Description
The purpose of this paper is to explore the impact of creditors’ undervaluing the total
expected cost of a borrower’s bankruptcy filing because a portion of the cost will be borne by other
lenders. Creditors who bear a smaller portion of the total cost of a personal bankruptcy would be
expected to take less care to avoid triggering one
Journal of Financial Economic Policy
Externalities among creditors and personal bankruptcy
Amanda E. Dawsey
Article information:
To cite this document:
Amanda E. Dawsey , (2014),"Externalities among creditors and personal bankruptcy", J ournal of Financial
Economic Policy, Vol. 6 Iss 1 pp. 2 - 24
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-09-2013-0037
Downloaded on: 24 January 2016, At: 21:48 (PT)
References: this document contains references to 38 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 353 times since 2014*
Users who downloaded this article also downloaded:
Wendy Ming-Yen Teoh, Siong-Choy Chong, Shi Mid Yong, (2013),"Exploring the factors influencing credit
card spending behavior among Malaysians", International J ournal of Bank Marketing, Vol. 31 Iss 6 pp.
481-500http://dx.doi.org/10.1108/IJ BM-04-2013-0037
Iftikhar Hussain, (2002),"Macroeconomic determinants of personal bankruptcies", Managerial Finance, Vol.
28 Iss 6 pp. 20-33http://dx.doi.org/10.1108/03074350210767898
Zafar U. Ahmed, Ishak Ismail, M. Sadiq Sohail, Ibrahim Tabsh, Hasbalaila Alias, (2010),"Malaysian
consumers' credit card usage behavior", Asia Pacific J ournal of Marketing and Logistics, Vol. 22 Iss 4 pp.
528-544http://dx.doi.org/10.1108/13555851011090547
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Externalities among creditors
and personal bankruptcy
Amanda E. Dawsey
Department of Economics, The University of Montana,
Missoula, Montana, USA
Abstract
Purpose – The purpose of this paper is to explore the impact of creditors’ undervaluing the total
expected cost of a borrower’s bankruptcy ?ling because a portion of the cost will be borne by other
lenders. Creditors who bear a smaller portion of the total cost of a personal bankruptcy would be
expected to take less care to avoid triggering one.
Design/methodology/approach – This paper presents a theoretical model of a creditor’s decision
of how aggressively to pursue collection. The model shows that because each lender’s collection
actions increase the probability of bankruptcy, each lender will collect more aggressively when a
borrower has many loans. The paper tests the predictions of the model using a large dataset of credit
card accounts.
Findings – The model highlights an important testable result: holding the level of debt constant,
a borrower with many loans is more likely to choose formal bankruptcy and less likely to choose
informal bankruptcy, i.e. chronic non-repayment absent a bankruptcy ?ling. This paper ?nds evidence
that strongly supports the predictions of the model. Laws that limit creditor collection actions do not
appear to mitigate the effects of increasing number of loans.
Originality/value – While a few papers have tested whether strategic interactions may impact
business bankruptcy, no paper of which the author is aware has provided clear empirical evidence of
the existence of common pool effects in the personal credit market. These effects point to an important
and potentially underappreciated source of risk for borrowers and creditors in this market.
Keywords Bankruptcy, Consumer protection, Financial policy, Informal institutions, Garnishment law,
Debt collections, Credit card debt, Default risk, Property exemptions
Paper type Research paper
I. Introduction
A major factor determining whether a borrower repays her loan, declares bankruptcy,
or remains in long-term default is how aggressively her creditors pursue repayment
through garnishment, property liens, or other collection methods. Bankruptcy is likely to
yield very little to a creditor aside from any collateral securing his loan[1], and the more
a creditor collects, the more likely a borrower is to ?le for bankruptcy to halt it. Thus,
creditors, particularly those with unsecured or under-secured loans, must account for the
risk of bankruptcy when deciding whether to use aggressive collection techniques.
However, whena borrower has borrowedfrommore thanone lender, the increased risk of
bankruptcyis sharedbyall creditors. Borrowers canbecome the objects of their creditors’
common pool problem: because each lender might lose out to competing creditors, they
seize property and garnish wages rather than using more cooperative strategies.
The common pool concept provides a useful framework for understanding the impact
of credit on personal bankruptcy. Borrowing and bankruptcy are clearly linked, and
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cation – k35, g21, c72
Journal of Financial Economic Policy
Vol. 6 No. 1, 2014
pp. 2-24
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-09-2013-0037
JFEP
6,1
2
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
a substantial economic literature examines whether increasing debt results in a greater
probability of personal bankruptcy, likely through increased exposure to risk
(Zywicki, 2005). The common pool model highlights an additional channel through
which credit might impact bankruptcy: borrowers with many creditors will experience
more aggressive collection than those with one or a few creditors, independent of the
total debt level. And so increasing debt levels, if accompanied by increasing number of
lenders, may have a two-fold effect: a direct effect (increased debt burden decreases
repayment) and an indirect effect (increased number of lenders leads to increased
collection activity, which leads to bankruptcy).
One way to disentangle the two potential effects is to compare the likelihood of
bankruptcy to that of “informal bankruptcy,” de?ned as non-repayment without a
formal bankruptcy ?ling (Dawsey and Ausubel, 2004). Note ?rst that the direct effect of
increasing a borrower’s debt burden will be to increase the likelihood a borrower chooses
either informal or formal bankruptcy rather than repayment. But the indirect effect will
result in a greater likelihood of formal bankruptcy but a decreased likelihood of informal
bankruptcy. Aborrower choosing informal bankruptcy forgoes bankruptcy’s automatic
stay, which means that he remains subject to all the collections efforts that creditors can
bring to bear. Thus, if creditors’ pressure is high due to common pool effects, the cost of
informal bankruptcy relative to formal bankruptcy will also be high, which will decrease
the probability of informal bankruptcy.
This paper provides a straightforward model of a creditor’s optimization problem in
the presence of one type of collection externality. It shows that shows that, under
reasonable assumptions and holding debt levels constant, increasing a borrower’s
number of loans will lead to increased collection, which will result in an increased
probability of formal bankruptcy and a decreased probability of informal bankruptcy.
The model’s conclusions are tested using a unique dataset containing accounts of
individual credit card users, and ?nds results consistent with the model’s predictions:
increasing a borrower’s number of lenders substantially increases the likelihood a
borrower will substitute formal bankruptcy for informal bankruptcy.
This result leads naturally to the following question: are common pool effects weaker
in states with laws that restrict collection? In particular, does increasing the number of
creditors cause less substitution of formal for informal bankruptcy in states with strict
garnishment and harassment laws, and what is the impact of homestead exemption law?
This paper ?nds that, while these laws do affect a borrower’s choice among repayment,
informal and formal bankruptcy, they do not change the impact of a borrower’s number
of loans[2]. So, for example, a borrower who lives in a state that allows creditors to
garnish a signi?cant percentage of his paycheck is less likely to choose informal
bankruptcy, but increasing this borrower’s number of loans does not decrease informal
bankruptcy more than it would if he lived in a state which forbids garnishment.
Section II reviews the law and economics literature related to creditor externalities
and discusses the relevant collection and bankruptcy laws. Section III presents the
theoretical model, and Section IV provides an empirical test. Section V concludes.
II. The theory of collection externalities
A. Related literature
Bankruptcy theory in the law and economics literature has focused primarily on
business bankruptcies, and the literature can be divided into two schools (Baird, 1998):
Creditors and
personal
bankruptcy
3
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the “proceduralist” and the “traditionalist.” Proceduralists contend that the goal of
bankruptcy law is to maximize ef?ciency by preserving the value of the ?rm, arguing
that collection leads to two sources of negative externality. Collectively, multiple
creditors may forgive less than the optimal proportion of the loan (the under-forgiveness
externality), and they may not provide borrowers with suf?cient time to repay
(the under-deferment externality), in comparison to a monopolist creditor. These effects
cause excessive diminution or devaluation of a ?rm’s assets and therefore lead to
economic inef?ciency ( Jackson, 1986).
Though the proceduralist argument is focused on business bankruptcies, as Hynes
(2004) points out, creditor externalities theoretically exist with individual collection as
well. Defaulting borrowers who are deprived of a vehicle or tools of trade may not be
able to work, and their marginal productivity may fall as garnishment decreases their
incentive to earn income. And though employers are banned from ?ring an employee
for a single garnishment order, these laws are dif?cult to enforce, and a borrower may
be ?red for more than one garnishment (Consumer Credit Protection Act, 1968).
A major critique of the proceduralists is that the common pool model is empirically
unsubstantiated. Kripke (1985) criticizes proceduralist papers as “neither premised on
nor tested by facts.” Block-Leib (1993) doubts that creditor externalities are widespread,
arguing that they only occur in very speci?c situations, not as a general rule in
insolvency. Warren (1993) argues that the proceduralists ignore the complexities of the
bankruptcy system and other empirical facts in order to imperfectly ?t their model to
real world evidence. In fact, her main contention is not with their conclusions, but rather
with their methodology. For example, Warren (1987) argues that in his proceduralist
works, Douglas Baird:
[. . .] may be right that unsecured creditors will push for any wild scheme that risks assets of
the estate in order to get some shot at repayment [. . .] But Baird uses his economic analysis to
limit the inquiry – and there is the ?nal rub.
In fact, only a few empirical studies have investigated collection externalities, and none
of these have established a link between these effects and personal bankruptcy[3].
B. Collection and bankruptcy law
The underlying objective of collection law is to reward creditors for diligence. From an
ef?ciency perspective, the argument is the following: if creditors bene?t from their own
collections efforts, collection will be more productive, and the cost of credit will be
minimized (Williams, 1998). An unsecured creditor is not required to consult with other
creditors before entering into a bilateral payment or preference agreement with a
defaulting borrower[4]. A second, less diligent creditor is prevented from placing a lien
on that property, even if his loan predates or exceeds the ?rst creditor’s loan. Thus,
creditors’ repayment depends not only on the borrower’s absolute willingness to repay,
but also on their efforts relative to other creditors.
One of the more effective methods of collection is through garnishment. Garnishment
laws regulate the percentage of a borrower’s intangible property, such as wages or bank
deposits, a creditor can collect directly from the borrower’s employer or bank. Federal
law shields 75 percent of a borrower’s weekly wages or 30 times the minimum wage,
whichever is greater. States can choose to protect a greater proportion of a borrower’s
wages; during the observation period of this paper’s data, 16 states had done so while
JFEP
6,1
4
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
six states had outlawed garnishment altogether[5]. A summary of state laws on
exemptions and garnishment can be found in Table I[6].
Finally, in addition to property seizure and garnishment, creditors can attempt to
persuade a defaulting borrower to repay through phone calls and letters, either on their
own or through a professional collections agent. Even without seizing property or
wages, a creditor may be able to exert considerable pressure on the delinquent
borrower. Collection tactics by third parties are limited by federal standards set in the
Fair Debt Collections Practices Act (FDCPA), though there is evidence that collections
agents often violate its provisions[7]. Some states set stricter standards or rules that
apply to lenders themselves, and some states go so far as to allow borrowers to sue
lenders who violate anti-harassment laws[8].
In contrast to laws governing creditor and borrower interactions outside of
bankruptcy, bankruptcy law explicitly forestalls competition among creditors through
the automatic stay, which disallows all collections efforts. If an unsecured creditor
obtained a lien before the bankruptcy ?ling, like any secured creditor, his right to that
collateral is upheld in the bankruptcy proceeding. After taxes, administrative costs and
secured creditors are paid, the value of remaining non-exempt assets is divided among
creditors of the same class on a pro rata basis.
State % Garn
a
Days lien
b
Home
c
State % Garn Days lien Home
AK FED Unlimited $54,000 MT FED Unlimited $40,000
AL FED Unlimited $5,000 NC 0% 0 $10,000
AR FED Unlimited Unlimited ND FED 180 $80,000
AZ FED Unlimited $100,000 NE 15% 90 $10,000
CA FED 90 $50,000 NH 0% . . . $30,000
CO FED 90 $30,000 NJ 10% Unlimited $15,000
CT FED Unlimited $75,000 NM FED Unlimited $30,000
DE 15% Unlimited $0 NV FED 120 $95,000
FL FED Unlimited Unlimited NY 10% Unlimited $10,000
GA FED 0 $5,000 OH FED 30 $5,000
HI 19% Unlimited $20,000 OK FED 180 Unlimited
IA FED Unlimited Unlimited OR FED 90 $25,000
ID FED 0 $50,000 PA 0% 0 $15,000
IL 15% 84 $7,500 RI FED Unlimited $15,000
IN FED Unlimited $7,500 SC 0% . . . $15,000
KS FED 0 Unlimited SD 20% 60 Unlimited
KY FED Unlimited $5,000 TN FED 180 $5,000
LA FED Unlimited $15,000 TX 0% . . . Unlimited
MA FED Unlimited $15,000 UT FED 120 $8,000
MD FED Unlimited $0 VA FED 90 $5,000
ME FED Unlimited $12,500 VT 0% . . . $30,000
MI FED 91 $15,000 WA FED 60 $30,000
MN FED 70 $200,000 WI 20% 90 $40,000
MO 10% 90 $8,000 WV 20% Unlimited $15,000
MS FED Unlimited $75,000 WY FED 90 $10,000
Notes:
a
“FED” indicates that garnishment up to the federal maximum of 25 percent of salary is
allowable;
b
days garn indicates the number of days a garnishment lien precludes collection by other
creditors;
c
if a state allows consumers to choose between the federal and state exemption, the table
entry is the higher of the federal or state exemption
Table I.
State homestead
exemptions and
garnishment laws
Creditors and
personal
bankruptcy
5
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Individuals canchoose to ?le under Chapter 7 or Chapter 13[9], andexemptionlaws affect
both. Borrowers ?ling for bankruptcy under Chapter 7 forfeit their non-exempt property
in return for protection of their future income from creditors’ claims. The amount of
property individuals are allowed to exempt is ostensibly determined under federal law,
but because they can opt out, states actually determine the exemptions borrowers are
allowed. The state homestead exemption can vary from almost nothing to virtually
unrestricted. Exemptions also impact the costs of a Chapter 13 bankruptcy, in which a
borrower submits a repayment plan in return for additional time to repay and a possible
partial discharge. Under Chapter 13, a borrower must repayat least as muchas she would
under Chapter 7, which naturally will vary according to Chapter 7 exemptions. Creditors
are also barred from attaching or seizing exempt property even outside of formal
bankruptcy, unless the creditor provided funding for that property’s purchase[10].
III. The economics of the common pool model
The central thesis of the common pool model is that multiple competing creditors
“over-collect” relative to a creditor who is a borrower’s unique lender, but it is not
obvious why a creditor would ever collect less forcefully than the law allows. Collection
is directed toward imposing a cost of non-payment on the borrower, and decreasing
that cost by limiting collection decreases the probability a borrower will repay. Though
creditors receive little in a formal bankruptcy, they should expect even less from
informal bankruptcy: what the borrower will pay the creditor in a bankruptcy should
set the upper bound of her willingness to pay outside of bankruptcy. The potential for a
negotiated solution will only exist if formal bankruptcy imposes additional costs.
In the following model, a creditor restrains collection to avoid “lost value,” analogous
to the economic concept of deadweight loss[11]. This model explores two potential
sources of lost value: ?rst, the borrower’s value of her assets is greater than the resale
value the creditor would recoup from seized property. Second, informal and formal
bankruptcy impose transactions costs both to the borrower and creditors. Note that
other potential channels for collection externalities exist; for example, garnishment itself
may result in diminished productivity and reduced ability to repay, a cost that is shared
among all creditors.
A. Single creditor model
The following two-period model of a negotiation begins with a borrower who has a
single creditor. In the ?rst period, the creditor announces the level of collection he will
pursue if the borrower chooses not to repay. In the second period, the borrower has three
options: formal bankruptcy, informal bankruptcy, or repayment. If he chooses formal
bankruptcy, the borrower loses his non-exempt assets and pays the transactions costs
such as ?ling or attorney fees. If he chooses informal bankruptcy, he is subject to the
collection the creditor proposed in period one. He does not submit his non-exempt assets
as in a formal bankruptcy, but there is a positive probability the creditor will ?nd and
seize his assets. If he chooses repayment, he repays the full value of the loan.
Borrowers have assets ( y) which a creditor could seize and sell, and the market
value is private information. Assume that y is uniformly distributed between [0,Y], and
this information is public knowledge. The borrower’s value of her assets is uy,
and because assets are subject to lost value in the resale market, u . 1. If the borrower
has non-exempt property and chooses formal bankruptcy, her cost is uy 2 E þ C
FB
,
JFEP
6,1
6
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
where C
FB
captures the borrower’s transactions costs of a formal bankruptcy, and E is
the relevant exemption level.
In informal bankruptcy, the borrower remains exposed to her creditor’s collection
actions. The unsecured creditor can attempt to secure her non-exempt assets, but
because the borrower is not required to itemize the property as she would in a formal
bankruptcy, assume that the creditor is only able to attach the property with probability
a, where a [ (0,1). The creditor also has the option of collecting intangible property
(wages, bene?ts, money in a bank account) through wage garnishment and other
collection actions. Let the intangible property the creditor claims be g. While the most
common form of collection is wage garnishment, where payments would accrue over
time as deductions from a borrower’s paycheck, for simplicity let g represent the total
amount collected. Thus, a borrower’s cost of informal bankruptcy is a(uy 2 E) þ g
þ C
IB
, where C
IB
captures the borrower’s other costs of informal bankruptcy, such as
disutility fromcollections calls. Note that C
FB
is likely to be signi?cantly larger than C
IB
:
the borrower choosing formal bankruptcy must bear administrative fees as well as
signi?cant stigma and future borrowing costs, which are largely avoided if the borrower
instead chooses informal bankruptcy[12].
The borrower therefore chooses formal bankruptcy when the value of her
non-exempt assets is suf?ciently low and the following holds:
y ,
Eð1 2aÞ þ g þ C
IB
2C
FB
uð1 2aÞ
Let L be the value of the loan a borrower must repay[13]. He chooses informal
bankruptcy when his cost of informal bankruptcy is less that the cost of formal
bankruptcy or repayment, which occurs when:
y [
Eð1 2aÞ þ g þ C
IB
2C
FB
uð1 2aÞ
;
L þaE 2g 2C
IB
ua
She chooses repayment when:
y .
L þaE 2g 2C
IB
ua
Note that if all of the borrower’s assets are exempt so that E . y, then his cost of formal
bankruptcy is simply C
FB
and his cost of informal bankruptcy is g þ C
IB
. The maximum
a creditor could collect from this borrower without triggering a formal bankruptcy is
C
FB
2 C
IB
. This model assumes that C
FB
2 C
IB
, g, i.e. that the creditor’s optimal g is
high enough to always induce formal bankruptcy for these borrowers[14].
When the borrower chooses to repay the loan, the creditor receives L. Assume that
property seizure is subject to lost value, so that the creditor’s value of the assets is only
the resale price. If the borrower has non-exempt assets, then the creditor’s bene?t of
formal bankruptcy is y 2 E 2 K
B,
and the creditor’s bene?t of informal bankruptcy is
a( y 2 E) þ g 2 K
IB,
where K
B
and K
IB
represent the transactions costs to the creditor
of participating in a formal or informal bankruptcy. Note that K
IB
is likely to be
substantially higher than K
B
. The creditor’s cost of informal bankruptcy includes all
administrative costs of collection, including the percentage paid to third-party
Creditors and
personal
bankruptcy
7
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
collectors or the fees and time spent ?ling garnishment requests or liens. A formal
bankruptcy is likely to impose very little if any administrative cost to the creditor[15].
The creditor’s maximization problem can be written as:
g
max
Z
ðEð12aÞþgþC
IB
2C
FB
Þ=uð12aÞ
E
½y 2E 2K
FB
?dy
þ
Z
ðLþaE2g2C
IB
Þ=ua
ðEð12aÞþgþC
IB
2C
FB
Þ=uð12aÞ
ðað y 2EÞ 2 K
IB
þ gÞdy
þ
Z
Y
ðLþaE2g2C
IB
Þ=ua
Ldy
Increasing g decreases the range of y over which the borrower chooses informal
bankruptcy. The creditor’s maximization problem requires him to balance his bene?t
of increasing g, the higher payoff from informal bankruptcy and increased likelihood of
repayment, against his cost of increasing g, the decreased likelihood a borrower will
choose informal bankruptcy and increased likelihood she will choose formal
bankruptcy.
Note that in this model, increasing E results in equal increases in the upper and
lower bounds of the range of y over which the borrower chooses informal bankruptcy.
Therefore, increasing the exemption level does not directly increase the likelihood of
informal bankruptcy. Increasing E does result in more borrowers choosing formal
bankruptcy and fewer borrowers choosing repayment.
The choice that maximizes the creditor’s expected payoff is given by the following:
g
*
¼
Lð1 2aÞð2u 21Þ þ ðu 21ÞðaC
FB
2 C
IB
Þ þuðK
IB
2aK
FB
Þ
½2u 21?
Increasing collection increases the bene?t of informal bankruptcy and increases the
probability of formal bankruptcy or loan repayment. And so the creditor optimizes by
increasing collection when:
.
the expected payoff from informal bankruptcy is low; or
.
the expected payoffs from other choices are high.
The assumption that the administrative costs of informal bankruptcy are low relative
to formal bankruptcy for a borrower and high for a creditor ensures that aC
FB
. C
IB,
K
IB
. aK
FB
and g . 0.
