Arbitration Preliminary Report

DECEMBER 12, 2013

Arbitration Study
Preliminary Results
Section 1028(a) Study Results To Date

Table of contents
1.

Introduction .......................................................................................................... 4
1.1

What is a pre-dispute arbitration clause? ................................................ 4

1.2 The Bureau’s mandate to study consumer arbitration ............................ 8
1.3 What does this presentation cover? ......................................................... 9
2.

Summary of results to date ............................................................................... 12

3.

Clause incidence and features ......................................................................... 16
3.1 Prior studies ............................................................................................ 17
3.2 Clause incidence ...................................................................................... 19
3.3 Clause length and complexity ................................................................ 28
3.4 Clause features ....................................................................................... 30
3.5 Clause changes ........................................................................................53

4.

Incidence and typology of consumer arbitration filings................................. 58
4.1 Data sources ........................................................................................... 58
4.2 Case incidence ........................................................................................ 62
4.3 Representation ....................................................................................... 70
4.4 Amounts at issue .....................................................................................76
4.5 Non-collection arbitrations .................................................................... 82
4.6 Debt collection arbitrations ................................................................... 92
4.7 Which consumers bring arbitrations? ................................................... 96

2

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4.8 Prior litigation ....................................................................................... 101
4.9 Initial fee allocation............................................................................... 110
5.

Small claims court data ................................................................................... 118
5.1 Previous studies ....................................................................................120
5.2 Data sources .......................................................................................... 121
5.3 Incidence ............................................................................................... 124

6.

Future work ...................................................................................................... 129

Appendix A.............................................................................................................. 132
Clause incidence methodology ...................................................................... 132
Appendix B.............................................................................................................. 136
Arbitration data procedures .......................................................................... 136
Appendix C.............................................................................................................. 149
Identifying credit card cases in federal district courts .................................. 149
Appendix D.............................................................................................................. 151
Additional arbitration data ............................................................................ 151
Appendix E .............................................................................................................. 155
Small claims court data and methodology .................................................... 155
Appendix F .............................................................................................................. 167
Defined terms ................................................................................................. 167

3

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

1. Introduction
In section 1028(a) of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010,
Congress instructs the Consumer Financial Protection Bureau (the “Bureau”) to study the use of
pre-dispute arbitration contract provisions in connection with the offering or providing of
consumer financial products or services, and to provide a report to Congress on the same topic.
The Bureau is in the process of conducting that study. This document presents preliminary
results reached in the Bureau’s study to date. These results are subject to revision when the
Bureau submits its statutorily-mandated report to Congress if further analysis so warrants.
Furthermore, the Bureau’s report to Congress will contain additional analyses that the Bureau is
planning to conduct but for which the Bureau does not yet have even preliminary results. A
concluding section identifies work that the Bureau may or expects to cover in later stages of its
study that will also inform its report to Congress.

1.1

What is a pre-dispute arbitration
clause?

Companies provide almost all consumer financial products and services subject to the terms of a
written contract.1 Whenever a consumer obtains a consumer financial product such as a credit
card, a checking account, or a payday loan, he or she typically receives the company’s standardform, written legal contract.
In addition to being governed by such contracts, the relationship between a consumer and a
financial service provider will generally be governed by one or more federal consumer protection

1

4

Section 1002(5) of the Dodd-Frank Act defines “consumer financial product[s] or service[s].” See 12 U.S.C. § 5481(5).

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

laws and often by state consumer protection laws. These laws create legal rights for consumers
and impose duties on financial service providers. Absent an agreement to the contrary, if a
dispute arises between a consumer and a company as to whether one side or the other is
adhering to its contractual or statutory duties, the aggrieved party generally has the right to seek
resolution of the dispute in a court of law (although some state and federal laws provide only for
public, and not private enforcement). Furthermore, the federal court system and most state
court systems provide for a class action process in which, in defined circumstances, one or more
plaintiffs may file suit on behalf of similarly-situated individuals. If such an action is certified by
the court as meeting the criteria for a class action and plaintiffs prevail or secure a settlement,
all members of the class—for example, customers of a company who have been adversely
affected by a particular practice—may be eligible to obtain relief without initiating their own
lawsuits. Conversely, if the defendant prevails in a certified class action, all members of the class
may be bound by the decision and thereby precluded from initiating their own lawsuits with
respect to the claims at issue in the class case.
As a general rule, the parties to a contract can agree to an alternative means of resolving any
disputes that arise between the parties, including both contractual disputes and disputes under
state or federal law. The most common form of alternative dispute resolution is final and
binding arbitration in which a privately-appointed individual—an arbitrator—is empowered to
resolve claims that arise between the parties.2
As discussed in detail in section 3, contract clauses that provide for pre-dispute arbitration
appear to be a common, but not a universal, feature of consumer financial contracts. These
arbitration clauses are sometimes described as “mandatory.” Under the terms of such
agreements, either side can mandate that a dispute that arises between the parties be resolved in
binding arbitration.3 The clauses are described as “pre-dispute” because they commit the parties
to this arrangement before there is a dispute between them.

