^{<a href="#2" name="2a">2</a>}</small>
The article's argument—that bankruptcy reform is a moral issue—is one that might be defended with
subtlety and precision. But the author instead used flawed empirical reasoning to stereotype consumer
debtors as schemers.

</p><p>The text of the author's reasoning is quoted in the footnote.<small>^{<a href="#3" name="3a">3</a>}</small> Having made his argument, the author
immediately draws the following stark conclusion: "<i>On the evidence of these numbers,</i> Americans are going
bankrupt not because they're economically hard-pressed, but because they have figured out that bankruptcy
today is neither uncomfortable nor embarrassing."<small>^{<a href="#4" name="4a">4</a>}</small>

</p><p>This conclusion is itself embarrassing for more reasons than one. Here we note just one of these: it
is not reasonable to suppose that annual bankruptcy filings will rise and fall with annual measures of
economic conditions <i>for the same year.</i> To use the absence of simultaneous co-variation to support a
rhetorically loaded conclusion is indefensible.

</p><p>It would not be unreasonable, though, to seek the presence or absence of such co-variation given a
lag of time between economic change and a subsequent change in bankruptcy filings. The technical rules
for doing this work are well-known and have been used by some economists in attempts to parse the
contributions of various factors to changes in consumer filings.<small>^{<a href="#5" name="5a">5</a>}</small> Below we present a simple graph that
indicates the kind of work that is possible.

</p><p>The graph compares the rate of change in bankruptcy filings with consumer debt levels<small>^{<a href="#6" name="6a">6</a>}</small> by year since
1970. When we compared bankruptcy filing trends to changes in debt for the same year, there was little
correlation. On the chart displayed here, however, we have lagged bankruptcy filing trends (<i>e.g.,</i> 1999
bankruptcy filing trends are compared with 1997 consumer debt trends). This results in a much stronger
relationship between the two variables. There have been five periods of peak bankruptcy growth since
1970. All of these peaks occurred about two years after a spike in the rate of increase in consumer debt.

</p><p>We choose not to draw a strong inference from the obvious relationship demonstrated by the graph. Our
goals are simpler: first, to urge avoidance of stereotypes about consumer debtors; second, to call attention
to rhetoric parading as scientific inference in support of such stereotypes; and third, to show that there are
apparent relationships between economic change and <i>subsequent</i> changes in consumer filings that deserve
closer examination.

</p><hr>
<h3>Footnotes</h3>

<p>^{<small><a name="1">1</a></small>} All views expressed in this article are those of the authors and do not necessarily represent the views of the Executive Office
for U.S. Trustees. <a href="#1a">Return to article</a>

</p><p>^{<small><a name="2">2</a></small>} Frum, David, "Bankruptcy Reform is a Moral Issue," <i>The Wall Street Journal,</i> Feb. 11, 2000. <a href="#2a">Return to article</a>

</p><p>^{<small><a name="3">3</a></small>} <i>Id.</i> "The four years after 1978 were economically very volatile: there were two recessions, inflation soared and plunged,
interest rates twice scraped 20 percent. You'd think bankruptcies would increase during such difficult times. In fact, the great
takeoff in the number of consumer bankruptcies began in 1985, a year of economic smooth sailing. Bankruptcy filings declined in
1992 and 1993, economically weak years, and then rocketed upward after 1994. (They did decline somewhat in the early part
of 1999.)" <a href="#3a">Return to article</a>

</p><p>^{<small><a name="4">4</a></small>} <i>Id.</i> Emphasis added. <a href="#4a">Return to article</a>

</p><p>^{<small><a name="5">5</a></small>} <i>See, e.g.,</i> White, Michelle, "It Pays to File for Bankruptcy: A Critical Look at U.S. Bankruptcy Laws and a <a href="http://www.westlaw.com/find/default.asp?rs=CLWP2.1&amp;vr=1.0&amp;cite=…
for Change." 14 J.L. Econ &amp; Org. 205</a>, October 1998; Gross, D. and Souleles, N., "Explaining the Increase in
Bankruptcy and Delinquency: Stigma versus Risk-composition." Unpublished ms., Graduate School of Business, Univ. Chicago and the
Wharton School, Univ. Pennsylvania (1998). <a href="#5a">Return to article</a>

</p><p>^{<small><a name="6">6</a></small>} Consumer debt figures are taken from Release G.19 of the Federal Reserve Board of Governors. This figure is seasonally
adjusted and includes most short- and intermediate-term credit extended to individuals, excluding loans secured by real estate. <a href="#6a">Return to article</a>