B. Multiple creditors
Now suppose the borrower has borrowed loan amount L from N creditors, so that each
creditor receives L/N if the borrower chooses to repay. Let G represent total collection,
the sum of all individual creditors’ collection. Assume that seized property is divided
on a pro rata basis among all creditors in a formal bankruptcy, and that an individual
creditor’s likelihood of seizing assets in informal bankruptcy is proportional to the
JFEP
6,1
8
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
number of creditors. However, collection bene?ts the collecting creditor alone. As a
result, some of the cost of collection, the increased probability of bankruptcy, is now
shared among all creditors.
Note that this model assumes that collection results in both positive and negative
externalities because one bene?t of collection, the increased likelihood of repayment,
is also shared. This assumption of strong complementarity between one creditor’s
collection and the repayment of other creditors’ loans is somewhat implausible, but it
will establish a lower bound on the impact of creditor externalities. An extension that
explores the results of relaxing this assumption is discussed in the next section.
Each creditor chooses collection level g to maximize the following:
g
max
Z
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
E
y 2E
N
2K
FB
dy
þ
Z
ðLþaE2G2C
IB
Þ=ua
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
a
N
ð y 2EÞ 2 K
IB
þ g
dy
þ
Z
Y
ðLþaE2G2C
IB
Þ=ua
L
N
dy
In the symmetric equilibrium, each creditor maximizes his expected payoff where:
g
**
¼
Lð1 2aÞ½uðN þ 1Þ 21? þ ðNu 21ÞðaC
FB
2 C
IB
Þ þ NuðK
IB
2aK
FB
Þ
N½ðN þ 1Þu 21?
Total collection is given by:
G ¼ Ng
**
¼
Lð1 2aÞ½uðN þ 1Þ 21? þ ðNu 21ÞðaC
FB
2 C
IB
Þ þ NuðK
IB
2aK
FB
Þ
½ðN þ 1Þu 21?
As in the single creditor model, the effect of exemptions on collection is neutral, even
though high exemptions reduce a creditor’s payoff from informal bankruptcy and
formal bankruptcy and therefore increase the relative bene?t of repayment. It might
seem that, in response to generous exemptions, the creditor should increase collection in
order to induce repayment. However, a creditor considering this tack would recognize
that high exemptions would mean that increasing collection would come with an
additional cost: a greater likelihood the borrower would choose formal bankruptcy in
response. The creditor’s costs and bene?ts of changing collection in response to higher
exemptions offset, and so higher exemptions do not provoke a change in the creditor’s
collection level.
The response of total collection to an increase in the number of creditors is given by
the following:
Creditors and
personal
bankruptcy
9
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
›G
›N
¼
u
2
ð aC
FB
2C
IB
Þ þuðu 21ÞðK
IB
2aK
FB
Þ
½ðN þ 1Þu 21?
2
The impact of increasing N on G is not a function of L, the borrower’s total debt. The
extension that follows in the next section demonstrates how, in the absence of the positive
externality, ›G/›N increases with L. Instead, in this model, the relative costs of informal
and formal bankruptcy determine how collection changes as N increases. For example,
if either K
IB
or C
FB
is high, each creditor increases collection accordingly, which results in
a lower probabilityof informal bankruptcyanda higher probabilityof formal bankruptcy.
The negative externality of collection, the creditors’ shared cost of the increased likelihood
that the borrower will enter formal bankruptcy, results in the cost of bankruptcy being
increasingly externalized as N increases. Creditors therefore under-respond to the cost,
which results in over-collection (›G/›N . 0). Recall that the transactions cost of formal
bankruptcy is likely to be relatively high for borrowers and low for creditors, and as long
as K
IB
. aK
FB
and aC
FB
. C
IB
holds, then increases in the number of creditors will lead
to increased total collection, the main result tested in this paper.
C. Multiple creditors, no complementarity in collection
In the previous section, one creditor’s collection was assumed to increase the likelihood
a borrower would pay off all loans. This introduced a positive externality in collection,
which the following extension will remove.
As in previous sections, a borrower’s cost of formal bankruptcy is uy 2 E þ C
FB
and her cost of informal bankruptcy is a(uy 2 E) þ G þ C
IB
, and therefore she chooses
formal bankruptcy whenever:
y ,
Eð1 2aÞ þ G þ C
IB
2C
FB
uð1 2aÞ
However, in this extension, the borrower decides whether to repay a creditor’s loan by
comparing the cost and bene?ts of repaying that particular loan. Assume again that
creditors are symmetric, so that the cost of repaying a particular loan is L/N. Assume
further that the borrower’s administrative cost and the probability of asset seizure in
informal bankruptcy resulting from a particular lender are proportionate to the size of
his loan. So a borrower’s cost of remaining informally bankrupt on a particular loan is
(a(uy 2 E) þ C
IB
/N) þ g, where g represents the collection level chosen by an
individual creditor. Note that in this model, as compared to the previous multiple
creditor model, one creditor’s collection does not affect the borrower’s likelihood of
repaying other creditors’ loans.
A borrower chooses informal bankruptcy when:
y [
Eð1 2aÞ þ G þ C
IB
2C
FB
uð1 2aÞ
;
L þaE 2Ng 2C
IB
ua
and chooses repayment when:
y .
L þaE 2Ng 2C
IB
ua
Assume the creditor’s administrative costs of informal and formal bankruptcy are
?xed costs. The creditor will choose g to maximize the following:
JFEP
6,1
10
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
g
max
Z
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
E
y 2E
N
2K
FB
f ð yÞdy
þ
Z
ðLþaE2Ng2C
IB
Þ=ua
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
a
N
ð y 2EÞ 2 K
IB
þ g
f ðyÞdy
þ
Z
Y
ðLþaE2Ng2C
IB
Þ=ua
L
N
f ð yÞdy
Total collection is given by:
Ng
*
¼
Lð1 2aÞð2u 21ÞN þ Eað1 2aÞðu 21ÞðN 21Þ þ C
IB
½a þ Nð1 2a 2uÞ?
þaC
FB
ðNu 21Þ 2NuaK
FB
þ NuK
IB
½Nð1 2aÞ þa?
½aðu 21Þð1 2NÞ þ Nð2u 21Þ?
The impact of increasing N is given by:
›Ng
**
›N
¼
ðu 21Þð2u 21Þð1 2aÞðL þEÞ þ ½u
2
a þ ð2u 21Þð1 2aÞ? aC
FB
2½au
2
?C
IB
þ½N
2
ð1 2aÞu þ ðu 21Þ{Nað1 2aÞ þa}
2
?uK
IB
þa
2
ð1 2uÞ uK
FB
½aðu 21Þð1 2NÞ þ Nð2u 21Þ?
2
As with the previous model, assuming that K
IB
. aK
FB
and aC
FB
. C
IB
is suf?cient
to ensure that the impact on total collection of increasing number of loans is
unambiguously positive. And note further that, unlike the previous model, higher
exemptions increase the impact of number of creditors on total collection. In this model,
a creditor’s own collection is more likely to induce repayment than in the previous model.
The creditor’s bene?t of increasing collection when exemptions are high is greater than
the potential cost, the increase in the probability a borrower will choose formal
bankruptcy in response. As a result, each creditor increases collection as exemptions
increase. And as the number of creditors increase and the cost of bankruptcy becomes
more externalized, higher exemptions lead to even higher total collection.
This means that the impact of higher exemptions on borrowers is twofold. First,
higher exemptions decrease the cost of formal bankruptcy and, to a lesser degree and
depending on the probability of property seizure, informal bankruptcy. Second, higher
exemptions also lead to increased collection, which increases the borrower’s cost of
informal bankruptcy. Both of these effects act to decrease the relative cost of formal
bankruptcy, and so higher exemptions should lead to a higher probability the borrower
chooses formal bankruptcy. However, the total impacts on the relative costs of informal
bankruptcy and repayment are ambiguous, and so increasing exemptions may increase
or decrease the probability a borrower chooses repayment or informal bankruptcy.
One ?nal note: underlying each of these models is a particular form of externality
problem, where lost value occurs due to property seizure and administrative costs. Each
model assumes that higher collection results in a higher relative bene?t of both formal
bankruptcyandrepayment, andtherefore increases incollectionincreases the probabilitya
Creditors and
personal
bankruptcy
11
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
borrower will choose either formal bankruptcy or repayment. However, as discussed
above, another potential negative externalitymayresult if collectiondecreases a borrower’s
ability to repay, such as a case where the borrower is ?red for multiple garnishments. As
number of lenders increase, this externality would, like the property seizure externality,
result inanincrease aborrower’s probabilityof choosingformal bankruptcyanda decrease
the probability of informal bankruptcy. However, unlike the externality that is the subject
of the model above, inthis case, as number of lenders increase, the borrower’s probabilityof
repayment could decrease. Which externality predominates will determine whether a
borrower’s number of loans will increase or decrease his likelihood of repayment.
D. Implications for garnishment and exemption laws
If empirical tests provide evidence of collection externalities, a second line of inquiry
concerns whether lenient garnishment, harassment, and exemption laws prevent their
effects. If borrower-friendly laws effectively set a cap on the level of collection creditors can
pursue throughgarnishment andharassment laws, it follows that the effects of externalities
would be weaker in states that restrict or disallow garnishment and other aggressive
creditor techniques. Inparticular, increasingaborrower’s number of loans shouldcause less
substitution away from informal bankruptcy in states with lenient garnishment and
anti-harassment laws. In the model without complementarities in collection, higher
exemptions are positively related to common pool effects, which would suggest greater
substitution away from informal bankruptcy in states with generous homestead
exemptions. We can test these predictions by including interactions between number of
loans and level of exemptions, percentage garnished, and anti-harassment statutes.
IV. Evidence of strategic over-collection
A. Data
This analysis uses a dataset comprising 52,721 pre-approved gold card users from
a large credit card issuer. The users were respondents to three solicitations of
observationally similar borrowers; the recipients were randomly assigned offers that
varied by length of introductory period and introductory interest rate. Borrowers who
responded to offers with shorter introductory periods and higher interest rates are more
likely to be credit constrained and to exhibit more risky borrowing behavior, and so the
speci?cations that follow include indicator variables for each mailing and controls for
length of the introductory period and interest rate (Table II).
The data includes variables fromeach borrower’s credit history, along with variables
describingeachborrower’s use of the goldcardfor at least 20 months. The borrower’s zip
code is recorded, allowing the addition of demographic information for their zip code and
state. Variables representing applicable state laws were also added: homestead
exemption represents the dollar value of the homestead exemption allowed in each state,
no garnishment is a dummy variable indicating whether the state prohibits garnishment
for ordinary loans, and anti-harassment statute indicates whether the state allows a
borrower the right to sue an abusive creditor[16].
The outcome measure speci?es borrowers’ choosing among three options:
repayment, formal bankruptcy, and informal bankruptcy. formal bankruptcy holds
if a borrower declared bankruptcy during the observation period. Borrowers were
classi?ed as choosing informal bankruptcy when they were charged off for long-term
delinquency (usually six months of non-payment) without a formal bankruptcy ?ling.
JFEP
6,1
12
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
All other borrowers were classi?ed under repayment. Some speci?cations dropped
non-delinquent borrowers, that is, borrowers who were never more than two months
late with a payment.
The major advantage of this dataset is that it includes very detailed, individual-level
observations of borrower behavior. Also, the dataset includes the lending terms of the
borrower’s contract, and so all of the speci?cations reported in this paper control for the
interest rate and length of the introductory period. One other unique aspect of the data is
its temporal range: solicitations were sent between late 1995 and early 1997, and so the
accounts provide information about borrowing during the period in the late 1990s when
personal bankruptcies were climbing most dramatically[17]. The data also predates the
passage of the Bankruptcy Abuse Prevention and Consumer Protection Act (BAPCPA)
of 2005 and the ?nancial crisis beginning in 2007, both of which caused short-run
distortions in bankruptcy rates and credit market behavior. The impact of changes that
have occurred after the observation period from this data is discussed in the conclusion.
One limitation of this dataset is that total balances and number of loans are recorded
at the time of solicitation, not at default. It is conceivable that a borrower could obtain
several loans or close loan accounts between solicitation and default (or the end of the
observation period). The usefulness of more up-to-date information is somewhat
limited, however, because the borrower has an incentive to increase borrowing in
anticipation of a default. For this reason, the speci?cations do not include information
describing the borrowers’ current balances on the account that is observed in this data.
If the correlation between loan balances and number of loans at solicitation and at
default is reasonably strong, the results will be consistent.
Repayment Informal bankruptcy Formal bankruptcy
Revolving loans 4.539 (3.472) 3.110 (2.481) 4.483 (3.002)
Installment loans 3.160 (3.315) 1.760 (2.671) 3.090 (3.247)
Number of mortgages 0.528 (0.745) 0.190 (0.451) 0.357 (0.568)
Revolving balances 3,512 (4,257) 3,429 (4,701) 6,237 (5,927)
Non-revolving balances 5,257 (11,859) 3,148 (7,355) 5,391 (8,713)
Mortgage balances 36,934 (64,519) 9,602 (29,097) 23,547 (43,459)
Years on ?le 1.178 (0.748) 0.788 (0.688) 1.030 (0.638)
Credit score 632.4 (86.12) 564.1 (92.25) 564.2 (97.17)
State unemployment rate 5.304 (1.045) 5.500 (0.910) 5.382 (1.042)
Transfer payments 11.68 (3.459) 11.48 (3.838) 11.19 (3.338)
Self-employed 10.52 (1.661) 10.53 (1.415) 10.71 (1.644)
Manufacturing 15.36 (4.606) 14.61 (4.668) 15.16 (4.637)
Homestead exemption 1.103 (1.461) 1.343 (1.644) 1.219 (1.528)
Anti-harassment statute 0.559 (0.496) 0.623 (0.485) 0.537 (0.499)
No garnishment 0.158 (0.365) 0.223 (0.417) 0.149 (0.356)
County median home value 131.8 (66.18) 123.3 (60.08) 122.2 (58.09)
Percent black 11.14 (12.05) 14.13 (13.56) 12.02 (13.56)
Percent hispanic 10.42 (12.63) 13.09 (15.22) 11.37 (13.78)
Percent divorced 9.729 (1.812) 9.867 (1.837) 10.04 (1.794)
Percent uninsured 13.27 (4.131) 14.55 (4.256) 14.04 (4.157)
High school 77.09 (7.848) 74.79 (8.092) 76.24 (8.285)
Income 46,959 (31,160) 36,243 (42,166) 39,588 (21,662)
Credit limit 7,168 (3,720) 4,615 (2,813) 5,569 (3,036)
Observations 51,069 747 905
Table II.
Selected variable means
standard deviations
by outcome
Creditors and
personal
bankruptcy
13
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
B. Empirical analysis
The borrower is assumed to choose among informal bankruptcy, formal bankruptcy
and repayment according to which yields higher utility. The stochastic component of
her utility function is assumed to follow an extreme-value distribution. This error
structure yields a multinomial logit, so that the probability of outcome k is given by:
P
k
¼
e
b
0
k
X
e
b
0
R
X
þ e
b
0
I
X
þ e
b
0
F
X
:
where k ¼ repayment, informal bankruptcy or formal bankruptcy. The coef?cient
estimates are not readily interpretable, and therefore the average marginal effects are
presented and discussed.
The borrower’s choice among her three options is modeled as a function of the
explanatory variables itemized in Table III. These variables include several measures,
including loan balances, number of loans, and types of loans, recorded in a borrower’s
credit report at the time of solicitation, and the lender recorded additional variables. In
addition to this list of individual-level characteristics, several variables at the zip code
and state level were included, such as the borrower’s state and county unemployment
rate and median home value. Also included were variables describing the bankruptcy
and collection laws in each borrower’s state.
C. Results
C1. Credit variables. The ?rst three columns of Table IV report the average marginal
effects of the explanatory variables on the probability a borrower chooses repayment,
informal or formal bankrutpcy; the ?nal three columns report the results of speci?cations
that exclude credit limit and credit score. As predicted above, many of the variables from
the borrower’s credit report affect informal and formal bankruptcy in the same direction.
For example, years on ?le, credit limit and credit score all decrease the probability a
borrower will choose either formal or informal bankruptcy and increase the probability
of repayment. Counter-intuitively, an increase in a borrower’s income actually increases
the likelihooda borrower chooses either informal or formal bankruptcy, andsigni?cantly
decreases the likelihood she chooses repayment. As the ?nal three columns of Table IV
show, income has the expected negative effect on repayment and positive effects on
informal and formal bankruptcy when credit score and credit limit are dropped, while the
impacts of other variables are largely unaffected. Credit score and limit, while providing
a useful control for the creditor’s assessment of credit-worthiness, are constructed from
credit history variables and can bias these variables’ coef?cient estimates.
Unlike the other credit history variables, the borrower’s number of loans signi?cantly
increases the probability she will choose formal bankruptcy and decreases the
probability she will choose informal bankruptcy. For example, holding all else equal,
increasing a borrower’s number of loans by 1 will increase her probability of formal
bankruptcy by 0.052 percentage points, while decreasing her probability of informal
bankruptcy by 0.042 percentage points. Given that the unconditional probability of
formal bankruptcy is 1.7 percent, the 0.052 percentage point increase indicates a
3.1 percent increase in the probability of formal bankruptcy. The probability of informal
bankruptcy is 1.4 percent in this dataset, which means that a 0.042 percentage point
decrease implies a 3 percent decrease in the probability of informal bankruptcy.
JFEP
6,1
14
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The effect of number of loans on repayment is small, negative and insigni?cant when
credit score and credit limit are included. When credit score and credit limit are not
included, the coef?cient on repayment is positive and statistically signi?cant,
indicating that common pool effects may be diminishing a borrower’s ability to repay.
An increase of one loan will increase the borrower’s probability of repayment by about
Credit record variables: data compiled immediately before solicitation
Revolving loans The borrower’s number of revolving loans open at the time of the credit report
Installment loans The borrower’s number of installment loans
Number of mortgages The borrower’s number of mortgages
Loan balances The total non-mortgage balances the borrower has on all loans
Revolving balances The total balances on revolving loans
Non-revolving
balances
The balances on non-mortgage, non-revolving loans
Mortgage balances Total mortgage balances at the time of the credit report
Years on ?le The number of years on ?le with the credit reporting agency
Credit score A measure, compiled by the credit reporting agency, meant to capture the
borrower’s probability of default
State level variables
a
Unemployment rate
b
The unemployment rate of the borrower’s state of residence lagged by one year
Median home value The median home value in the borrower’s state
Transfer payments
b
The value of the average transfer (food stamps, housing subsidies, disability
and unemployment payments, Medicaid and TANF) to low income families
Self-employed
b
Percentage of workers who are self-employed
Manufacturing
b
Percentage of workers who work in manufacturing
Homestead
exemption
State homestead exemption
Anti-harassment
statute
A dummy that equals 1 if the state allows a private right of action against
harassing creditors
No garnishment A dummy that equals 1 if the state does not allow garnishment for ordinary
loans
County level variables
a
Unemployment rate The county unemployment rate, lagged by one year
Median home value The median home value in the borrower’s county
Percent black Percentage of residents in a borrower’s county that are black
Percent hispanic Percentage of residents in a borrower’s county that are hispanic
Percent divorced Percentage of adults in a borrower’s county that are divorced
Percent uninsured Percentage of residents that do not have insurance coverage
High school Percentage of adults who graduated from high school
Issuer variables: recorded by the credit card issuer at the beginning of the experiment period
Mailing Indicator variables for each round of mailing
Introductory rate Teaser interest rate offered in solicitation
Intro period duration Length in months of introductory period
Log income The log of the borrower’s self-reported income
Credit limit The borrower’s limit on the card
Notes:
a
If a borrower moved during the experiment period and defaulted, the borrower’s residence
was de?ned as the state and zip code where the default occurred; if the borrower moved and did not
default, the borrower’s residence was de?ned as the state where she lived longest during the
experiment period; aside from the state laws and unless noted otherwise, the data was merged from the
2000 Census based on the account’s zip code;
b
constructed from the 1996-2000 March supplement to
the Current Population Survey
Table III.
De?nitions
Creditors and
personal
bankruptcy
15
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
V
a
r
i
a
b
l
e
s
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
M
e
d
i
a
n
h
o
m
e
v
a
l
u
e
4
.
8
5
£
1
0
2
0
5
*
*
*
2
1
.
4
5
£
1
0
2
0
5
*
2
3
.
4
0
£
1
0
2
0
5
*
*
*
6
.
8
9
£
1
0
2
0
5
*
*
*
2
2
.
1
7
£
1
0
2
0
5
*
*
2
4
.
7
2
£
1
0
2
0
5
*
*
*
(
1
.
0
9
£
1
0
2
0
5
)
(
7
.
5
7
£
1
0
2
0
6
)
(
1
.
0
2
£
1
0
2
0
5
)
(
1
.
3
4
£
1
0
2
0
5
)
(
9
.
6
9
£
1
0
2
0
6
)
(
1
.
3
1
£
1
0
2
0
5
)
P
e
r
c
e
n
t
d
i
v
o
r
c
e
d
2
0
.
0
0
0
9
1
1
*
*
*
0
.
0
0
0
5
2
5
*
*
*
0
.
0
0
0
3
8
5
2
0
.
0
0
1
3
4
*
*
*
0
.
0
0
0
7
2
6
*
*
*
0
.
0
0
0
6
1
6
(
0
.
0
0
0
3
2
2
)
(
0
.
0
0
0
1
6
5
)
(
0
.
0
0
0
3
0
6
)
(
0
.
0
0
0
3
9
9
)
(
0
.
0
0
0
2
0
4
)
(
0
.
0
0
0
3
8
3
)
P
e
r
c
e
n
t
u
n
i
n
s
u
r
e
d
2
5
.
6
9
£
1
0
2
0
5
2
7
.
3
8
£
1
0
2
0
6
6
.
4
3
£
1
0
2
0
5
2
0
.
0
0
0
2
6
5
5
.
0
2
£
1
0
2
0
5
0
.
0
0
0
2
1
5
(
0
.
0
0
0
2
6
3
)
(
0
.
0
0
0
1
5
3
)
(
0
.
0
0
0
2
5
4
)
(
0
.
0
0
0
3
3
1
)
(
0
.
0
0
0
1
9
5
)
(
0
.
0
0
0
3
1
5
)
L
o
g
i
n
c
o
m
e
2
0
.
0
0
4
2
1
*
*
*
0
.
0
0
0
9
6
9
*
*
*
0
.
0
0
3
2
4
*
*
*
0
.
0
0
1
3
6
*
*
*
2
0
.
0
0
0
6
9
0
*
*
*
2
0
.
0
0
0
6
6
7
*
*
(
0
.
0
0
0
9
6
4
)
(
0
.
0
0
0
3
7
0
)
(
0
.
0
0
0
7
9
6
)
(
0
.
0
0
0
4
0
6
)
(
0
.
0
0
0
2
1
4
)
(
0
.
0
0
0
2
7
4
)
Y
e
a
r
s
o
n
?
l
e
0
.
0
0
2
0
5
*
*
2
0
.
0
0
1
8
1
*
*
2
0
.
0
0
0
2
4
1
0
.
0
0
8
3
0
*
*
*
2
0
.
0
0
4
5
0
*
*
*
2
0
.
0
0
3
8
0
*
*
*
(
0
.
0
0
0
9
0
5
)
(
0
.
0
0
0
7
2
7
)
(
0
.
0
0
0
6
3
7
)
(
0
.
0
0
1
2
4
)
(
0
.
0
0
0
9
1
8
)
(
0
.
0
0
0
9
4
5
)
C
r
e
d
i
t
s
c
o
r
e
0
.
0
0
0
1
1
1
*
*
*
2
3
.
4
0
£
1
0
2
0
5
*
*
*
2
7
.
7
2
£
1
0
2
0
5
*
*
*
(
5
.
2
0
£
1
0
2
0
6
)
(
3
.
4
0
£
1
0
2
0
6
)
(
3
.
8
2
£
1
0
2
0
6
)
C
r
e
d
i
t
l
i
m
i
t
3
.
6
5
£
1
0
2
0
6
*
*
*
2
1
.
4
9
£
1
0
2
0
6
*
*
*
2
2
.
1
6
£
1
0
2
0
6
*
*
*
(
2
.
2
0
£
1
0
2
0
7
)
(
1
.
3
9
£
1
0
2
0
7
)
(
1
.
7
6
£
1
0
2
0
7
)
I
n
t
r
o
d
u
c
t
o
r
y
p
e
r
i
o
d
d
u
r
a
t
i
o
n
0
.
0
0
0
4
3
5
*
*
*
2
0
.
0
0
0
2
7
5
*
*
*
2
0
.
0
0
0
1
6
0
0
.
0
0
0
7
9
8
*
*
*
2
0
.
0
0
0
4
3
6
*
*
*
2
0
.
0
0
0
3
6
2
*
*
(
0
.
0
0
0
1
6
6
)
(
6
.
6
5
£
1
0
2
0
5
)
(
0
.
0
0
0
1
3
2
)
(
0
.
0
0
0
2
1
1
)
(
7
.
9
9
£
1
0
2
0
5
)
(
0
.
0
0
0
1
7
8
)
I
n
t
r
o
d
u
c
t
o
r
y
r
a
t
e
2
0
.
0
0
2
5
1
*
*
0
.
0
0
1
8
3
*
*
*
0
.
0
0
0
6
7
2
2
0
.
0
0
5
6
6
*
*
*
0
.
0
0
3
1
5
*
*
*
0
.
0
0
2
5
1
*
(
0
.
0
0
1
1
5
)
(
0
.
0
0
0
4
1
2
)
(
0
.
0
0
1
0
4
)
(
0
.
0
0
1
4
6
)
(
0
.
0
0
0
5
1
9
)
(
0
.
0
0
1
3
8
)
H
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
(
$
1
0
,
0
0
0
s
)
2
0
.
0
0
0
5
6
4
2
2
.
4
1
£
1
0
2
0
5
0
.
0
0
0
5
8
9
*
2
0
.
0
0
0
4
6
4
2
0
.
0
0
0
1
2
3
0
.
0
0
0
5
8
7
(
0
.
0
0
0
4
1
6
)
(
0
.
0
0
0
2
7
2
)
(
0
.
0
0
0
3
2
4
)
(
0
.
0
0
0
5
1
4
)
(
0
.
0
0
0
3
5
7
)
(
0
.
0
0
0
3
8
0
)
A
n
t
i
-
h
a
r
a
s
s
m
e
n
t
s
t
a
t
u
t
e
0
.
0
0
1
7
5
0
.
0
0
1
0
6
2
0
.
0
0
2
8
1
*
*
0
.
0
0
2
5
7
0
.
0
0
1
2
9
2
0
.
0
0
3
8
7
*
*
*
(
0
.
0
0
1
3
6
)
(
0
.
0
0
0
6
9
6
)
(
0
.
0
0
1
2
5
)
(
0
.
0
0
1
6
0
)
(
0
.
0
0
0
8
8
0
)
(
0
.
0
0
1
4
8
)
N
o
g
a
r
n
i
s
h
m
e
n
t
2
0
.
0
0
3
7
4
*
*
0
.
0
0
3
9
1
*
*
*
2
0
.
0
0
0
1
7
3
2
0
.
0
0
5
0
9
*
*
0
.
0
0
4
9
5
*
*
*
0
.
0
0
0
1
3
7
(
0
.
0
0
1
5
3
)
(
0
.
0
0
1
0
7
)
(
0
.
0
0
1
3
4
)
(
0
.
0
0
1
9
8
)
(
0
.
0
0
1
4
6
)
(
0
.
0
0
1
5
5
)
L
o
a
n
b
a
l
a
n
c
e
s
2
2
.
9
1
£
1
0
2
0
8
2
2
.
3
7
£
1
0
2
0
8
5
.
2
9
£
1
0
2
0
8
*
*
*
2
1
.
4
1
£
1
0
2
0
7
*
*
3
.
5
1
£
1
0
2
0
8
1
.
0
6
£
1
0
2
0
7
*
*
*
(
4
.
8
9
£
1
0
2
0
8
)
(
4
.
3
5
£
1
0
2
0
8
)
(
1
.
6
5
£
1
0
2
0
8
)
(
5
.
5
5
£
1
0
2
0
8
)
(
2
.
5
8
£
1
0
2
0
8
)
(
3
.
3
1
£
1
0
2
0
8
)
N
u
m
b
e
r
o
f
l
o
a
n
s
2
9
.
9
4
£
1
0
2
0
5
2
0
.
0
0
0
4
2
0
*
*
*
0
.
0
0
0
5
1
9
*
*
*
0
.
0
0
0
4
3
3
*
*
*
2
0
.
0
0
0
8
0
3
*
*
*
0
.
0
0
0
3
7
0
*
*
*
(
9
.
4
5
£
1
0
2
0
5
)
(
9
.
4
5
£
1
0
2
0
5
)
(
8
.
5
5
£
1
0
2
0
5
)
(
0
.
0
0
0
1
3
4
)
(
0
.
0
0
0
1
0
6
)
(
0
.
0
0
0
1
1
3
)
O
b
s
e
r
v
a
t
i
o
n
s
5
2
,
7
2
1
5
2
,
7
2
1
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
p
,
0
.
1
,
*
*
p
,
0
.
0
5
a
n
d
*
*
*
p
,
0
.
0
1
;
s
t
a
n
d
a
r
d
e
r
r
o
r
s
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
a
n
d
a
r
e
c
l
u
s
t
e
r
e
d
b
y
s
t
a
t
e
;
e
a
c
h
s
p
e
c
i
?
c
a
t
i
o
n
i
n
T
a
b
l
e
s
I
V
-
V
I
a
l
s
o
c
o
n
t
r
o
l
l
e
d
f
o
r
t
h
e
s
t
a
t
e
a
n
d
c
o
u
n
t
y
d
e
m
o
g
r
a
p
h
i
c
v
a
r
i
a
b
l
e
s
l
i
s
t
e
d
i
n
T
a
b
l
e
I
I
I
;
s
p
e
c
i
?
c
a
t
i
o
n
s
w
i
t
h
o
u
t
t
h
e
s
e
c
o
n
t
r
o
l
s
y
i
e
l
d
e
d
s
i
m
i
l
a
r
r
e
s
u
l
t
s
;
f
u
l
l
r
e
s
u
l
t
s
a
r
e
a
v
a
i
l
a
b
l
e
u
p
o
n
r
e
q
u
e
s
t
Table IV.
Multinomial logit
results – marginal effects
JFEP
6,1
16
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
0.043 percentage points. Given that the average charged-off balance is $4,604, the mean
expected cost of each extra loan is about $2 per borrower in this sample. As we will see
below, the creditor’s expected cost is signi?cantly higher for borrowers who have
missed more than two consecutive payments.
C2. State laws. Most of the coef?cients on homestead exemptions in Tables IV-VI
fail the test of statistical signi?cance. The signs of the coef?cients are consistent with
the model, however: increasing the homestead exemption allowance decreases the
likelihood of repayment, increases the likelihood of formal bankruptcy, and has a small
negative effect on informal bankruptcy. The coef?cients on garnishment laws are also
consistent with the model, and have greater precision: borrowers living in states
without garnishment are signi?cantly less likely to repay, more likely to remain in
informal bankruptcy, and less likely to ?le a formal bankruptcy (though the coef?cient
on formal bankruptcy is not statistically signi?cant). Anti-harassment laws also
signi?cantly decrease the likelihood of formal bankruptcy.
The impact of harassment and garnishment laws seems to be substantial. For example,
the coef?cient in the second column of Table IV implies that moving a borrower from a
state without garnishment to a state where garnishment is permitted will decrease the
probability of informal bankruptcy by 0.39 percentage points, which is roughly equal to
the impact of increasing the borrower’s number of loans by 9. The same move to a state
that allows garnishment will increase the probability of repayment by 0.37 percentage
points, which is similar to the effect of increasing the borrower’s time on ?le by 1.8 years.
The ?rst three columns of Table V report the results of an estimation that included
interactions between state laws and the borrower’s number of loans. If state laws
effectively limit over-collection, then the borrower’s number of loans should have less
impact in states that restrict collection. So, for example, the garnishment and
anti-harassment interaction coef?cients would be positive for informal bankruptcy and
negative for formal bankruptcy if borrowers living in states that restrict collection are
less likely to substitute formal bankruptcy for informal bankruptcy as number of loans
increase. The signs of these coef?cients do not follow this pattern and are statistically
insigni?cant. These results do not support the hypothesis that garnishment and
anti-harassment laws limit the impact of common pool effects.
In the model presented in IIIB where creditor’s collection increased repayment on all
loans, exemptions had no effect on each creditor’s optimal collection level. However,
in the extension presented in Section III(C), increasing the exemption level not only
increased total collection, the impact of number of loans on total collection also increased
with the exemption level. This would imply that there should be greater substitution
towards repayment and formal bankruptcy and away from informal bankruptcy in
states with higher homestead exemptions. Instead, the coef?cients on the homestead
interaction terms for formal and informal bankruptcy are statistically insigni?cant,
and the coef?cient for repayment is negative rather than positive. Again, the homestead
exemption laws do not appear to counteract the impact of common pool effects.
C3. Delinquent borrowers. The credit negotiation model focuses on distressed
borrowers. For that reason, the last three columns of Table V report the results of a
speci?cation that only includes borrowers who were at least two months delinquent;
borrowers are designated as having chosen “repayment” if they return to good
standing during the observation period. In this speci?cation, increasing a borrower’s
number of loans again signi?cantly increases the probability of formal and
Creditors and
personal
bankruptcy
17
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A
l
l
b
o
r
r
o
w
e
r
s
D
e
l
i
n
q
u
e
n
t
b
o
r
r
o
w
e
r
s
o
n
l
y
V
a
r
i
a
b
l
e
s
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
L
o
g
i
n
c
o
m
e
2
0
.
0
0
4
2
0
*
*
*
0
.
0
0
0
9
6
0
*
*
*
0
.
0
0
3
2
4
*
*
*
2
0
.
0
3
6
4
*
*
*
0
.
0
0
6
9
1
0
.
0
2
9
5
*
*
*
(
0
.
0
0
0
9
6
5
)
(
0
.
0
0
0
3
6
4
)
(
0
.
0
0
0
7
9
9
)
(
0
.
0
0
9
6
0
)
(
0
.
0
0
5
2
1
)
(
0
.
0
0
9
0
6
)
Y
e
a
r
s
o
n
?
l
e
0
.
0
0
2
0
1
*
*
2
0
.
0
0
1
7
9
*
*
2
0
.
0
0
0
2
2
7
2
0
.
0
1
1
5
2
0
.
0
1
4
2
0
.
0
2
5
6
*
*
(
0
.
0
0
0
9
0
0
)
(
0
.
0
0
0
7
1
6
)
(
0
.
0
0
0
6
3
5
)
(
0
.
0
1
6
5
)
(
0
.
0
1
2
8
)
(
0
.
0
1
1
6
)
C
r
e
d
i
t
s
c
o
r
e
0
.
0
0
0
1
1
1
*
*
*
2
3
.
3
9
£
1
0
2
0
5
*
*
*
2
7
.
7
1
£
1
0
2
0
5
*
*
*
0
.
0
0
0
4
3
9
*
*
*
2
7
.
6
5
£
1
0
2
0
5
2
0
.
0
0
0
3
6
2
*
*
*
(
5
.
2
1
£
1
0
2
0
6
)
(
3
.
3
6
£
1
0
2
0
6
)
(
3
.
8
2
£
1
0
2
0
6
)
(
9
.
8
7
£
1
0
2
0
5
)
(
5
.
4
5
£
1
0
2
0
5
)
(
8
.
2
1
£
1
0
2
0
5
)
C
r
e
d
i
t
l
i
m
i
t
3
.
6
3
£
1
0
2
0
6
*
*
*
2
1
.
4
8
£
1
0
2
0
6
*
*
*
2
2
.
1
5
£
1
0
2
0
6
*
*
*
2
.
4
3
£
1
0
2
0
5
*
*
*
2
1
.
2
4
£
1
0
2
0
5
*
*
*
2
1
.
1
9
£
1
0
2
0
5
*
*
*
(
2
.
2
1
£
1
0
2
0
7
)
(
1
.
4
3
£
1
0
2
0
7
)
(
1
.
7
4
£
1
0
2
0
7
)
(
3
.
0
6
£
1
0
2
0
6
)
(
2
.
7
2
£
1
0
2
0
6
)
(
2
.
1
3
£
1
0
2
0
6
)
H
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
(
$
1
0
,
0
0
0
s
)
2
0
.
0
0
0
1
7
9
2
0
.
0
0
0
2
6
4
0
.
0
0
0
4
4
4
2
0
.
0
0
4
4
6
2
0
.
0
0
4
2
1
0
.
0
0
8
6
7
*
(
0
.
0
0
0
4
6
4
)
(
0
.
0
0
0
3
2
3
)
(
0
.
0
0
0
3
6
1
)
(
0
.
0
0
6
2
9
)
(
0
.
0
0
5
4
8
)
(
0
.
0
0
4
6
5
)
A
n
t
i
-
h
a
r
a
s
s
m
e
n
t
s
t
a
t
u
t
e
0
.
0
0
3
8
4
0
.
0
0
0
1
4
8
2
0
.
0
0
3
9
9
*
0
.
0
2
3
1
0
.
0
2
2
5
*
2
0
.
0
4
5
7
*
*
*
(
0
.
0
0
2
3
4
)
(
0
.
0
0
1
2
1
)
(
0
.
0
0
2
1
1
)
(
0
.
0
1
7
1
)
(
0
.
0
1
3
5
)
(
0
.
0
1
6
3
)
N
o
g
a
r
n
i
s
h
m
e
n
t
2
0
.
0
0
3
6
9
*
*
0
.
0
0
5
0
0
*
*
*
2
0
.
0
0
1
3
1
2
0
.
0
5
9
2
*
*
*
0
.
0
7
5
8
*
*
*
2
0
.
0
1
6
6
(
0
.
0
0
1
8
3
)
(
0
.
0
0
1
3
6
)
(
0
.
0
0
1
4
3
)
(
0
.
0
2
0
7
)
(
0
.
0
2
0
2
)
(
0
.
0
1
9
7
)
L
o
a
n
b
a
l
a
n
c
e
s
2
2
.
8
6
£
1
0
2
0
8
2
2
.
4
0
£
1
0
2
0
8
5
.
2
6
£
1
0
2
0
8
*
*
*
2
5
.
8
6
£
1
0
2
0
7
2
1
.
3
1
£
1
0
2
0
6
1
.
9
0
£
1
0
2
0
6
*
*
*
(
4
.
8
5
£
1
0
2
0
8
)
(
4
.
3
2
£
1
0
2
0
8
)
(
1
.
6
5
£
1
0
2
0
8
)
(
1
.
3
2
£
1
0
2
0
6
)
(
9
.
7
1
£
1
0
2
0
7
)
(
7
.
1
6
£
1
0
2
0
7
)
N
u
m
b
e
r
o
f
l
o
a
n
s
0
.
0
0
0
1
7
0
2
0
.
0
0
0
5
5
9
*
*
*
0
.
0
0
0
3
8
9
*
*
2
0
.
0
0
5
1
3
*
*
*
2
0
.
0
0
5
2
5
*
*
*
0
.
0
1
0
4
*
*
*
(
0
.
0
0
0
1
9
2
)
(
0
.
0
0
0
1
7
8
)
(
0
.
0
0
0
1
8
3
)
(
0
.
0
0
1
5
3
)
(
0
.
0
0
1
6
6
)
(
0
.
0
0
1
5
6
)
L
o
a
n
s
*
h
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
2
6
.
9
4
£
1
0
2
0
5
*
4
.
8
5
£
1
0
2
0
5
2
.
0
9
£
1
0
2
0
5
(
3
.
9
8
£
1
0
2
0
5
)
(
3
.
2
8
£
1
0
2
0
5
)
(
2
.
8
9
£
1
0
2
0
5
)
L
o
a
n
s
*
a
n
t
i
-
h
a
r
a
s
s
m
e
n
t
2
0
.
0
0
0
3
3
4
0
.
0
0
0
1
8
9
0
.
0
0
0
1
4
5
(
0
.
0
0
0
2
2
9
)
(
0
.
0
0
0
1
8
8
)
(
0
.
0
0
0
1
7
6
)
L
o
a
n
s
*
g
a
r
n
i
s
h
m
e
n
t
1
.
4
5
£
1
0
2
0
5
2
0
.
0
0
0
1
5
8
0
.
0
0
0
1
4
4
(
0
.
0
0
0
1
7
1
)
(
0
.
0
0
0
1
2
4
)
(
0
.
0
0
0
1
0
7
)
O
b
s
e
r
v
a
t
i
o
n
s
5
2
,
7
2
1
4
,
6
6
9
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
p
,
0
.
1
,
*
*
p
,
0
.
0
5
a
n
d
*
*
*
p
,
0
.
0
1
;
s
t
a
n
d
a
r
d
e
r
r
o
r
s
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
a
n
d
a
r
e
c
l
u
s
t
e
r
e
d
b
y
s
t
a
t
e
;
e
a
c
h
s
p
e
c
i
?
c
a
t
i
o
n
a
l
s
o
i
n
c
l
u
d
e
d
a
l
l
o
f
t
h
e
c
o
n
t
r
o
l
v
a
r
i
a
b
l
e
s
l
i
s
t
e
d
i
n
T
a
b
l
e
I
V
Table V.
Multinomial logit
results – marginal effects
(standard errors
in parentheses)
JFEP
6,1
18
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
informal bankruptcy. The repayment coef?cient is negative and statistically
signi?cant, and it indicates that an additional loan decreases the probability a
delinquent borrower will repay by 0.513 percentage points. Thus, the addition of one
more loan to the record of a delinquent borrower increases the creditor’s expected
charged-off balance by $23.62.
C4. Revolving and non-revolving loans. The ?rst three columns of Table VI report
the results of a speci?cation that separates revolving and non-revolving loan balances
and number of loans. These results show that revolving balances have a signi?cantly
positive effect on formal and informal bankruptcy, while non-revolving balances have
negative effects on both informal and formal bankruptcy. These results suggest that
high non-revolving balances might indicate a less risky borrower and revolving loans
may imply the opposite. Note, however, that the number of loans coef?cients do not
follow this pattern, and instead correspond to our common pool model: higher numbers
of revolving and non-revolving loans both signi?cantly increase the probability of
formal bankruptcy and decrease the probability of informal bankruptcy.
The common pool model addresses the strategic problem faced by lenders of
unsecured loans. Note that, while revolving loans are almost always unsecured, and
installment loans may or may not be secured, mortgages are secured by a valuable
asset. Thus, mortgage lenders, because they have the option to foreclose, have less
incentive to react to the possibility of losing attachable assets to other lenders. Though
the ?nal three columns of Table VI show that increasing a borrower’s number of
mortgages results in similar coef?cient estimates as increasing other types of loans,
these coef?cient estimates are not statistically signi?cant.
V Conclusion
Correcting for strategic over-collection has been cited as a central tenant of the US
business bankruptcy system. This paper provides the ?rst systematic examination of
collection externalities in personal bankruptcy, and the results suggest that the common
pool effects are real phenomena. The credit card data utilized in this paper allowed an
evaluation of the distinct impacts of number of loans oninformal andformal bankruptcy.
If number of loans were simply an indicator of credit-worthiness, increasing number of
loans should either decrease or increase both informal and formal bankruptcy. However,
increasing a borrower’s number of loans consistently increases her probability of
choosing formal bankruptcy while decreasing the likelihood she chooses informal
bankruptcy, a result that is in accordance with the common pool model of collection. The
effects are not only statistically signi?cant, but also economically signi?cant, potentially
contributing dramatically to the outcome a distressed borrower experiences.
This paper ?nds strong evidence that anti-harassment, garnishment, and exemption
laws have an impact on a borrower’s choice among informal bankruptcy, formal
bankruptcy, and repayment. However, there is little evidence that these laws mitigate
the effects of creditor externalities. Unfortunately, these laws change very rarely, so this
paper, like most investigations into these laws’ effects, relies on cross-state variation,
and the small sample size may prevent a complete accounting of the laws’ impacts.
Personal bankruptcy law faced a major overhaul in via the enactment of BAPCPA in
2005, which, along with the Dodd-Frank Wall Street Reform and Consumer Protection
Act of 2010, signi?cantly changed the consumer borrowing landscape in ways that
could alleviate or exacerbate collection externalities. For example, Dodd-Frank created
Creditors and
personal
bankruptcy
19
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
V
a
r
i
a
b
l
e
s
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
H
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
(
$
1
0
,
0
0
0
s
)
2
0
.
0
0
0
3
8
8
2
4
.
8
3
£
1
0
2
0
5
0
.
0
0
0
4
3
6
2
0
.
0
0
0
4
0
2
2
4
.
4
6
£
1
0
2
0
5
0
.
0
0
0
4
4
7
(
0
.
0
0
0
3
7
9
)
(
0
.
0
0
0
2
5
3
)
(
0
.
0
0
0
2
9
8
)
(
0
.
0
0
0
3
8
3
)
(
0
.
0
0
0
2
6
9
)
(
0
.
0
0
0
2
9
8
)
A
n
t
i
-
h
a
r
a
s
s
m
e
n
t
s
t
a
t
u
t
e
0
.
0
0
1
5
2
0
.
0
0
0
9
8
5
2
0
.
0
0
2
5
1
*
*
0
.
0
0
1
5
0
0
.
0
0
1
0
5
2
0
.
0
0
2
5
6
*
*
(
0
.
0
0
1
2
7
)
(
0
.
0
0
0
6
4
7
)
(
0
.
0
0
1
1
5
)
(
0
.
0
0
1
2
6
)
(
0
.
0
0
0
6
8
9
)
(
0
.
0
0
1
1
5
)
N
o
g
a
r
n
i
s
h
m
e
n
t
2
0
.
0
0
3
1
4
*
*
0
.
0
0
3
5
2
*
*
*
2
0
.
0
0
0
3
8
0
2
0
.
0
0
3
4
5
*
*
0
.
0
0
3
8
3
*
*
*
2
0
.
0
0
0
3
8
2
(
0
.
0
0
1
3
9
)
(
0
.
0
0
1
0
1
)
(
0
.
0
0
1
1
6
)
(
0
.
0
0
1
4
1
)
(
0
.
0
0
1
0
6
)
(
0
.
0
0
1
1
6
)
R
e
v
o
l
v
i
n
g
b
a
l
a
n
c
e
s
2
9
.
6
6
£
1
0
2
0
7
*
*
*
1
.
9
1
£
1
0
2
0
7
*
7
.
7
5
£
1
0
2
0
7
*
*
*
2
9
.
2
9
£
1
0
2
0
7
*
*
*
1
.
6
9
£
1
0
2
0
7
7
.
6
0
£
1
0
2
0
7
*
*
*
(
1
.
3
9
£
1
0
2
0
7
)
(
1
.
0
2
£
1
0
2
0
7
)
(
6
.
4
2
£
1
0
2
0
8
)
(
1
.
4
2
£
1
0
2
0
7
)
(
1
.
0
7
£
1
0
2
0
7
)
(
6
.
3
3
£
1
0
2
0
8
)
N
o
n
-
r
e
v
o
l
v
i
n
g
b
a
l
a
n
c
e
s
1
.
4
8
£
1
0
2
0
7
*
*
*
2
8
.
7
9
£
1
0
2
0
8
*
*
2
6
.
0
6
£
1
0
2
0
8
1
.
4
1
£
1
0
2
0
7
*
*
*
2
8
.
4
1
£
1
0
2
0
8
*
2
5
.
7
1
£
1
0
2
0
5
(
5
.
3
5
£
1
0
2
0
8
)
(
4
.
0
9
£
1
0
2
0
8
)
(
4
.
0
1
£
1
0
2
0
8
)
(
5
.
4
0
£
1
0
2
0
8
)
(
4
.
4
0
£
1
0
2
0
5
)
(
4
.
0
4
£
1
0
2
0
8
)
N
u
m
b
e
r
r
e
v
o
l
v
i
n
g
l
o
a
n
s
2
0
.
0
0
0
2
6
1
*
2
0
.
0
0
0
3
6
6
*
*
*
0
.
0
0
0
6
2
6
*
*
*
2
0
.
0
0
0
2
2
0
2
0
.
0
0
0
4
0
6
*
*
*
0
.
0
0
0
6
2
6
*
*
*
(
0
.
0
0
0
1
4
8
)
(
0
.
0
0
0
1
0
7
)
(
0
.
0
0
0
1
2
3
)
(
0
.
0
0
0
1
5
8
)
(
0
.
0
0
0
1
2
1
)
(
0
.
0
0
0
1
2
3
)
N
u
m
b
e
r
i
n
s
t
a
l
l
m
e
n
t
l
o
a
n
s
2
7
.
2
7
£
1
0
2
0
5
2
0
.
0
0
0
2
7
0
*
0
.
0
0
0
3
4
2
*
*
*
0
.
0
0
0
2
1
7
2
0
.
0
0
0
4
4
2
*
*
*
0
.
0
0
0
2
2
6
*
*
(
0
.
0
0
0
1
5
7
)
(
0
.
0
0
0
1
4
0
)
(
0
.
0
0
0
1
0
7
)
(
0
.
0
0
0
1
7
2
)
(
0
.
0
0
0
1
5
7
)
(
0
.
0
0
0
1
0
3
)
M
o
r
t
g
a
g
e
b
a
l
a
n
c
e
s
6
.
9
6
£
1
0
2
0
8
*
*
*
2
5
.
6
2
£
1
0
2
0
8
*
*
*
2
1
.
3
4
£
1
0
2
0
8
(
2
.
0
0
£
1
0
2
0
8
)
(
2
.
1
2
£
1
0
2
0
8
)
(
1
.
0
1
£
1
0
2
0
8
)
N
u
m
b
e
r
o
f
m
o
r
t
g
a
g
e
s
0
.
0
0
0
3
8
3
0
.
0
0
0
9
6
2
2
0
.
0
0
1
3
4
(
0
.
0
0
1
6
2
)
(
0
.
0
0
1
2
2
)
(
0
.
0
0
0
8
2
4
)
O
b
s
e
r
v
a
t
i
o
n
s
5
2
,
7
2
1
5
2
,
7
2
1
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
p
,
0
.
1
,
*
*
p
,
0
.
0
5
a
n
d
*
*
*
p
,
0
.
0
1
;
s
t
a
n
d
a
r
d
e
r
r
o
r
s
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
a
n
d
a
r
e
c
l
u
s
t
e
r
e
d
b
y
s
t
a
t
e
;
e
a
c
h
s
p
e
c
i
?
c
a
t
i
o
n
a
l
s
o
i
n
c
l
u
d
e
d
t
h
e
c
o
n
t
r
o
l
v
a
r
i
a
b
l
e
s
l
i
s
t
e
d
i
n
T
a
b
l
e
I
V
Table VI.
Multinomial logit
results – marginal effects
JFEP
6,1
20
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the Consumer Financial Protection Bureau, and BAPCPA includes a requirement that
debtors receive counseling from an agency approved by the bankruptcy administrator
or US trustee. If counseling facilitates a negotiated settlement with competing creditors,
this requirement could reduce the rush to collect that defaulting borrowers often
experience and diminish the incentive of borrowers with multiple creditors to opt for
formal bankruptcy. In addition, BAPCPA decreased federal property and homestead
exemptions, a change that could also mitigate the effects of the negative externality that
resulted from property seizure. A fruitful avenue for future research will be measuring
whether these legal changes signi?cantly altered the impact of externalities on creditor
and borrower behavior.
Notes
1. According to The Preliminary Report on Chapter 7 Asset Cases 1994 to 2000 (2001), “ the vast
majority (about 95 to 97 percent) of Chapter 7 cases yield no assets.” Chapter 7 bankruptcy is
the most common type of personal bankruptcy; more than 70 percent of non-business
bankruptcies were ?led under Chapter 7 during the late 1990s, the observation period of this
paper.
2. The data used in this paper was collected over a period of three years, during which only a
few states experienced any changes in laws governing bankruptcy and garnishment. As a
result, the empirical model is able to utilize only cross-state variation in these laws.
Garnishment law appears to be a signi?cant factor in cross-state variation in bankruptcy
rates: see Lefgren and McIntyre (2009) for an overview.
3. Brunner and Krahnen (2004) ?nd that German ?rms are more likely to negotiate a workout
and avoid bankruptcy when they have borrowed from fewer banks, facilitating lender
coordination. Williams (1998) ?nds evidence that credit counseling services mitigate
destructive creditor collection.
4. However, a preference granted soon before the borrower ?les may be nulli?ed during a
bankruptcy proceeding.
5. Since the passage of the Federal Debt Collection Procedures Practices Act in 1977,
garnishment for the collection of federal debts is determined by federal law, and therefore
legal in all states.
6. The observations in our dataset were collected between 1995 and 1997, and so the entries in
Table I re?ect exemption levels in 1997.
7. For a discussion of these violations, see Schulman (1985) and Araki (1995).
8. See Dawsey et al. (2013) for a summary of these laws.
9. Individuals with substantial debts and assets can choose to ?le under Chapter 11, and family
farmers can ?le a Chapter 12 reorganization. These chapters are rarely used by individuals
and will not be considered here.
10. Bankruptcy law underwent a major overhaul in 2005 with the enactment of the Bankruptcy
Abuse Prevention and Consumer Protection Act (BAPCPA). BAPCPA instituted means
testing and mandatory credit counseling and other hurdles to personal bankruptcy
(and Chapter 7 bankruptcy in particular). The data in this paper was collected from 1995
through 1997 and predates these changes.
11. See Scott (1989) for a discussion of the literature on lost value.
12. The administrative costs of a formal bankruptcy ?ling might include attorney fees and
?ling fees, as well as less quanti?able costs associated with social stigma or reduced ability
Creditors and
personal
bankruptcy
21
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
to borrow. In 2005, after the observation period of this paper but previous to signi?cant
changes in US bankruptcy law that increased the cost of ?ling, the GAO estimated that for a
Chapter 7 personal bankruptcy ?ling, the average attorney’s fee was $712 and the average
?ling fee was $209 (GAO, 2008). And in a small study of bankrupt married couples, 35 of 37
participants statedthat “they felt shame and stigmatization as theynegotiated the bankruptcy
process.” (Thorne and Anderson, 2006). An individual’s ?xed cost of an informal bankruptcy
would likely also include some stigma costs, but because the borrower does not make a public
declaration of bankruptcy through the legal system, these costs would likely be less than what
would follow a formal bankruptcy. The borrower in informal bankruptcy would also face
some future borrowing constraints, along with the disutility of engaging with or avoiding
collections agents.
13. In this model, the size of the loan is given, though creditors clearly take account of borrower
characteristics, including number of lenders, in assessing credit-worthiness and assigning the
terms of a credit offer. In theory, a creditor would anticipate the negative externality of
collection and therefore lend less when a borrower has many loans. However, a lender,
particularly a lender of revolving debt, balances the risk of default or bankruptcy against the
potential for fee and interest rate revenue, and a borrower with many loans may be pro?table
in the short run even if bankruptcy eventually results. Rather than modeling the complex loan
determination process, the econometric speci?cations control, for the lender’s assessment of
the borrower’s credit-worthiness using credit score and credit limit. See Miller (2011) for an
example of a two-sided model of both default and a creditor’s choice of loan and interest rate.
14. A suf?cient condition for this assumption would be that L is above a minimum threshold.
15. If all of the borrower’s assets are exempt, then, given our assumption that these borrowers
choose formal bankruptcy, the creditor receives 2 K
B
. Thus, the creditor’s payoff when
E . y does not depend on g, which removes the very low asset borrowers from the creditor’s
strategic problem that is the focus of this model.
16. See Dawsey et al. (2013) for a more detailed description of anti-harassment statutes. Other
speci?cations of state garnishment laws (e.g. using percentage garnishment or a dummy
that indicated if garnishments were restricted to less than 25 percent) yielded similar results.
State exemption laws generally include both a homestead exemption and a personal
property exemption. While the legal value of homestead exemptions are, to a large degree,
clearly speci?ed in state laws, personal exemptions are often simply described in general
terms. Estimates of the value of personal property exemptions failed to produce statistically
signi?cant results or to alter the main results of this article, and were therefore excluded
from the speci?cations reported in the tables that follow. Complete results are available upon
request.
17. According to the of?cial statistics of the Administrative Of?ce of the United States Courts,
the number of personal bankruptcies increased by nearly 73 percent between 1994 and 1997.
References
Araki, L.A.S. (1995), “Rx for abusive debt collection practices: amend the FDCPA”, University of
Hawaii Law Review, Vol. 17, pp. 69-111.
Baird, D.G. (1998), “Bankruptcy’s uncontested axioms”, Yale Law Journal, Vol. 108 No. 3,
pp. 573-599.
Block-Leib, S. (1993), “Fishing in muddy waters: clarifying the common pool analogy as applied
to the standard for commencement of a bankruptcy case”, American University Law
Review, Vol. 42 No. 2, pp. 337-431.
JFEP
6,1
22
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Brunner, A. and Krahnen, J.P. (2004), “Multiple lenders and corporate distress: evidence on debt
restructuring”, CEPR Discussion Paper 4287.
Dawsey, A. and Ausubel, L.M. (2004), Informal Bankruptcy, University of Maryland, College
Park, MD, September.
Dawsey, A., Hynes, R.M. and Ausubel, L.M. (2013), “Non-judicial debt collection and the
consumer’s choice among repayment, bankruptcy, and informal bankruptcy”, American
Bankruptcy Law Journal, Vol. 87 No. 1, pp. 1-26.
GAO (2008), Bankruptcy Reform: Dollar Costs Associated with the Bankruptcy Abuse Prevention
and Consumer Protection Act of 2005, United States Government Accountability Of?ce,
Washington, DC.
Hynes, R.M. (2004), “Why (consumer) bankruptcy?”, Alabama Law Review, Vol. 56 No. 1,
pp. 121-179.
Jackson, T.H. (1986), The Logic and Limits of Bankruptcy Law, Harvard University Press,
Cambridge, MA.
Kripke, H. (1985), “Law and economics: measuring the economic ef?ciency of commercial
law in a vacuum of fact”, University of Pennsylvania Law Review, Vol. 133 No. 5,
pp. 929-985.
Lefgren, L. and McIntyre, F. (2009), “Explaining the puzzle of cross-state differences in
bankruptcy rates”, Journal of Law and Economics, Vol. 52 No. 2, pp. 367-393.
Miller, M.M. (2011), Who Files for Bankruptcy? State Laws and the Characteristics of Bankrupt
Households, The State University of New Jersey, Rutgers, December.
Schulman, D.A. (1985), “The effectiveness of the federal fair debt collection practices act
(FDCPA)”, Emory Bankruptcy Developments Journal, Vol. 2, pp. 171-199.
Scott, R.E. (1989), “Rethinking the regulation of coercive creditor remedies”, Columbia Law
Review, Vol. 89 No. 4, pp. 730-788.
Warren, E. (1987), “Bankruptcy policy”, University of Chicago Law Review, Vol. 54 No. 3,
pp. 775-814.
Warren, E. (1993), “Bankruptcy policymaking in an imperfect world”, Michigan Law Review,
Vol. 92 No. 2, pp. 336-387.
Williams, W. (1998), Games Creditors Play: Collecting from Overextended Consumers, Carolina
Academic Press, Durham, NC.
Zywicki, T.J. (2005), “An economic analysis of the consumer bankruptcy crisis”, Northwestern
University Law Review, Vol. 99 No. 4, pp. 1463-1541.
Further reading
Ausubel, L.M. (1991), “The failure of competition in the credit card market”, American Economic
Review, Vol. 81 No. 1, pp. 50-81.
Ausubel, L.M. (1997), “Credit card defaults, credit card pro?ts, and bankruptcy”, American
Bankruptcy Law Journal, Vol. 71, pp. 249-270.
Ausubel, L.M. (1999), Adverse Selection in the Credit Card Market, University of Maryland,
College Park, MD, June.
Baird, D.G. and Jackson, T.H. (1983), “Possession and ownership: an examination of the scope of
Article 9”, Stanford Law Review, Vol. 35 No. 2, pp. 175-212.
Barth, J.R., Cordes, J.J. and Yezer, A.M.J. (1986), “Bene?ts and costs of legal restrictions on
personal loan markets”, Journal of Law & Economics, Vol. 25, pp. 357-380.
Creditors and
personal
bankruptcy
23
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Domowitz, I. and Sartain, R.L. (1999), “Determinants of the consumer bankruptcy decision”,
Journal of Finance, Vol. 54 No. 1, pp. 403-420.
Fishman, C. (1998), “Deconstructing the debt head”, American Demographics, Vol. 20 No. 12,
pp. 60-65.
Gross, K. (1997), Failure and Forgiveness: Rebalancing the Bankrutpcy System, Yale University
Press, New Haven, CT.
Hausman, J. and McFadden, D. (1984), “Speci?cation tests for the multinomial logit model”,
Econometrica, Vol. 52 No. 5, pp. 1219-1240.
Hensher, D.A. (1986), “Sequential and full information estimation maximum likelihood
estimation of a nested logit model”, The Review of Economics and Statistics, Vol. 68 No. 4,
pp. 657-667.
Hynes, R.M. (2008), “Broke but not bankrupt: consumer debt collection in state court”, Florida
Law Review, Vol. 60 No. 1, pp. 1-62.
Korobkin, D.R. (1991), “Rehabilitating values: a jurisprudence of bankruptcy”, Columbia Law
Review, Vol. 91 No. 4, pp. 717-789.
Lawrence, D.B. (1992), Handbook of Consumer Lending, Prentice-Hall, Englewood Cliffs, NJ.
Letsou, P.V. (1995), “The political economy of consumer credit regulation”, Emory Law Journal,
Vol. 44 No. 2, pp. 587-677.
Maddala, G.S. (1983), Limited-Dependent and Qualitative Variables in Econometrics, Cambridge
University Press, Cambridge.
Moss, D.A. and Johnson, G.A. (1999), “The rise of consumer bankruptcy: evolution, revolution, or
both?”, American Bankruptcy Law Journal, Vol. 73 No. 1, pp. 311-351.
Nelson, J.P. (1999), “Consumer bankruptcy and chapter choice: state panel evidence”,
Contemporary Economic Policy, Vol. 17 No. 4, pp. 552-566.
Peterson, R.L. and Aoki, K. (1984), “Bankruptcy ?lings before and after implementation of the
bankruptcy reform law”, Journal of Economics and Business, Vol. 36 No. 1, pp. 95-105.
Train, K. (1993), Qualitative Choice Analysis, The MIT Press, Cambridge, MA.
Warren, C. (1935), Bankruptcy inUnited States History, HarvardUniversityPress, Cambridge, MA.
Corresponding author
Amanda E. Dawsey can be contacted at: [email protected]
JFEP
6,1
24
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Ilona Alisauskaite-Seskiene, Rita Remeikiene, Ligita Gaspareniene. 2015. The Factors that Determine
Physical Entities’ Borrowing: Lithuanian Case. Procedia Economics and Finance 26, 616-622. [CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_780941622.pdf
The purpose of this paper is to explore the impact of creditors’ undervaluing the total
expected cost of a borrower’s bankruptcy filing because a portion of the cost will be borne by other
lenders. Creditors who bear a smaller portion of the total cost of a personal bankruptcy would be
expected to take less care to avoid triggering one
Journal of Financial Economic Policy
Externalities among creditors and personal bankruptcy
Amanda E. Dawsey
Article information:
To cite this document:
Amanda E. Dawsey , (2014),"Externalities among creditors and personal bankruptcy", J ournal of Financial
Economic Policy, Vol. 6 Iss 1 pp. 2 - 24
Permanent link to this document:http://dx.doi.org/10.1108/J FEP-09-2013-0037
Downloaded on: 24 January 2016, At: 21:48 (PT)
References: this document contains references to 38 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 353 times since 2014*
Users who downloaded this article also downloaded:
Wendy Ming-Yen Teoh, Siong-Choy Chong, Shi Mid Yong, (2013),"Exploring the factors influencing credit
card spending behavior among Malaysians", International J ournal of Bank Marketing, Vol. 31 Iss 6 pp.
481-500http://dx.doi.org/10.1108/IJ BM-04-2013-0037
Iftikhar Hussain, (2002),"Macroeconomic determinants of personal bankruptcies", Managerial Finance, Vol.
28 Iss 6 pp. 20-33http://dx.doi.org/10.1108/03074350210767898
Zafar U. Ahmed, Ishak Ismail, M. Sadiq Sohail, Ibrahim Tabsh, Hasbalaila Alias, (2010),"Malaysian
consumers' credit card usage behavior", Asia Pacific J ournal of Marketing and Logistics, Vol. 22 Iss 4 pp.
528-544http://dx.doi.org/10.1108/13555851011090547
Access to this document was granted through an Emerald subscription provided by emerald-srm:115632 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Externalities among creditors
and personal bankruptcy
Amanda E. Dawsey
Department of Economics, The University of Montana,
Missoula, Montana, USA
Abstract
Purpose – The purpose of this paper is to explore the impact of creditors’ undervaluing the total
expected cost of a borrower’s bankruptcy ?ling because a portion of the cost will be borne by other
lenders. Creditors who bear a smaller portion of the total cost of a personal bankruptcy would be
expected to take less care to avoid triggering one.
Design/methodology/approach – This paper presents a theoretical model of a creditor’s decision
of how aggressively to pursue collection. The model shows that because each lender’s collection
actions increase the probability of bankruptcy, each lender will collect more aggressively when a
borrower has many loans. The paper tests the predictions of the model using a large dataset of credit
card accounts.
Findings – The model highlights an important testable result: holding the level of debt constant,
a borrower with many loans is more likely to choose formal bankruptcy and less likely to choose
informal bankruptcy, i.e. chronic non-repayment absent a bankruptcy ?ling. This paper ?nds evidence
that strongly supports the predictions of the model. Laws that limit creditor collection actions do not
appear to mitigate the effects of increasing number of loans.
Originality/value – While a few papers have tested whether strategic interactions may impact
business bankruptcy, no paper of which the author is aware has provided clear empirical evidence of
the existence of common pool effects in the personal credit market. These effects point to an important
and potentially underappreciated source of risk for borrowers and creditors in this market.
Keywords Bankruptcy, Consumer protection, Financial policy, Informal institutions, Garnishment law,
Debt collections, Credit card debt, Default risk, Property exemptions
Paper type Research paper
I. Introduction
A major factor determining whether a borrower repays her loan, declares bankruptcy,
or remains in long-term default is how aggressively her creditors pursue repayment
through garnishment, property liens, or other collection methods. Bankruptcy is likely to
yield very little to a creditor aside from any collateral securing his loan[1], and the more
a creditor collects, the more likely a borrower is to ?le for bankruptcy to halt it. Thus,
creditors, particularly those with unsecured or under-secured loans, must account for the
risk of bankruptcy when deciding whether to use aggressive collection techniques.
However, whena borrower has borrowedfrommore thanone lender, the increased risk of
bankruptcyis sharedbyall creditors. Borrowers canbecome the objects of their creditors’
common pool problem: because each lender might lose out to competing creditors, they
seize property and garnish wages rather than using more cooperative strategies.
The common pool concept provides a useful framework for understanding the impact
of credit on personal bankruptcy. Borrowing and bankruptcy are clearly linked, and
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-6385.htm
JEL classi?cation – k35, g21, c72
Journal of Financial Economic Policy
Vol. 6 No. 1, 2014
pp. 2-24
qEmerald Group Publishing Limited
1757-6385
DOI 10.1108/JFEP-09-2013-0037
JFEP
6,1
2
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
a substantial economic literature examines whether increasing debt results in a greater
probability of personal bankruptcy, likely through increased exposure to risk
(Zywicki, 2005). The common pool model highlights an additional channel through
which credit might impact bankruptcy: borrowers with many creditors will experience
more aggressive collection than those with one or a few creditors, independent of the
total debt level. And so increasing debt levels, if accompanied by increasing number of
lenders, may have a two-fold effect: a direct effect (increased debt burden decreases
repayment) and an indirect effect (increased number of lenders leads to increased
collection activity, which leads to bankruptcy).
One way to disentangle the two potential effects is to compare the likelihood of
bankruptcy to that of “informal bankruptcy,” de?ned as non-repayment without a
formal bankruptcy ?ling (Dawsey and Ausubel, 2004). Note ?rst that the direct effect of
increasing a borrower’s debt burden will be to increase the likelihood a borrower chooses
either informal or formal bankruptcy rather than repayment. But the indirect effect will
result in a greater likelihood of formal bankruptcy but a decreased likelihood of informal
bankruptcy. Aborrower choosing informal bankruptcy forgoes bankruptcy’s automatic
stay, which means that he remains subject to all the collections efforts that creditors can
bring to bear. Thus, if creditors’ pressure is high due to common pool effects, the cost of
informal bankruptcy relative to formal bankruptcy will also be high, which will decrease
the probability of informal bankruptcy.
This paper provides a straightforward model of a creditor’s optimization problem in
the presence of one type of collection externality. It shows that shows that, under
reasonable assumptions and holding debt levels constant, increasing a borrower’s
number of loans will lead to increased collection, which will result in an increased
probability of formal bankruptcy and a decreased probability of informal bankruptcy.
The model’s conclusions are tested using a unique dataset containing accounts of
individual credit card users, and ?nds results consistent with the model’s predictions:
increasing a borrower’s number of lenders substantially increases the likelihood a
borrower will substitute formal bankruptcy for informal bankruptcy.
This result leads naturally to the following question: are common pool effects weaker
in states with laws that restrict collection? In particular, does increasing the number of
creditors cause less substitution of formal for informal bankruptcy in states with strict
garnishment and harassment laws, and what is the impact of homestead exemption law?
This paper ?nds that, while these laws do affect a borrower’s choice among repayment,
informal and formal bankruptcy, they do not change the impact of a borrower’s number
of loans[2]. So, for example, a borrower who lives in a state that allows creditors to
garnish a signi?cant percentage of his paycheck is less likely to choose informal
bankruptcy, but increasing this borrower’s number of loans does not decrease informal
bankruptcy more than it would if he lived in a state which forbids garnishment.
Section II reviews the law and economics literature related to creditor externalities
and discusses the relevant collection and bankruptcy laws. Section III presents the
theoretical model, and Section IV provides an empirical test. Section V concludes.
II. The theory of collection externalities
A. Related literature
Bankruptcy theory in the law and economics literature has focused primarily on
business bankruptcies, and the literature can be divided into two schools (Baird, 1998):
Creditors and
personal
bankruptcy
3
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the “proceduralist” and the “traditionalist.” Proceduralists contend that the goal of
bankruptcy law is to maximize ef?ciency by preserving the value of the ?rm, arguing
that collection leads to two sources of negative externality. Collectively, multiple
creditors may forgive less than the optimal proportion of the loan (the under-forgiveness
externality), and they may not provide borrowers with suf?cient time to repay
(the under-deferment externality), in comparison to a monopolist creditor. These effects
cause excessive diminution or devaluation of a ?rm’s assets and therefore lead to
economic inef?ciency ( Jackson, 1986).
Though the proceduralist argument is focused on business bankruptcies, as Hynes
(2004) points out, creditor externalities theoretically exist with individual collection as
well. Defaulting borrowers who are deprived of a vehicle or tools of trade may not be
able to work, and their marginal productivity may fall as garnishment decreases their
incentive to earn income. And though employers are banned from ?ring an employee
for a single garnishment order, these laws are dif?cult to enforce, and a borrower may
be ?red for more than one garnishment (Consumer Credit Protection Act, 1968).
A major critique of the proceduralists is that the common pool model is empirically
unsubstantiated. Kripke (1985) criticizes proceduralist papers as “neither premised on
nor tested by facts.” Block-Leib (1993) doubts that creditor externalities are widespread,
arguing that they only occur in very speci?c situations, not as a general rule in
insolvency. Warren (1993) argues that the proceduralists ignore the complexities of the
bankruptcy system and other empirical facts in order to imperfectly ?t their model to
real world evidence. In fact, her main contention is not with their conclusions, but rather
with their methodology. For example, Warren (1987) argues that in his proceduralist
works, Douglas Baird:
[. . .] may be right that unsecured creditors will push for any wild scheme that risks assets of
the estate in order to get some shot at repayment [. . .] But Baird uses his economic analysis to
limit the inquiry – and there is the ?nal rub.
In fact, only a few empirical studies have investigated collection externalities, and none
of these have established a link between these effects and personal bankruptcy[3].
B. Collection and bankruptcy law
The underlying objective of collection law is to reward creditors for diligence. From an
ef?ciency perspective, the argument is the following: if creditors bene?t from their own
collections efforts, collection will be more productive, and the cost of credit will be
minimized (Williams, 1998). An unsecured creditor is not required to consult with other
creditors before entering into a bilateral payment or preference agreement with a
defaulting borrower[4]. A second, less diligent creditor is prevented from placing a lien
on that property, even if his loan predates or exceeds the ?rst creditor’s loan. Thus,
creditors’ repayment depends not only on the borrower’s absolute willingness to repay,
but also on their efforts relative to other creditors.
One of the more effective methods of collection is through garnishment. Garnishment
laws regulate the percentage of a borrower’s intangible property, such as wages or bank
deposits, a creditor can collect directly from the borrower’s employer or bank. Federal
law shields 75 percent of a borrower’s weekly wages or 30 times the minimum wage,
whichever is greater. States can choose to protect a greater proportion of a borrower’s
wages; during the observation period of this paper’s data, 16 states had done so while
JFEP
6,1
4
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
six states had outlawed garnishment altogether[5]. A summary of state laws on
exemptions and garnishment can be found in Table I[6].
Finally, in addition to property seizure and garnishment, creditors can attempt to
persuade a defaulting borrower to repay through phone calls and letters, either on their
own or through a professional collections agent. Even without seizing property or
wages, a creditor may be able to exert considerable pressure on the delinquent
borrower. Collection tactics by third parties are limited by federal standards set in the
Fair Debt Collections Practices Act (FDCPA), though there is evidence that collections
agents often violate its provisions[7]. Some states set stricter standards or rules that
apply to lenders themselves, and some states go so far as to allow borrowers to sue
lenders who violate anti-harassment laws[8].
In contrast to laws governing creditor and borrower interactions outside of
bankruptcy, bankruptcy law explicitly forestalls competition among creditors through
the automatic stay, which disallows all collections efforts. If an unsecured creditor
obtained a lien before the bankruptcy ?ling, like any secured creditor, his right to that
collateral is upheld in the bankruptcy proceeding. After taxes, administrative costs and
secured creditors are paid, the value of remaining non-exempt assets is divided among
creditors of the same class on a pro rata basis.
State % Garn
a
Days lien
b
Home
c
State % Garn Days lien Home
AK FED Unlimited $54,000 MT FED Unlimited $40,000
AL FED Unlimited $5,000 NC 0% 0 $10,000
AR FED Unlimited Unlimited ND FED 180 $80,000
AZ FED Unlimited $100,000 NE 15% 90 $10,000
CA FED 90 $50,000 NH 0% . . . $30,000
CO FED 90 $30,000 NJ 10% Unlimited $15,000
CT FED Unlimited $75,000 NM FED Unlimited $30,000
DE 15% Unlimited $0 NV FED 120 $95,000
FL FED Unlimited Unlimited NY 10% Unlimited $10,000
GA FED 0 $5,000 OH FED 30 $5,000
HI 19% Unlimited $20,000 OK FED 180 Unlimited
IA FED Unlimited Unlimited OR FED 90 $25,000
ID FED 0 $50,000 PA 0% 0 $15,000
IL 15% 84 $7,500 RI FED Unlimited $15,000
IN FED Unlimited $7,500 SC 0% . . . $15,000
KS FED 0 Unlimited SD 20% 60 Unlimited
KY FED Unlimited $5,000 TN FED 180 $5,000
LA FED Unlimited $15,000 TX 0% . . . Unlimited
MA FED Unlimited $15,000 UT FED 120 $8,000
MD FED Unlimited $0 VA FED 90 $5,000
ME FED Unlimited $12,500 VT 0% . . . $30,000
MI FED 91 $15,000 WA FED 60 $30,000
MN FED 70 $200,000 WI 20% 90 $40,000
MO 10% 90 $8,000 WV 20% Unlimited $15,000
MS FED Unlimited $75,000 WY FED 90 $10,000
Notes:
a
“FED” indicates that garnishment up to the federal maximum of 25 percent of salary is
allowable;
b
days garn indicates the number of days a garnishment lien precludes collection by other
creditors;
c
if a state allows consumers to choose between the federal and state exemption, the table
entry is the higher of the federal or state exemption
Table I.
State homestead
exemptions and
garnishment laws
Creditors and
personal
bankruptcy
5
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Individuals canchoose to ?le under Chapter 7 or Chapter 13[9], andexemptionlaws affect
both. Borrowers ?ling for bankruptcy under Chapter 7 forfeit their non-exempt property
in return for protection of their future income from creditors’ claims. The amount of
property individuals are allowed to exempt is ostensibly determined under federal law,
but because they can opt out, states actually determine the exemptions borrowers are
allowed. The state homestead exemption can vary from almost nothing to virtually
unrestricted. Exemptions also impact the costs of a Chapter 13 bankruptcy, in which a
borrower submits a repayment plan in return for additional time to repay and a possible
partial discharge. Under Chapter 13, a borrower must repayat least as muchas she would
under Chapter 7, which naturally will vary according to Chapter 7 exemptions. Creditors
are also barred from attaching or seizing exempt property even outside of formal
bankruptcy, unless the creditor provided funding for that property’s purchase[10].
III. The economics of the common pool model
The central thesis of the common pool model is that multiple competing creditors
“over-collect” relative to a creditor who is a borrower’s unique lender, but it is not
obvious why a creditor would ever collect less forcefully than the law allows. Collection
is directed toward imposing a cost of non-payment on the borrower, and decreasing
that cost by limiting collection decreases the probability a borrower will repay. Though
creditors receive little in a formal bankruptcy, they should expect even less from
informal bankruptcy: what the borrower will pay the creditor in a bankruptcy should
set the upper bound of her willingness to pay outside of bankruptcy. The potential for a
negotiated solution will only exist if formal bankruptcy imposes additional costs.
In the following model, a creditor restrains collection to avoid “lost value,” analogous
to the economic concept of deadweight loss[11]. This model explores two potential
sources of lost value: ?rst, the borrower’s value of her assets is greater than the resale
value the creditor would recoup from seized property. Second, informal and formal
bankruptcy impose transactions costs both to the borrower and creditors. Note that
other potential channels for collection externalities exist; for example, garnishment itself
may result in diminished productivity and reduced ability to repay, a cost that is shared
among all creditors.
A. Single creditor model
The following two-period model of a negotiation begins with a borrower who has a
single creditor. In the ?rst period, the creditor announces the level of collection he will
pursue if the borrower chooses not to repay. In the second period, the borrower has three
options: formal bankruptcy, informal bankruptcy, or repayment. If he chooses formal
bankruptcy, the borrower loses his non-exempt assets and pays the transactions costs
such as ?ling or attorney fees. If he chooses informal bankruptcy, he is subject to the
collection the creditor proposed in period one. He does not submit his non-exempt assets
as in a formal bankruptcy, but there is a positive probability the creditor will ?nd and
seize his assets. If he chooses repayment, he repays the full value of the loan.
Borrowers have assets ( y) which a creditor could seize and sell, and the market
value is private information. Assume that y is uniformly distributed between [0,Y], and
this information is public knowledge. The borrower’s value of her assets is uy,
and because assets are subject to lost value in the resale market, u . 1. If the borrower
has non-exempt property and chooses formal bankruptcy, her cost is uy 2 E þ C
FB
,
JFEP
6,1
6
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
where C
FB
captures the borrower’s transactions costs of a formal bankruptcy, and E is
the relevant exemption level.
In informal bankruptcy, the borrower remains exposed to her creditor’s collection
actions. The unsecured creditor can attempt to secure her non-exempt assets, but
because the borrower is not required to itemize the property as she would in a formal
bankruptcy, assume that the creditor is only able to attach the property with probability
a, where a [ (0,1). The creditor also has the option of collecting intangible property
(wages, bene?ts, money in a bank account) through wage garnishment and other
collection actions. Let the intangible property the creditor claims be g. While the most
common form of collection is wage garnishment, where payments would accrue over
time as deductions from a borrower’s paycheck, for simplicity let g represent the total
amount collected. Thus, a borrower’s cost of informal bankruptcy is a(uy 2 E) þ g
þ C
IB
, where C
IB
captures the borrower’s other costs of informal bankruptcy, such as
disutility fromcollections calls. Note that C
FB
is likely to be signi?cantly larger than C
IB
:
the borrower choosing formal bankruptcy must bear administrative fees as well as
signi?cant stigma and future borrowing costs, which are largely avoided if the borrower
instead chooses informal bankruptcy[12].
The borrower therefore chooses formal bankruptcy when the value of her
non-exempt assets is suf?ciently low and the following holds:
y ,
Eð1 2aÞ þ g þ C
IB
2C
FB
uð1 2aÞ
Let L be the value of the loan a borrower must repay[13]. He chooses informal
bankruptcy when his cost of informal bankruptcy is less that the cost of formal
bankruptcy or repayment, which occurs when:
y [
Eð1 2aÞ þ g þ C
IB
2C
FB
uð1 2aÞ
;
L þaE 2g 2C
IB
ua
She chooses repayment when:
y .
L þaE 2g 2C
IB
ua
Note that if all of the borrower’s assets are exempt so that E . y, then his cost of formal
bankruptcy is simply C
FB
and his cost of informal bankruptcy is g þ C
IB
. The maximum
a creditor could collect from this borrower without triggering a formal bankruptcy is
C
FB
2 C
IB
. This model assumes that C
FB
2 C
IB
, g, i.e. that the creditor’s optimal g is
high enough to always induce formal bankruptcy for these borrowers[14].
When the borrower chooses to repay the loan, the creditor receives L. Assume that
property seizure is subject to lost value, so that the creditor’s value of the assets is only
the resale price. If the borrower has non-exempt assets, then the creditor’s bene?t of
formal bankruptcy is y 2 E 2 K
B,
and the creditor’s bene?t of informal bankruptcy is
a( y 2 E) þ g 2 K
IB,
where K
B
and K
IB
represent the transactions costs to the creditor
of participating in a formal or informal bankruptcy. Note that K
IB
is likely to be
substantially higher than K
B
. The creditor’s cost of informal bankruptcy includes all
administrative costs of collection, including the percentage paid to third-party
Creditors and
personal
bankruptcy
7
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
collectors or the fees and time spent ?ling garnishment requests or liens. A formal
bankruptcy is likely to impose very little if any administrative cost to the creditor[15].
The creditor’s maximization problem can be written as:
g
max
Z
ðEð12aÞþgþC
IB
2C
FB
Þ=uð12aÞ
E
½y 2E 2K
FB
?dy
þ
Z
ðLþaE2g2C
IB
Þ=ua
ðEð12aÞþgþC
IB
2C
FB
Þ=uð12aÞ
ðað y 2EÞ 2 K
IB
þ gÞdy
þ
Z
Y
ðLþaE2g2C
IB
Þ=ua
Ldy
Increasing g decreases the range of y over which the borrower chooses informal
bankruptcy. The creditor’s maximization problem requires him to balance his bene?t
of increasing g, the higher payoff from informal bankruptcy and increased likelihood of
repayment, against his cost of increasing g, the decreased likelihood a borrower will
choose informal bankruptcy and increased likelihood she will choose formal
bankruptcy.
Note that in this model, increasing E results in equal increases in the upper and
lower bounds of the range of y over which the borrower chooses informal bankruptcy.
Therefore, increasing the exemption level does not directly increase the likelihood of
informal bankruptcy. Increasing E does result in more borrowers choosing formal
bankruptcy and fewer borrowers choosing repayment.
The choice that maximizes the creditor’s expected payoff is given by the following:
g
*
¼
Lð1 2aÞð2u 21Þ þ ðu 21ÞðaC
FB
2 C
IB
Þ þuðK
IB
2aK
FB
Þ
½2u 21?
Increasing collection increases the bene?t of informal bankruptcy and increases the
probability of formal bankruptcy or loan repayment. And so the creditor optimizes by
increasing collection when:
.
the expected payoff from informal bankruptcy is low; or
.
the expected payoffs from other choices are high.
The assumption that the administrative costs of informal bankruptcy are low relative
to formal bankruptcy for a borrower and high for a creditor ensures that aC
FB
. C
IB,
K
IB
. aK
FB
and g . 0.
B. Multiple creditors
Now suppose the borrower has borrowed loan amount L from N creditors, so that each
creditor receives L/N if the borrower chooses to repay. Let G represent total collection,
the sum of all individual creditors’ collection. Assume that seized property is divided
on a pro rata basis among all creditors in a formal bankruptcy, and that an individual
creditor’s likelihood of seizing assets in informal bankruptcy is proportional to the
JFEP
6,1
8
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
number of creditors. However, collection bene?ts the collecting creditor alone. As a
result, some of the cost of collection, the increased probability of bankruptcy, is now
shared among all creditors.
Note that this model assumes that collection results in both positive and negative
externalities because one bene?t of collection, the increased likelihood of repayment,
is also shared. This assumption of strong complementarity between one creditor’s
collection and the repayment of other creditors’ loans is somewhat implausible, but it
will establish a lower bound on the impact of creditor externalities. An extension that
explores the results of relaxing this assumption is discussed in the next section.
Each creditor chooses collection level g to maximize the following:
g
max
Z
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
E
y 2E
N
2K
FB
dy
þ
Z
ðLþaE2G2C
IB
Þ=ua
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
a
N
ð y 2EÞ 2 K
IB
þ g
dy
þ
Z
Y
ðLþaE2G2C
IB
Þ=ua
L
N
dy
In the symmetric equilibrium, each creditor maximizes his expected payoff where:
g
**
¼
Lð1 2aÞ½uðN þ 1Þ 21? þ ðNu 21ÞðaC
FB
2 C
IB
Þ þ NuðK
IB
2aK
FB
Þ
N½ðN þ 1Þu 21?
Total collection is given by:
G ¼ Ng
**
¼
Lð1 2aÞ½uðN þ 1Þ 21? þ ðNu 21ÞðaC
FB
2 C
IB
Þ þ NuðK
IB
2aK
FB
Þ
½ðN þ 1Þu 21?
As in the single creditor model, the effect of exemptions on collection is neutral, even
though high exemptions reduce a creditor’s payoff from informal bankruptcy and
formal bankruptcy and therefore increase the relative bene?t of repayment. It might
seem that, in response to generous exemptions, the creditor should increase collection in
order to induce repayment. However, a creditor considering this tack would recognize
that high exemptions would mean that increasing collection would come with an
additional cost: a greater likelihood the borrower would choose formal bankruptcy in
response. The creditor’s costs and bene?ts of changing collection in response to higher
exemptions offset, and so higher exemptions do not provoke a change in the creditor’s
collection level.
The response of total collection to an increase in the number of creditors is given by
the following:
Creditors and
personal
bankruptcy
9
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
›G
›N
¼
u
2
ð aC
FB
2C
IB
Þ þuðu 21ÞðK
IB
2aK
FB
Þ
½ðN þ 1Þu 21?
2
The impact of increasing N on G is not a function of L, the borrower’s total debt. The
extension that follows in the next section demonstrates how, in the absence of the positive
externality, ›G/›N increases with L. Instead, in this model, the relative costs of informal
and formal bankruptcy determine how collection changes as N increases. For example,
if either K
IB
or C
FB
is high, each creditor increases collection accordingly, which results in
a lower probabilityof informal bankruptcyanda higher probabilityof formal bankruptcy.
The negative externality of collection, the creditors’ shared cost of the increased likelihood
that the borrower will enter formal bankruptcy, results in the cost of bankruptcy being
increasingly externalized as N increases. Creditors therefore under-respond to the cost,
which results in over-collection (›G/›N . 0). Recall that the transactions cost of formal
bankruptcy is likely to be relatively high for borrowers and low for creditors, and as long
as K
IB
. aK
FB
and aC
FB
. C
IB
holds, then increases in the number of creditors will lead
to increased total collection, the main result tested in this paper.
C. Multiple creditors, no complementarity in collection
In the previous section, one creditor’s collection was assumed to increase the likelihood
a borrower would pay off all loans. This introduced a positive externality in collection,
which the following extension will remove.
As in previous sections, a borrower’s cost of formal bankruptcy is uy 2 E þ C
FB
and her cost of informal bankruptcy is a(uy 2 E) þ G þ C
IB
, and therefore she chooses
formal bankruptcy whenever:
y ,
Eð1 2aÞ þ G þ C
IB
2C
FB
uð1 2aÞ
However, in this extension, the borrower decides whether to repay a creditor’s loan by
comparing the cost and bene?ts of repaying that particular loan. Assume again that
creditors are symmetric, so that the cost of repaying a particular loan is L/N. Assume
further that the borrower’s administrative cost and the probability of asset seizure in
informal bankruptcy resulting from a particular lender are proportionate to the size of
his loan. So a borrower’s cost of remaining informally bankrupt on a particular loan is
(a(uy 2 E) þ C
IB
/N) þ g, where g represents the collection level chosen by an
individual creditor. Note that in this model, as compared to the previous multiple
creditor model, one creditor’s collection does not affect the borrower’s likelihood of
repaying other creditors’ loans.
A borrower chooses informal bankruptcy when:
y [
Eð1 2aÞ þ G þ C
IB
2C
FB
uð1 2aÞ
;
L þaE 2Ng 2C
IB
ua
and chooses repayment when:
y .
L þaE 2Ng 2C
IB
ua
Assume the creditor’s administrative costs of informal and formal bankruptcy are
?xed costs. The creditor will choose g to maximize the following:
JFEP
6,1
10
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
g
max
Z
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
E
y 2E
N
2K
FB
f ð yÞdy
þ
Z
ðLþaE2Ng2C
IB
Þ=ua
ðEð12aÞþGþC
IB
2C
FB
Þ=uð12aÞ
a
N
ð y 2EÞ 2 K
IB
þ g
f ðyÞdy
þ
Z
Y
ðLþaE2Ng2C
IB
Þ=ua
L
N
f ð yÞdy
Total collection is given by:
Ng
*
¼
Lð1 2aÞð2u 21ÞN þ Eað1 2aÞðu 21ÞðN 21Þ þ C
IB
½a þ Nð1 2a 2uÞ?
þaC
FB
ðNu 21Þ 2NuaK
FB
þ NuK
IB
½Nð1 2aÞ þa?
½aðu 21Þð1 2NÞ þ Nð2u 21Þ?
The impact of increasing N is given by:
›Ng
**
›N
¼
ðu 21Þð2u 21Þð1 2aÞðL þEÞ þ ½u
2
a þ ð2u 21Þð1 2aÞ? aC
FB
2½au
2
?C
IB
þ½N
2
ð1 2aÞu þ ðu 21Þ{Nað1 2aÞ þa}
2
?uK
IB
þa
2
ð1 2uÞ uK
FB
½aðu 21Þð1 2NÞ þ Nð2u 21Þ?
2
As with the previous model, assuming that K
IB
. aK
FB
and aC
FB
. C
IB
is suf?cient
to ensure that the impact on total collection of increasing number of loans is
unambiguously positive. And note further that, unlike the previous model, higher
exemptions increase the impact of number of creditors on total collection. In this model,
a creditor’s own collection is more likely to induce repayment than in the previous model.
The creditor’s bene?t of increasing collection when exemptions are high is greater than
the potential cost, the increase in the probability a borrower will choose formal
bankruptcy in response. As a result, each creditor increases collection as exemptions
increase. And as the number of creditors increase and the cost of bankruptcy becomes
more externalized, higher exemptions lead to even higher total collection.
This means that the impact of higher exemptions on borrowers is twofold. First,
higher exemptions decrease the cost of formal bankruptcy and, to a lesser degree and
depending on the probability of property seizure, informal bankruptcy. Second, higher
exemptions also lead to increased collection, which increases the borrower’s cost of
informal bankruptcy. Both of these effects act to decrease the relative cost of formal
bankruptcy, and so higher exemptions should lead to a higher probability the borrower
chooses formal bankruptcy. However, the total impacts on the relative costs of informal
bankruptcy and repayment are ambiguous, and so increasing exemptions may increase
or decrease the probability a borrower chooses repayment or informal bankruptcy.
One ?nal note: underlying each of these models is a particular form of externality
problem, where lost value occurs due to property seizure and administrative costs. Each
model assumes that higher collection results in a higher relative bene?t of both formal
bankruptcyandrepayment, andtherefore increases incollectionincreases the probabilitya
Creditors and
personal
bankruptcy
11
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
borrower will choose either formal bankruptcy or repayment. However, as discussed
above, another potential negative externalitymayresult if collectiondecreases a borrower’s
ability to repay, such as a case where the borrower is ?red for multiple garnishments. As
number of lenders increase, this externality would, like the property seizure externality,
result inanincrease aborrower’s probabilityof choosingformal bankruptcyanda decrease
the probability of informal bankruptcy. However, unlike the externality that is the subject
of the model above, inthis case, as number of lenders increase, the borrower’s probabilityof
repayment could decrease. Which externality predominates will determine whether a
borrower’s number of loans will increase or decrease his likelihood of repayment.
D. Implications for garnishment and exemption laws
If empirical tests provide evidence of collection externalities, a second line of inquiry
concerns whether lenient garnishment, harassment, and exemption laws prevent their
effects. If borrower-friendly laws effectively set a cap on the level of collection creditors can
pursue throughgarnishment andharassment laws, it follows that the effects of externalities
would be weaker in states that restrict or disallow garnishment and other aggressive
creditor techniques. Inparticular, increasingaborrower’s number of loans shouldcause less
substitution away from informal bankruptcy in states with lenient garnishment and
anti-harassment laws. In the model without complementarities in collection, higher
exemptions are positively related to common pool effects, which would suggest greater
substitution away from informal bankruptcy in states with generous homestead
exemptions. We can test these predictions by including interactions between number of
loans and level of exemptions, percentage garnished, and anti-harassment statutes.
IV. Evidence of strategic over-collection
A. Data
This analysis uses a dataset comprising 52,721 pre-approved gold card users from
a large credit card issuer. The users were respondents to three solicitations of
observationally similar borrowers; the recipients were randomly assigned offers that
varied by length of introductory period and introductory interest rate. Borrowers who
responded to offers with shorter introductory periods and higher interest rates are more
likely to be credit constrained and to exhibit more risky borrowing behavior, and so the
speci?cations that follow include indicator variables for each mailing and controls for
length of the introductory period and interest rate (Table II).
The data includes variables fromeach borrower’s credit history, along with variables
describingeachborrower’s use of the goldcardfor at least 20 months. The borrower’s zip
code is recorded, allowing the addition of demographic information for their zip code and
state. Variables representing applicable state laws were also added: homestead
exemption represents the dollar value of the homestead exemption allowed in each state,
no garnishment is a dummy variable indicating whether the state prohibits garnishment
for ordinary loans, and anti-harassment statute indicates whether the state allows a
borrower the right to sue an abusive creditor[16].
The outcome measure speci?es borrowers’ choosing among three options:
repayment, formal bankruptcy, and informal bankruptcy. formal bankruptcy holds
if a borrower declared bankruptcy during the observation period. Borrowers were
classi?ed as choosing informal bankruptcy when they were charged off for long-term
delinquency (usually six months of non-payment) without a formal bankruptcy ?ling.
JFEP
6,1
12
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
All other borrowers were classi?ed under repayment. Some speci?cations dropped
non-delinquent borrowers, that is, borrowers who were never more than two months
late with a payment.
The major advantage of this dataset is that it includes very detailed, individual-level
observations of borrower behavior. Also, the dataset includes the lending terms of the
borrower’s contract, and so all of the speci?cations reported in this paper control for the
interest rate and length of the introductory period. One other unique aspect of the data is
its temporal range: solicitations were sent between late 1995 and early 1997, and so the
accounts provide information about borrowing during the period in the late 1990s when
personal bankruptcies were climbing most dramatically[17]. The data also predates the
passage of the Bankruptcy Abuse Prevention and Consumer Protection Act (BAPCPA)
of 2005 and the ?nancial crisis beginning in 2007, both of which caused short-run
distortions in bankruptcy rates and credit market behavior. The impact of changes that
have occurred after the observation period from this data is discussed in the conclusion.
One limitation of this dataset is that total balances and number of loans are recorded
at the time of solicitation, not at default. It is conceivable that a borrower could obtain
several loans or close loan accounts between solicitation and default (or the end of the
observation period). The usefulness of more up-to-date information is somewhat
limited, however, because the borrower has an incentive to increase borrowing in
anticipation of a default. For this reason, the speci?cations do not include information
describing the borrowers’ current balances on the account that is observed in this data.
If the correlation between loan balances and number of loans at solicitation and at
default is reasonably strong, the results will be consistent.
Repayment Informal bankruptcy Formal bankruptcy
Revolving loans 4.539 (3.472) 3.110 (2.481) 4.483 (3.002)
Installment loans 3.160 (3.315) 1.760 (2.671) 3.090 (3.247)
Number of mortgages 0.528 (0.745) 0.190 (0.451) 0.357 (0.568)
Revolving balances 3,512 (4,257) 3,429 (4,701) 6,237 (5,927)
Non-revolving balances 5,257 (11,859) 3,148 (7,355) 5,391 (8,713)
Mortgage balances 36,934 (64,519) 9,602 (29,097) 23,547 (43,459)
Years on ?le 1.178 (0.748) 0.788 (0.688) 1.030 (0.638)
Credit score 632.4 (86.12) 564.1 (92.25) 564.2 (97.17)
State unemployment rate 5.304 (1.045) 5.500 (0.910) 5.382 (1.042)
Transfer payments 11.68 (3.459) 11.48 (3.838) 11.19 (3.338)
Self-employed 10.52 (1.661) 10.53 (1.415) 10.71 (1.644)
Manufacturing 15.36 (4.606) 14.61 (4.668) 15.16 (4.637)
Homestead exemption 1.103 (1.461) 1.343 (1.644) 1.219 (1.528)
Anti-harassment statute 0.559 (0.496) 0.623 (0.485) 0.537 (0.499)
No garnishment 0.158 (0.365) 0.223 (0.417) 0.149 (0.356)
County median home value 131.8 (66.18) 123.3 (60.08) 122.2 (58.09)
Percent black 11.14 (12.05) 14.13 (13.56) 12.02 (13.56)
Percent hispanic 10.42 (12.63) 13.09 (15.22) 11.37 (13.78)
Percent divorced 9.729 (1.812) 9.867 (1.837) 10.04 (1.794)
Percent uninsured 13.27 (4.131) 14.55 (4.256) 14.04 (4.157)
High school 77.09 (7.848) 74.79 (8.092) 76.24 (8.285)
Income 46,959 (31,160) 36,243 (42,166) 39,588 (21,662)
Credit limit 7,168 (3,720) 4,615 (2,813) 5,569 (3,036)
Observations 51,069 747 905
Table II.
Selected variable means
standard deviations
by outcome
Creditors and
personal
bankruptcy
13
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
B. Empirical analysis
The borrower is assumed to choose among informal bankruptcy, formal bankruptcy
and repayment according to which yields higher utility. The stochastic component of
her utility function is assumed to follow an extreme-value distribution. This error
structure yields a multinomial logit, so that the probability of outcome k is given by:
P
k
¼
e
b
0
k
X
e
b
0
R
X
þ e
b
0
I
X
þ e
b
0
F
X
:
where k ¼ repayment, informal bankruptcy or formal bankruptcy. The coef?cient
estimates are not readily interpretable, and therefore the average marginal effects are
presented and discussed.
The borrower’s choice among her three options is modeled as a function of the
explanatory variables itemized in Table III. These variables include several measures,
including loan balances, number of loans, and types of loans, recorded in a borrower’s
credit report at the time of solicitation, and the lender recorded additional variables. In
addition to this list of individual-level characteristics, several variables at the zip code
and state level were included, such as the borrower’s state and county unemployment
rate and median home value. Also included were variables describing the bankruptcy
and collection laws in each borrower’s state.
C. Results
C1. Credit variables. The ?rst three columns of Table IV report the average marginal
effects of the explanatory variables on the probability a borrower chooses repayment,
informal or formal bankrutpcy; the ?nal three columns report the results of speci?cations
that exclude credit limit and credit score. As predicted above, many of the variables from
the borrower’s credit report affect informal and formal bankruptcy in the same direction.
For example, years on ?le, credit limit and credit score all decrease the probability a
borrower will choose either formal or informal bankruptcy and increase the probability
of repayment. Counter-intuitively, an increase in a borrower’s income actually increases
the likelihooda borrower chooses either informal or formal bankruptcy, andsigni?cantly
decreases the likelihood she chooses repayment. As the ?nal three columns of Table IV
show, income has the expected negative effect on repayment and positive effects on
informal and formal bankruptcy when credit score and credit limit are dropped, while the
impacts of other variables are largely unaffected. Credit score and limit, while providing
a useful control for the creditor’s assessment of credit-worthiness, are constructed from
credit history variables and can bias these variables’ coef?cient estimates.
Unlike the other credit history variables, the borrower’s number of loans signi?cantly
increases the probability she will choose formal bankruptcy and decreases the
probability she will choose informal bankruptcy. For example, holding all else equal,
increasing a borrower’s number of loans by 1 will increase her probability of formal
bankruptcy by 0.052 percentage points, while decreasing her probability of informal
bankruptcy by 0.042 percentage points. Given that the unconditional probability of
formal bankruptcy is 1.7 percent, the 0.052 percentage point increase indicates a
3.1 percent increase in the probability of formal bankruptcy. The probability of informal
bankruptcy is 1.4 percent in this dataset, which means that a 0.042 percentage point
decrease implies a 3 percent decrease in the probability of informal bankruptcy.
JFEP
6,1
14
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
The effect of number of loans on repayment is small, negative and insigni?cant when
credit score and credit limit are included. When credit score and credit limit are not
included, the coef?cient on repayment is positive and statistically signi?cant,
indicating that common pool effects may be diminishing a borrower’s ability to repay.
An increase of one loan will increase the borrower’s probability of repayment by about
Credit record variables: data compiled immediately before solicitation
Revolving loans The borrower’s number of revolving loans open at the time of the credit report
Installment loans The borrower’s number of installment loans
Number of mortgages The borrower’s number of mortgages
Loan balances The total non-mortgage balances the borrower has on all loans
Revolving balances The total balances on revolving loans
Non-revolving
balances
The balances on non-mortgage, non-revolving loans
Mortgage balances Total mortgage balances at the time of the credit report
Years on ?le The number of years on ?le with the credit reporting agency
Credit score A measure, compiled by the credit reporting agency, meant to capture the
borrower’s probability of default
State level variables
a
Unemployment rate
b
The unemployment rate of the borrower’s state of residence lagged by one year
Median home value The median home value in the borrower’s state
Transfer payments
b
The value of the average transfer (food stamps, housing subsidies, disability
and unemployment payments, Medicaid and TANF) to low income families
Self-employed
b
Percentage of workers who are self-employed
Manufacturing
b
Percentage of workers who work in manufacturing
Homestead
exemption
State homestead exemption
Anti-harassment
statute
A dummy that equals 1 if the state allows a private right of action against
harassing creditors
No garnishment A dummy that equals 1 if the state does not allow garnishment for ordinary
loans
County level variables
a
Unemployment rate The county unemployment rate, lagged by one year
Median home value The median home value in the borrower’s county
Percent black Percentage of residents in a borrower’s county that are black
Percent hispanic Percentage of residents in a borrower’s county that are hispanic
Percent divorced Percentage of adults in a borrower’s county that are divorced
Percent uninsured Percentage of residents that do not have insurance coverage
High school Percentage of adults who graduated from high school
Issuer variables: recorded by the credit card issuer at the beginning of the experiment period
Mailing Indicator variables for each round of mailing
Introductory rate Teaser interest rate offered in solicitation
Intro period duration Length in months of introductory period
Log income The log of the borrower’s self-reported income
Credit limit The borrower’s limit on the card
Notes:
a
If a borrower moved during the experiment period and defaulted, the borrower’s residence
was de?ned as the state and zip code where the default occurred; if the borrower moved and did not
default, the borrower’s residence was de?ned as the state where she lived longest during the
experiment period; aside from the state laws and unless noted otherwise, the data was merged from the
2000 Census based on the account’s zip code;
b
constructed from the 1996-2000 March supplement to
the Current Population Survey
Table III.
De?nitions
Creditors and
personal
bankruptcy
15
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
V
a
r
i
a
b
l
e
s
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
M
e
d
i
a
n
h
o
m
e
v
a
l
u
e
4
.
8
5
£
1
0
2
0
5
*
*
*
2
1
.
4
5
£
1
0
2
0
5
*
2
3
.
4
0
£
1
0
2
0
5
*
*
*
6
.
8
9
£
1
0
2
0
5
*
*
*
2
2
.
1
7
£
1
0
2
0
5
*
*
2
4
.
7
2
£
1
0
2
0
5
*
*
*
(
1
.
0
9
£
1
0
2
0
5
)
(
7
.
5
7
£
1
0
2
0
6
)
(
1
.
0
2
£
1
0
2
0
5
)
(
1
.
3
4
£
1
0
2
0
5
)
(
9
.
6
9
£
1
0
2
0
6
)
(
1
.
3
1
£
1
0
2
0
5
)
P
e
r
c
e
n
t
d
i
v
o
r
c
e
d
2
0
.
0
0
0
9
1
1
*
*
*
0
.
0
0
0
5
2
5
*
*
*
0
.
0
0
0
3
8
5
2
0
.
0
0
1
3
4
*
*
*
0
.
0
0
0
7
2
6
*
*
*
0
.
0
0
0
6
1
6
(
0
.
0
0
0
3
2
2
)
(
0
.
0
0
0
1
6
5
)
(
0
.
0
0
0
3
0
6
)
(
0
.
0
0
0
3
9
9
)
(
0
.
0
0
0
2
0
4
)
(
0
.
0
0
0
3
8
3
)
P
e
r
c
e
n
t
u
n
i
n
s
u
r
e
d
2
5
.
6
9
£
1
0
2
0
5
2
7
.
3
8
£
1
0
2
0
6
6
.
4
3
£
1
0
2
0
5
2
0
.
0
0
0
2
6
5
5
.
0
2
£
1
0
2
0
5
0
.
0
0
0
2
1
5
(
0
.
0
0
0
2
6
3
)
(
0
.
0
0
0
1
5
3
)
(
0
.
0
0
0
2
5
4
)
(
0
.
0
0
0
3
3
1
)
(
0
.
0
0
0
1
9
5
)
(
0
.
0
0
0
3
1
5
)
L
o
g
i
n
c
o
m
e
2
0
.
0
0
4
2
1
*
*
*
0
.
0
0
0
9
6
9
*
*
*
0
.
0
0
3
2
4
*
*
*
0
.
0
0
1
3
6
*
*
*
2
0
.
0
0
0
6
9
0
*
*
*
2
0
.
0
0
0
6
6
7
*
*
(
0
.
0
0
0
9
6
4
)
(
0
.
0
0
0
3
7
0
)
(
0
.
0
0
0
7
9
6
)
(
0
.
0
0
0
4
0
6
)
(
0
.
0
0
0
2
1
4
)
(
0
.
0
0
0
2
7
4
)
Y
e
a
r
s
o
n
?
l
e
0
.
0
0
2
0
5
*
*
2
0
.
0
0
1
8
1
*
*
2
0
.
0
0
0
2
4
1
0
.
0
0
8
3
0
*
*
*
2
0
.
0
0
4
5
0
*
*
*
2
0
.
0
0
3
8
0
*
*
*
(
0
.
0
0
0
9
0
5
)
(
0
.
0
0
0
7
2
7
)
(
0
.
0
0
0
6
3
7
)
(
0
.
0
0
1
2
4
)
(
0
.
0
0
0
9
1
8
)
(
0
.
0
0
0
9
4
5
)
C
r
e
d
i
t
s
c
o
r
e
0
.
0
0
0
1
1
1
*
*
*
2
3
.
4
0
£
1
0
2
0
5
*
*
*
2
7
.
7
2
£
1
0
2
0
5
*
*
*
(
5
.
2
0
£
1
0
2
0
6
)
(
3
.
4
0
£
1
0
2
0
6
)
(
3
.
8
2
£
1
0
2
0
6
)
C
r
e
d
i
t
l
i
m
i
t
3
.
6
5
£
1
0
2
0
6
*
*
*
2
1
.
4
9
£
1
0
2
0
6
*
*
*
2
2
.
1
6
£
1
0
2
0
6
*
*
*
(
2
.
2
0
£
1
0
2
0
7
)
(
1
.
3
9
£
1
0
2
0
7
)
(
1
.
7
6
£
1
0
2
0
7
)
I
n
t
r
o
d
u
c
t
o
r
y
p
e
r
i
o
d
d
u
r
a
t
i
o
n
0
.
0
0
0
4
3
5
*
*
*
2
0
.
0
0
0
2
7
5
*
*
*
2
0
.
0
0
0
1
6
0
0
.
0
0
0
7
9
8
*
*
*
2
0
.
0
0
0
4
3
6
*
*
*
2
0
.
0
0
0
3
6
2
*
*
(
0
.
0
0
0
1
6
6
)
(
6
.
6
5
£
1
0
2
0
5
)
(
0
.
0
0
0
1
3
2
)
(
0
.
0
0
0
2
1
1
)
(
7
.
9
9
£
1
0
2
0
5
)
(
0
.
0
0
0
1
7
8
)
I
n
t
r
o
d
u
c
t
o
r
y
r
a
t
e
2
0
.
0
0
2
5
1
*
*
0
.
0
0
1
8
3
*
*
*
0
.
0
0
0
6
7
2
2
0
.
0
0
5
6
6
*
*
*
0
.
0
0
3
1
5
*
*
*
0
.
0
0
2
5
1
*
(
0
.
0
0
1
1
5
)
(
0
.
0
0
0
4
1
2
)
(
0
.
0
0
1
0
4
)
(
0
.
0
0
1
4
6
)
(
0
.
0
0
0
5
1
9
)
(
0
.
0
0
1
3
8
)
H
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
(
$
1
0
,
0
0
0
s
)
2
0
.
0
0
0
5
6
4
2
2
.
4
1
£
1
0
2
0
5
0
.
0
0
0
5
8
9
*
2
0
.
0
0
0
4
6
4
2
0
.
0
0
0
1
2
3
0
.
0
0
0
5
8
7
(
0
.
0
0
0
4
1
6
)
(
0
.
0
0
0
2
7
2
)
(
0
.
0
0
0
3
2
4
)
(
0
.
0
0
0
5
1
4
)
(
0
.
0
0
0
3
5
7
)
(
0
.
0
0
0
3
8
0
)
A
n
t
i
-
h
a
r
a
s
s
m
e
n
t
s
t
a
t
u
t
e
0
.
0
0
1
7
5
0
.
0
0
1
0
6
2
0
.
0
0
2
8
1
*
*
0
.
0
0
2
5
7
0
.
0
0
1
2
9
2
0
.
0
0
3
8
7
*
*
*
(
0
.
0
0
1
3
6
)
(
0
.
0
0
0
6
9
6
)
(
0
.
0
0
1
2
5
)
(
0
.
0
0
1
6
0
)
(
0
.
0
0
0
8
8
0
)
(
0
.
0
0
1
4
8
)
N
o
g
a
r
n
i
s
h
m
e
n
t
2
0
.
0
0
3
7
4
*
*
0
.
0
0
3
9
1
*
*
*
2
0
.
0
0
0
1
7
3
2
0
.
0
0
5
0
9
*
*
0
.
0
0
4
9
5
*
*
*
0
.
0
0
0
1
3
7
(
0
.
0
0
1
5
3
)
(
0
.
0
0
1
0
7
)
(
0
.
0
0
1
3
4
)
(
0
.
0
0
1
9
8
)
(
0
.
0
0
1
4
6
)
(
0
.
0
0
1
5
5
)
L
o
a
n
b
a
l
a
n
c
e
s
2
2
.
9
1
£
1
0
2
0
8
2
2
.
3
7
£
1
0
2
0
8
5
.
2
9
£
1
0
2
0
8
*
*
*
2
1
.
4
1
£
1
0
2
0
7
*
*
3
.
5
1
£
1
0
2
0
8
1
.
0
6
£
1
0
2
0
7
*
*
*
(
4
.
8
9
£
1
0
2
0
8
)
(
4
.
3
5
£
1
0
2
0
8
)
(
1
.
6
5
£
1
0
2
0
8
)
(
5
.
5
5
£
1
0
2
0
8
)
(
2
.
5
8
£
1
0
2
0
8
)
(
3
.
3
1
£
1
0
2
0
8
)
N
u
m
b
e
r
o
f
l
o
a
n
s
2
9
.
9
4
£
1
0
2
0
5
2
0
.
0
0
0
4
2
0
*
*
*
0
.
0
0
0
5
1
9
*
*
*
0
.
0
0
0
4
3
3
*
*
*
2
0
.
0
0
0
8
0
3
*
*
*
0
.
0
0
0
3
7
0
*
*
*
(
9
.
4
5
£
1
0
2
0
5
)
(
9
.
4
5
£
1
0
2
0
5
)
(
8
.
5
5
£
1
0
2
0
5
)
(
0
.
0
0
0
1
3
4
)
(
0
.
0
0
0
1
0
6
)
(
0
.
0
0
0
1
1
3
)
O
b
s
e
r
v
a
t
i
o
n
s
5
2
,
7
2
1
5
2
,
7
2
1
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
p
,
0
.
1
,
*
*
p
,
0
.
0
5
a
n
d
*
*
*
p
,
0
.
0
1
;
s
t
a
n
d
a
r
d
e
r
r
o
r
s
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
a
n
d
a
r
e
c
l
u
s
t
e
r
e
d
b
y
s
t
a
t
e
;
e
a
c
h
s
p
e
c
i
?
c
a
t
i
o
n
i
n
T
a
b
l
e
s
I
V
-
V
I
a
l
s
o
c
o
n
t
r
o
l
l
e
d
f
o
r
t
h
e
s
t
a
t
e
a
n
d
c
o
u
n
t
y
d
e
m
o
g
r
a
p
h
i
c
v
a
r
i
a
b
l
e
s
l
i
s
t
e
d
i
n
T
a
b
l
e
I
I
I
;
s
p
e
c
i
?
c
a
t
i
o
n
s
w
i
t
h
o
u
t
t
h
e
s
e
c
o
n
t
r
o
l
s
y
i
e
l
d
e
d
s
i
m
i
l
a
r
r
e
s
u
l
t
s
;
f
u
l
l
r
e
s
u
l
t
s
a
r
e
a
v
a
i
l
a
b
l
e
u
p
o
n
r
e
q
u
e
s
t
Table IV.
Multinomial logit
results – marginal effects
JFEP
6,1
16
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
0.043 percentage points. Given that the average charged-off balance is $4,604, the mean
expected cost of each extra loan is about $2 per borrower in this sample. As we will see
below, the creditor’s expected cost is signi?cantly higher for borrowers who have
missed more than two consecutive payments.
C2. State laws. Most of the coef?cients on homestead exemptions in Tables IV-VI
fail the test of statistical signi?cance. The signs of the coef?cients are consistent with
the model, however: increasing the homestead exemption allowance decreases the
likelihood of repayment, increases the likelihood of formal bankruptcy, and has a small
negative effect on informal bankruptcy. The coef?cients on garnishment laws are also
consistent with the model, and have greater precision: borrowers living in states
without garnishment are signi?cantly less likely to repay, more likely to remain in
informal bankruptcy, and less likely to ?le a formal bankruptcy (though the coef?cient
on formal bankruptcy is not statistically signi?cant). Anti-harassment laws also
signi?cantly decrease the likelihood of formal bankruptcy.
The impact of harassment and garnishment laws seems to be substantial. For example,
the coef?cient in the second column of Table IV implies that moving a borrower from a
state without garnishment to a state where garnishment is permitted will decrease the
probability of informal bankruptcy by 0.39 percentage points, which is roughly equal to
the impact of increasing the borrower’s number of loans by 9. The same move to a state
that allows garnishment will increase the probability of repayment by 0.37 percentage
points, which is similar to the effect of increasing the borrower’s time on ?le by 1.8 years.
The ?rst three columns of Table V report the results of an estimation that included
interactions between state laws and the borrower’s number of loans. If state laws
effectively limit over-collection, then the borrower’s number of loans should have less
impact in states that restrict collection. So, for example, the garnishment and
anti-harassment interaction coef?cients would be positive for informal bankruptcy and
negative for formal bankruptcy if borrowers living in states that restrict collection are
less likely to substitute formal bankruptcy for informal bankruptcy as number of loans
increase. The signs of these coef?cients do not follow this pattern and are statistically
insigni?cant. These results do not support the hypothesis that garnishment and
anti-harassment laws limit the impact of common pool effects.
In the model presented in IIIB where creditor’s collection increased repayment on all
loans, exemptions had no effect on each creditor’s optimal collection level. However,
in the extension presented in Section III(C), increasing the exemption level not only
increased total collection, the impact of number of loans on total collection also increased
with the exemption level. This would imply that there should be greater substitution
towards repayment and formal bankruptcy and away from informal bankruptcy in
states with higher homestead exemptions. Instead, the coef?cients on the homestead
interaction terms for formal and informal bankruptcy are statistically insigni?cant,
and the coef?cient for repayment is negative rather than positive. Again, the homestead
exemption laws do not appear to counteract the impact of common pool effects.
C3. Delinquent borrowers. The credit negotiation model focuses on distressed
borrowers. For that reason, the last three columns of Table V report the results of a
speci?cation that only includes borrowers who were at least two months delinquent;
borrowers are designated as having chosen “repayment” if they return to good
standing during the observation period. In this speci?cation, increasing a borrower’s
number of loans again signi?cantly increases the probability of formal and
Creditors and
personal
bankruptcy
17
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
A
l
l
b
o
r
r
o
w
e
r
s
D
e
l
i
n
q
u
e
n
t
b
o
r
r
o
w
e
r
s
o
n
l
y
V
a
r
i
a
b
l
e
s
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
L
o
g
i
n
c
o
m
e
2
0
.
0
0
4
2
0
*
*
*
0
.
0
0
0
9
6
0
*
*
*
0
.
0
0
3
2
4
*
*
*
2
0
.
0
3
6
4
*
*
*
0
.
0
0
6
9
1
0
.
0
2
9
5
*
*
*
(
0
.
0
0
0
9
6
5
)
(
0
.
0
0
0
3
6
4
)
(
0
.
0
0
0
7
9
9
)
(
0
.
0
0
9
6
0
)
(
0
.
0
0
5
2
1
)
(
0
.
0
0
9
0
6
)
Y
e
a
r
s
o
n
?
l
e
0
.
0
0
2
0
1
*
*
2
0
.
0
0
1
7
9
*
*
2
0
.
0
0
0
2
2
7
2
0
.
0
1
1
5
2
0
.
0
1
4
2
0
.
0
2
5
6
*
*
(
0
.
0
0
0
9
0
0
)
(
0
.
0
0
0
7
1
6
)
(
0
.
0
0
0
6
3
5
)
(
0
.
0
1
6
5
)
(
0
.
0
1
2
8
)
(
0
.
0
1
1
6
)
C
r
e
d
i
t
s
c
o
r
e
0
.
0
0
0
1
1
1
*
*
*
2
3
.
3
9
£
1
0
2
0
5
*
*
*
2
7
.
7
1
£
1
0
2
0
5
*
*
*
0
.
0
0
0
4
3
9
*
*
*
2
7
.
6
5
£
1
0
2
0
5
2
0
.
0
0
0
3
6
2
*
*
*
(
5
.
2
1
£
1
0
2
0
6
)
(
3
.
3
6
£
1
0
2
0
6
)
(
3
.
8
2
£
1
0
2
0
6
)
(
9
.
8
7
£
1
0
2
0
5
)
(
5
.
4
5
£
1
0
2
0
5
)
(
8
.
2
1
£
1
0
2
0
5
)
C
r
e
d
i
t
l
i
m
i
t
3
.
6
3
£
1
0
2
0
6
*
*
*
2
1
.
4
8
£
1
0
2
0
6
*
*
*
2
2
.
1
5
£
1
0
2
0
6
*
*
*
2
.
4
3
£
1
0
2
0
5
*
*
*
2
1
.
2
4
£
1
0
2
0
5
*
*
*
2
1
.
1
9
£
1
0
2
0
5
*
*
*
(
2
.
2
1
£
1
0
2
0
7
)
(
1
.
4
3
£
1
0
2
0
7
)
(
1
.
7
4
£
1
0
2
0
7
)
(
3
.
0
6
£
1
0
2
0
6
)
(
2
.
7
2
£
1
0
2
0
6
)
(
2
.
1
3
£
1
0
2
0
6
)
H
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
(
$
1
0
,
0
0
0
s
)
2
0
.
0
0
0
1
7
9
2
0
.
0
0
0
2
6
4
0
.
0
0
0
4
4
4
2
0
.
0
0
4
4
6
2
0
.
0
0
4
2
1
0
.
0
0
8
6
7
*
(
0
.
0
0
0
4
6
4
)
(
0
.
0
0
0
3
2
3
)
(
0
.
0
0
0
3
6
1
)
(
0
.
0
0
6
2
9
)
(
0
.
0
0
5
4
8
)
(
0
.
0
0
4
6
5
)
A
n
t
i
-
h
a
r
a
s
s
m
e
n
t
s
t
a
t
u
t
e
0
.
0
0
3
8
4
0
.
0
0
0
1
4
8
2
0
.
0
0
3
9
9
*
0
.
0
2
3
1
0
.
0
2
2
5
*
2
0
.
0
4
5
7
*
*
*
(
0
.
0
0
2
3
4
)
(
0
.
0
0
1
2
1
)
(
0
.
0
0
2
1
1
)
(
0
.
0
1
7
1
)
(
0
.
0
1
3
5
)
(
0
.
0
1
6
3
)
N
o
g
a
r
n
i
s
h
m
e
n
t
2
0
.
0
0
3
6
9
*
*
0
.
0
0
5
0
0
*
*
*
2
0
.
0
0
1
3
1
2
0
.
0
5
9
2
*
*
*
0
.
0
7
5
8
*
*
*
2
0
.
0
1
6
6
(
0
.
0
0
1
8
3
)
(
0
.
0
0
1
3
6
)
(
0
.
0
0
1
4
3
)
(
0
.
0
2
0
7
)
(
0
.
0
2
0
2
)
(
0
.
0
1
9
7
)
L
o
a
n
b
a
l
a
n
c
e
s
2
2
.
8
6
£
1
0
2
0
8
2
2
.
4
0
£
1
0
2
0
8
5
.
2
6
£
1
0
2
0
8
*
*
*
2
5
.
8
6
£
1
0
2
0
7
2
1
.
3
1
£
1
0
2
0
6
1
.
9
0
£
1
0
2
0
6
*
*
*
(
4
.
8
5
£
1
0
2
0
8
)
(
4
.
3
2
£
1
0
2
0
8
)
(
1
.
6
5
£
1
0
2
0
8
)
(
1
.
3
2
£
1
0
2
0
6
)
(
9
.
7
1
£
1
0
2
0
7
)
(
7
.
1
6
£
1
0
2
0
7
)
N
u
m
b
e
r
o
f
l
o
a
n
s
0
.
0
0
0
1
7
0
2
0
.
0
0
0
5
5
9
*
*
*
0
.
0
0
0
3
8
9
*
*
2
0
.
0
0
5
1
3
*
*
*
2
0
.
0
0
5
2
5
*
*
*
0
.
0
1
0
4
*
*
*
(
0
.
0
0
0
1
9
2
)
(
0
.
0
0
0
1
7
8
)
(
0
.
0
0
0
1
8
3
)
(
0
.
0
0
1
5
3
)
(
0
.
0
0
1
6
6
)
(
0
.
0
0
1
5
6
)
L
o
a
n
s
*
h
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
2
6
.
9
4
£
1
0
2
0
5
*
4
.
8
5
£
1
0
2
0
5
2
.
0
9
£
1
0
2
0
5
(
3
.
9
8
£
1
0
2
0
5
)
(
3
.
2
8
£
1
0
2
0
5
)
(
2
.
8
9
£
1
0
2
0
5
)
L
o
a
n
s
*
a
n
t
i
-
h
a
r
a
s
s
m
e
n
t
2
0
.
0
0
0
3
3
4
0
.
0
0
0
1
8
9
0
.
0
0
0
1
4
5
(
0
.
0
0
0
2
2
9
)
(
0
.
0
0
0
1
8
8
)
(
0
.
0
0
0
1
7
6
)
L
o
a
n
s
*
g
a
r
n
i
s
h
m
e
n
t
1
.
4
5
£
1
0
2
0
5
2
0
.
0
0
0
1
5
8
0
.
0
0
0
1
4
4
(
0
.
0
0
0
1
7
1
)
(
0
.
0
0
0
1
2
4
)
(
0
.
0
0
0
1
0
7
)
O
b
s
e
r
v
a
t
i
o
n
s
5
2
,
7
2
1
4
,
6
6
9
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
p
,
0
.
1
,
*
*
p
,
0
.
0
5
a
n
d
*
*
*
p
,
0
.
0
1
;
s
t
a
n
d
a
r
d
e
r
r
o
r
s
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
a
n
d
a
r
e
c
l
u
s
t
e
r
e
d
b
y
s
t
a
t
e
;
e
a
c
h
s
p
e
c
i
?
c
a
t
i
o
n
a
l
s
o
i
n
c
l
u
d
e
d
a
l
l
o
f
t
h
e
c
o
n
t
r
o
l
v
a
r
i
a
b
l
e
s
l
i
s
t
e
d
i
n
T
a
b
l
e
I
V
Table V.
Multinomial logit
results – marginal effects
(standard errors
in parentheses)
JFEP
6,1
18
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
informal bankruptcy. The repayment coef?cient is negative and statistically
signi?cant, and it indicates that an additional loan decreases the probability a
delinquent borrower will repay by 0.513 percentage points. Thus, the addition of one
more loan to the record of a delinquent borrower increases the creditor’s expected
charged-off balance by $23.62.
C4. Revolving and non-revolving loans. The ?rst three columns of Table VI report
the results of a speci?cation that separates revolving and non-revolving loan balances
and number of loans. These results show that revolving balances have a signi?cantly
positive effect on formal and informal bankruptcy, while non-revolving balances have
negative effects on both informal and formal bankruptcy. These results suggest that
high non-revolving balances might indicate a less risky borrower and revolving loans
may imply the opposite. Note, however, that the number of loans coef?cients do not
follow this pattern, and instead correspond to our common pool model: higher numbers
of revolving and non-revolving loans both signi?cantly increase the probability of
formal bankruptcy and decrease the probability of informal bankruptcy.
The common pool model addresses the strategic problem faced by lenders of
unsecured loans. Note that, while revolving loans are almost always unsecured, and
installment loans may or may not be secured, mortgages are secured by a valuable
asset. Thus, mortgage lenders, because they have the option to foreclose, have less
incentive to react to the possibility of losing attachable assets to other lenders. Though
the ?nal three columns of Table VI show that increasing a borrower’s number of
mortgages results in similar coef?cient estimates as increasing other types of loans,
these coef?cient estimates are not statistically signi?cant.
V Conclusion
Correcting for strategic over-collection has been cited as a central tenant of the US
business bankruptcy system. This paper provides the ?rst systematic examination of
collection externalities in personal bankruptcy, and the results suggest that the common
pool effects are real phenomena. The credit card data utilized in this paper allowed an
evaluation of the distinct impacts of number of loans oninformal andformal bankruptcy.
If number of loans were simply an indicator of credit-worthiness, increasing number of
loans should either decrease or increase both informal and formal bankruptcy. However,
increasing a borrower’s number of loans consistently increases her probability of
choosing formal bankruptcy while decreasing the likelihood she chooses informal
bankruptcy, a result that is in accordance with the common pool model of collection. The
effects are not only statistically signi?cant, but also economically signi?cant, potentially
contributing dramatically to the outcome a distressed borrower experiences.
This paper ?nds strong evidence that anti-harassment, garnishment, and exemption
laws have an impact on a borrower’s choice among informal bankruptcy, formal
bankruptcy, and repayment. However, there is little evidence that these laws mitigate
the effects of creditor externalities. Unfortunately, these laws change very rarely, so this
paper, like most investigations into these laws’ effects, relies on cross-state variation,
and the small sample size may prevent a complete accounting of the laws’ impacts.
Personal bankruptcy law faced a major overhaul in via the enactment of BAPCPA in
2005, which, along with the Dodd-Frank Wall Street Reform and Consumer Protection
Act of 2010, signi?cantly changed the consumer borrowing landscape in ways that
could alleviate or exacerbate collection externalities. For example, Dodd-Frank created
Creditors and
personal
bankruptcy
19
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
V
a
r
i
a
b
l
e
s
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
R
e
p
a
y
I
n
f
o
r
m
a
l
b
a
n
k
r
u
p
t
c
y
B
a
n
k
r
u
p
t
c
y
H
o
m
e
s
t
e
a
d
e
x
e
m
p
t
i
o
n
(
$
1
0
,
0
0
0
s
)
2
0
.
0
0
0
3
8
8
2
4
.
8
3
£
1
0
2
0
5
0
.
0
0
0
4
3
6
2
0
.
0
0
0
4
0
2
2
4
.
4
6
£
1
0
2
0
5
0
.
0
0
0
4
4
7
(
0
.
0
0
0
3
7
9
)
(
0
.
0
0
0
2
5
3
)
(
0
.
0
0
0
2
9
8
)
(
0
.
0
0
0
3
8
3
)
(
0
.
0
0
0
2
6
9
)
(
0
.
0
0
0
2
9
8
)
A
n
t
i
-
h
a
r
a
s
s
m
e
n
t
s
t
a
t
u
t
e
0
.
0
0
1
5
2
0
.
0
0
0
9
8
5
2
0
.
0
0
2
5
1
*
*
0
.
0
0
1
5
0
0
.
0
0
1
0
5
2
0
.
0
0
2
5
6
*
*
(
0
.
0
0
1
2
7
)
(
0
.
0
0
0
6
4
7
)
(
0
.
0
0
1
1
5
)
(
0
.
0
0
1
2
6
)
(
0
.
0
0
0
6
8
9
)
(
0
.
0
0
1
1
5
)
N
o
g
a
r
n
i
s
h
m
e
n
t
2
0
.
0
0
3
1
4
*
*
0
.
0
0
3
5
2
*
*
*
2
0
.
0
0
0
3
8
0
2
0
.
0
0
3
4
5
*
*
0
.
0
0
3
8
3
*
*
*
2
0
.
0
0
0
3
8
2
(
0
.
0
0
1
3
9
)
(
0
.
0
0
1
0
1
)
(
0
.
0
0
1
1
6
)
(
0
.
0
0
1
4
1
)
(
0
.
0
0
1
0
6
)
(
0
.
0
0
1
1
6
)
R
e
v
o
l
v
i
n
g
b
a
l
a
n
c
e
s
2
9
.
6
6
£
1
0
2
0
7
*
*
*
1
.
9
1
£
1
0
2
0
7
*
7
.
7
5
£
1
0
2
0
7
*
*
*
2
9
.
2
9
£
1
0
2
0
7
*
*
*
1
.
6
9
£
1
0
2
0
7
7
.
6
0
£
1
0
2
0
7
*
*
*
(
1
.
3
9
£
1
0
2
0
7
)
(
1
.
0
2
£
1
0
2
0
7
)
(
6
.
4
2
£
1
0
2
0
8
)
(
1
.
4
2
£
1
0
2
0
7
)
(
1
.
0
7
£
1
0
2
0
7
)
(
6
.
3
3
£
1
0
2
0
8
)
N
o
n
-
r
e
v
o
l
v
i
n
g
b
a
l
a
n
c
e
s
1
.
4
8
£
1
0
2
0
7
*
*
*
2
8
.
7
9
£
1
0
2
0
8
*
*
2
6
.
0
6
£
1
0
2
0
8
1
.
4
1
£
1
0
2
0
7
*
*
*
2
8
.
4
1
£
1
0
2
0
8
*
2
5
.
7
1
£
1
0
2
0
5
(
5
.
3
5
£
1
0
2
0
8
)
(
4
.
0
9
£
1
0
2
0
8
)
(
4
.
0
1
£
1
0
2
0
8
)
(
5
.
4
0
£
1
0
2
0
8
)
(
4
.
4
0
£
1
0
2
0
5
)
(
4
.
0
4
£
1
0
2
0
8
)
N
u
m
b
e
r
r
e
v
o
l
v
i
n
g
l
o
a
n
s
2
0
.
0
0
0
2
6
1
*
2
0
.
0
0
0
3
6
6
*
*
*
0
.
0
0
0
6
2
6
*
*
*
2
0
.
0
0
0
2
2
0
2
0
.
0
0
0
4
0
6
*
*
*
0
.
0
0
0
6
2
6
*
*
*
(
0
.
0
0
0
1
4
8
)
(
0
.
0
0
0
1
0
7
)
(
0
.
0
0
0
1
2
3
)
(
0
.
0
0
0
1
5
8
)
(
0
.
0
0
0
1
2
1
)
(
0
.
0
0
0
1
2
3
)
N
u
m
b
e
r
i
n
s
t
a
l
l
m
e
n
t
l
o
a
n
s
2
7
.
2
7
£
1
0
2
0
5
2
0
.
0
0
0
2
7
0
*
0
.
0
0
0
3
4
2
*
*
*
0
.
0
0
0
2
1
7
2
0
.
0
0
0
4
4
2
*
*
*
0
.
0
0
0
2
2
6
*
*
(
0
.
0
0
0
1
5
7
)
(
0
.
0
0
0
1
4
0
)
(
0
.
0
0
0
1
0
7
)
(
0
.
0
0
0
1
7
2
)
(
0
.
0
0
0
1
5
7
)
(
0
.
0
0
0
1
0
3
)
M
o
r
t
g
a
g
e
b
a
l
a
n
c
e
s
6
.
9
6
£
1
0
2
0
8
*
*
*
2
5
.
6
2
£
1
0
2
0
8
*
*
*
2
1
.
3
4
£
1
0
2
0
8
(
2
.
0
0
£
1
0
2
0
8
)
(
2
.
1
2
£
1
0
2
0
8
)
(
1
.
0
1
£
1
0
2
0
8
)
N
u
m
b
e
r
o
f
m
o
r
t
g
a
g
e
s
0
.
0
0
0
3
8
3
0
.
0
0
0
9
6
2
2
0
.
0
0
1
3
4
(
0
.
0
0
1
6
2
)
(
0
.
0
0
1
2
2
)
(
0
.
0
0
0
8
2
4
)
O
b
s
e
r
v
a
t
i
o
n
s
5
2
,
7
2
1
5
2
,
7
2
1
N
o
t
e
s
:
S
i
g
n
i
?
c
a
n
t
a
t
:
*
p
,
0
.
1
,
*
*
p
,
0
.
0
5
a
n
d
*
*
*
p
,
0
.
0
1
;
s
t
a
n
d
a
r
d
e
r
r
o
r
s
a
r
e
i
n
p
a
r
e
n
t
h
e
s
e
s
a
n
d
a
r
e
c
l
u
s
t
e
r
e
d
b
y
s
t
a
t
e
;
e
a
c
h
s
p
e
c
i
?
c
a
t
i
o
n
a
l
s
o
i
n
c
l
u
d
e
d
t
h
e
c
o
n
t
r
o
l
v
a
r
i
a
b
l
e
s
l
i
s
t
e
d
i
n
T
a
b
l
e
I
V
Table VI.
Multinomial logit
results – marginal effects
JFEP
6,1
20
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
the Consumer Financial Protection Bureau, and BAPCPA includes a requirement that
debtors receive counseling from an agency approved by the bankruptcy administrator
or US trustee. If counseling facilitates a negotiated settlement with competing creditors,
this requirement could reduce the rush to collect that defaulting borrowers often
experience and diminish the incentive of borrowers with multiple creditors to opt for
formal bankruptcy. In addition, BAPCPA decreased federal property and homestead
exemptions, a change that could also mitigate the effects of the negative externality that
resulted from property seizure. A fruitful avenue for future research will be measuring
whether these legal changes signi?cantly altered the impact of externalities on creditor
and borrower behavior.
Notes
1. According to The Preliminary Report on Chapter 7 Asset Cases 1994 to 2000 (2001), “ the vast
majority (about 95 to 97 percent) of Chapter 7 cases yield no assets.” Chapter 7 bankruptcy is
the most common type of personal bankruptcy; more than 70 percent of non-business
bankruptcies were ?led under Chapter 7 during the late 1990s, the observation period of this
paper.
2. The data used in this paper was collected over a period of three years, during which only a
few states experienced any changes in laws governing bankruptcy and garnishment. As a
result, the empirical model is able to utilize only cross-state variation in these laws.
Garnishment law appears to be a signi?cant factor in cross-state variation in bankruptcy
rates: see Lefgren and McIntyre (2009) for an overview.
3. Brunner and Krahnen (2004) ?nd that German ?rms are more likely to negotiate a workout
and avoid bankruptcy when they have borrowed from fewer banks, facilitating lender
coordination. Williams (1998) ?nds evidence that credit counseling services mitigate
destructive creditor collection.
4. However, a preference granted soon before the borrower ?les may be nulli?ed during a
bankruptcy proceeding.
5. Since the passage of the Federal Debt Collection Procedures Practices Act in 1977,
garnishment for the collection of federal debts is determined by federal law, and therefore
legal in all states.
6. The observations in our dataset were collected between 1995 and 1997, and so the entries in
Table I re?ect exemption levels in 1997.
7. For a discussion of these violations, see Schulman (1985) and Araki (1995).
8. See Dawsey et al. (2013) for a summary of these laws.
9. Individuals with substantial debts and assets can choose to ?le under Chapter 11, and family
farmers can ?le a Chapter 12 reorganization. These chapters are rarely used by individuals
and will not be considered here.
10. Bankruptcy law underwent a major overhaul in 2005 with the enactment of the Bankruptcy
Abuse Prevention and Consumer Protection Act (BAPCPA). BAPCPA instituted means
testing and mandatory credit counseling and other hurdles to personal bankruptcy
(and Chapter 7 bankruptcy in particular). The data in this paper was collected from 1995
through 1997 and predates these changes.
11. See Scott (1989) for a discussion of the literature on lost value.
12. The administrative costs of a formal bankruptcy ?ling might include attorney fees and
?ling fees, as well as less quanti?able costs associated with social stigma or reduced ability
Creditors and
personal
bankruptcy
21
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
to borrow. In 2005, after the observation period of this paper but previous to signi?cant
changes in US bankruptcy law that increased the cost of ?ling, the GAO estimated that for a
Chapter 7 personal bankruptcy ?ling, the average attorney’s fee was $712 and the average
?ling fee was $209 (GAO, 2008). And in a small study of bankrupt married couples, 35 of 37
participants statedthat “they felt shame and stigmatization as theynegotiated the bankruptcy
process.” (Thorne and Anderson, 2006). An individual’s ?xed cost of an informal bankruptcy
would likely also include some stigma costs, but because the borrower does not make a public
declaration of bankruptcy through the legal system, these costs would likely be less than what
would follow a formal bankruptcy. The borrower in informal bankruptcy would also face
some future borrowing constraints, along with the disutility of engaging with or avoiding
collections agents.
13. In this model, the size of the loan is given, though creditors clearly take account of borrower
characteristics, including number of lenders, in assessing credit-worthiness and assigning the
terms of a credit offer. In theory, a creditor would anticipate the negative externality of
collection and therefore lend less when a borrower has many loans. However, a lender,
particularly a lender of revolving debt, balances the risk of default or bankruptcy against the
potential for fee and interest rate revenue, and a borrower with many loans may be pro?table
in the short run even if bankruptcy eventually results. Rather than modeling the complex loan
determination process, the econometric speci?cations control, for the lender’s assessment of
the borrower’s credit-worthiness using credit score and credit limit. See Miller (2011) for an
example of a two-sided model of both default and a creditor’s choice of loan and interest rate.
14. A suf?cient condition for this assumption would be that L is above a minimum threshold.
15. If all of the borrower’s assets are exempt, then, given our assumption that these borrowers
choose formal bankruptcy, the creditor receives 2 K
B
. Thus, the creditor’s payoff when
E . y does not depend on g, which removes the very low asset borrowers from the creditor’s
strategic problem that is the focus of this model.
16. See Dawsey et al. (2013) for a more detailed description of anti-harassment statutes. Other
speci?cations of state garnishment laws (e.g. using percentage garnishment or a dummy
that indicated if garnishments were restricted to less than 25 percent) yielded similar results.
State exemption laws generally include both a homestead exemption and a personal
property exemption. While the legal value of homestead exemptions are, to a large degree,
clearly speci?ed in state laws, personal exemptions are often simply described in general
terms. Estimates of the value of personal property exemptions failed to produce statistically
signi?cant results or to alter the main results of this article, and were therefore excluded
from the speci?cations reported in the tables that follow. Complete results are available upon
request.
17. According to the of?cial statistics of the Administrative Of?ce of the United States Courts,
the number of personal bankruptcies increased by nearly 73 percent between 1994 and 1997.
References
Araki, L.A.S. (1995), “Rx for abusive debt collection practices: amend the FDCPA”, University of
Hawaii Law Review, Vol. 17, pp. 69-111.
Baird, D.G. (1998), “Bankruptcy’s uncontested axioms”, Yale Law Journal, Vol. 108 No. 3,
pp. 573-599.
Block-Leib, S. (1993), “Fishing in muddy waters: clarifying the common pool analogy as applied
to the standard for commencement of a bankruptcy case”, American University Law
Review, Vol. 42 No. 2, pp. 337-431.
JFEP
6,1
22
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Brunner, A. and Krahnen, J.P. (2004), “Multiple lenders and corporate distress: evidence on debt
restructuring”, CEPR Discussion Paper 4287.
Dawsey, A. and Ausubel, L.M. (2004), Informal Bankruptcy, University of Maryland, College
Park, MD, September.
Dawsey, A., Hynes, R.M. and Ausubel, L.M. (2013), “Non-judicial debt collection and the
consumer’s choice among repayment, bankruptcy, and informal bankruptcy”, American
Bankruptcy Law Journal, Vol. 87 No. 1, pp. 1-26.
GAO (2008), Bankruptcy Reform: Dollar Costs Associated with the Bankruptcy Abuse Prevention
and Consumer Protection Act of 2005, United States Government Accountability Of?ce,
Washington, DC.
Hynes, R.M. (2004), “Why (consumer) bankruptcy?”, Alabama Law Review, Vol. 56 No. 1,
pp. 121-179.
Jackson, T.H. (1986), The Logic and Limits of Bankruptcy Law, Harvard University Press,
Cambridge, MA.
Kripke, H. (1985), “Law and economics: measuring the economic ef?ciency of commercial
law in a vacuum of fact”, University of Pennsylvania Law Review, Vol. 133 No. 5,
pp. 929-985.
Lefgren, L. and McIntyre, F. (2009), “Explaining the puzzle of cross-state differences in
bankruptcy rates”, Journal of Law and Economics, Vol. 52 No. 2, pp. 367-393.
Miller, M.M. (2011), Who Files for Bankruptcy? State Laws and the Characteristics of Bankrupt
Households, The State University of New Jersey, Rutgers, December.
Schulman, D.A. (1985), “The effectiveness of the federal fair debt collection practices act
(FDCPA)”, Emory Bankruptcy Developments Journal, Vol. 2, pp. 171-199.
Scott, R.E. (1989), “Rethinking the regulation of coercive creditor remedies”, Columbia Law
Review, Vol. 89 No. 4, pp. 730-788.
Warren, E. (1987), “Bankruptcy policy”, University of Chicago Law Review, Vol. 54 No. 3,
pp. 775-814.
Warren, E. (1993), “Bankruptcy policymaking in an imperfect world”, Michigan Law Review,
Vol. 92 No. 2, pp. 336-387.
Williams, W. (1998), Games Creditors Play: Collecting from Overextended Consumers, Carolina
Academic Press, Durham, NC.
Zywicki, T.J. (2005), “An economic analysis of the consumer bankruptcy crisis”, Northwestern
University Law Review, Vol. 99 No. 4, pp. 1463-1541.
Further reading
Ausubel, L.M. (1991), “The failure of competition in the credit card market”, American Economic
Review, Vol. 81 No. 1, pp. 50-81.
Ausubel, L.M. (1997), “Credit card defaults, credit card pro?ts, and bankruptcy”, American
Bankruptcy Law Journal, Vol. 71, pp. 249-270.
Ausubel, L.M. (1999), Adverse Selection in the Credit Card Market, University of Maryland,
College Park, MD, June.
Baird, D.G. and Jackson, T.H. (1983), “Possession and ownership: an examination of the scope of
Article 9”, Stanford Law Review, Vol. 35 No. 2, pp. 175-212.
Barth, J.R., Cordes, J.J. and Yezer, A.M.J. (1986), “Bene?ts and costs of legal restrictions on
personal loan markets”, Journal of Law & Economics, Vol. 25, pp. 357-380.
Creditors and
personal
bankruptcy
23
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
Domowitz, I. and Sartain, R.L. (1999), “Determinants of the consumer bankruptcy decision”,
Journal of Finance, Vol. 54 No. 1, pp. 403-420.
Fishman, C. (1998), “Deconstructing the debt head”, American Demographics, Vol. 20 No. 12,
pp. 60-65.
Gross, K. (1997), Failure and Forgiveness: Rebalancing the Bankrutpcy System, Yale University
Press, New Haven, CT.
Hausman, J. and McFadden, D. (1984), “Speci?cation tests for the multinomial logit model”,
Econometrica, Vol. 52 No. 5, pp. 1219-1240.
Hensher, D.A. (1986), “Sequential and full information estimation maximum likelihood
estimation of a nested logit model”, The Review of Economics and Statistics, Vol. 68 No. 4,
pp. 657-667.
Hynes, R.M. (2008), “Broke but not bankrupt: consumer debt collection in state court”, Florida
Law Review, Vol. 60 No. 1, pp. 1-62.
Korobkin, D.R. (1991), “Rehabilitating values: a jurisprudence of bankruptcy”, Columbia Law
Review, Vol. 91 No. 4, pp. 717-789.
Lawrence, D.B. (1992), Handbook of Consumer Lending, Prentice-Hall, Englewood Cliffs, NJ.
Letsou, P.V. (1995), “The political economy of consumer credit regulation”, Emory Law Journal,
Vol. 44 No. 2, pp. 587-677.
Maddala, G.S. (1983), Limited-Dependent and Qualitative Variables in Econometrics, Cambridge
University Press, Cambridge.
Moss, D.A. and Johnson, G.A. (1999), “The rise of consumer bankruptcy: evolution, revolution, or
both?”, American Bankruptcy Law Journal, Vol. 73 No. 1, pp. 311-351.
Nelson, J.P. (1999), “Consumer bankruptcy and chapter choice: state panel evidence”,
Contemporary Economic Policy, Vol. 17 No. 4, pp. 552-566.
Peterson, R.L. and Aoki, K. (1984), “Bankruptcy ?lings before and after implementation of the
bankruptcy reform law”, Journal of Economics and Business, Vol. 36 No. 1, pp. 95-105.
Train, K. (1993), Qualitative Choice Analysis, The MIT Press, Cambridge, MA.
Warren, C. (1935), Bankruptcy inUnited States History, HarvardUniversityPress, Cambridge, MA.
Corresponding author
Amanda E. Dawsey can be contacted at: [email protected]
JFEP
6,1
24
To purchase reprints of this article please e-mail: [email protected]
Or visit our web site for further details: www.emeraldinsight.com/reprints
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
This article has been cited by:
1. Ilona Alisauskaite-Seskiene, Rita Remeikiene, Ligita Gaspareniene. 2015. The Factors that Determine
Physical Entities’ Borrowing: Lithuanian Case. Procedia Economics and Finance 26, 616-622. [CrossRef]
D
o
w
n
l
o
a
d
e
d
b
y
P
O
N
D
I
C
H
E
R
R
Y
U
N
I
V
E
R
S
I
T
Y
A
t
2
1
:
4
8
2
4
J
a
n
u
a
r
y
2
0
1
6
(
P
T
)
doc_780941622.pdf