2

In some cases, more than one arbitrator may be involved in resolving a dispute.

3

Alternatively, the term “mandatory,” when used to describe arbitration clauses in the consumer context, may derive
from the nature of consumer contracts. When a consumer uses a consumer financial product, he or she is usually
bound by the terms of a consumer contract. The terms of that contract are not generally open to negotiation by the
consumer, but are instead offered on a take-it-or-leave-it basis, meaning that the consumer either accepts those terms
or instead shops for another product with different standard-form terms. In legal terms, the contract is one of

5

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

These arbitration clauses generally give each party to the contract two distinct contractual
rights. First, either side can file claims against the other in arbitration and obtain a binding
decision from the arbitrator. Second, if one side sues the other in court, the party that has been
sued in court can invoke the arbitration clause to require that the party that went to court
instead proceed, if at all, in arbitration.
Arbitration clauses almost always specify the private arbitration organization that will
administer the dispute if and when the clause is invoked by the consumer or the company. The
American Arbitration Association (or “AAA”) is one such “arbitration administrator”; JAMS,
Inc. is another. For the consumer financial contracts we have reviewed to date, our research
shows that the AAA is the most commonly specified organization in such pre-dispute arbitration
clauses. We review certain AAA data in section 4 below.4
Arbitration clauses may have a number of other features that address where disputes may be
adjudicated, the timing for filing disputes, and the remedies that may be awarded. Additionally,
arbitration clauses may include limitations on the use of class proceedings and “carve-outs” for
small claims court proceedings.5
Pre-dispute arbitration has become a contentious legal and policy issue. Following a series of
Supreme Court cases interpreting the 1925 Federal Arbitration Act (the “FAA”), courts regularly
enforce pre-dispute arbitration clauses in consumer, employment, and other contexts in which
the relevant contract is not subject to negotiation between the contracting parties.6 At least from

adhesion, making the clause “mandatory” in contrast to the voluntary clauses that may be reached by negotiation
between commercial parties. See, e.g., Jean R. Sternlight, Creeping Mandatory Arbitration: Is It Just?, 57 STAN. L.
REV. 1631, 1632 n.1 (2005). Other scholars argue that the term “mandatory arbitration” may be better reserved for
arbitration that is mandated by statute or regulation. See, e.g., Ian R. Macneil et al., FEDERAL ARBITRATION LAW
§ 17.1.2.2, at 17:8-17:9 (Supp. 1999).
4

The AAA is a non-profit organization. It was founded in 1926, following enactment of the Federal Arbitration Act,
with the specific goal of helping to implement arbitration as an out-of-court solution to resolving disputes. It
describes itself as being dedicated to “the development and widespread use of prompt, effective and economical
methods of dispute resolution.” Its mission statement and other information are available at the AAA’s website,
www.adr.org.
5

We discuss these two features at sections 3.4.2 and 3.4.5.

6

Chapter 1 of the FAA is codified at 9 U.S.C. §§ 1-16. It provides that an arbitration award is final and binding, with
limited grounds available for judicial review. See 9 U.S.C. §§ 9-10. There has been an active scholarly and judicial
debate over the meaning of the FAA, particularly as it applies to consumer contracts and state court proceedings. See

6

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

the late 1990s, the inclusion of arbitration clauses in consumer financial contracts appears to
have become more common.7 In 2011, the Supreme Court issued its decision in AT&T Mobility
LLC v. Concepcion.8 In that case, the Court held that the FAA preempted state law that would
have prohibited the enforcement of a consumer arbitration clause with a “no-class” provision.
Prior to that decision, courts were divided on state law challenges to the enforceability of noclass provisions in arbitration clauses.
Some commenters take the view that pre-dispute arbitration clauses contained in standard-form
contracts are unfair to consumers. Critics generally focus on three areas. First, they attack
arbitration as a dispute resolution process. They contend that it reduces or eliminates
procedural protections—such as a right of appeal or access to discovery—that are generally
available in court.9 There are also claims that arbitration may be biased against consumers, and
that it may not be as fast or cheap as its proponents claim.10 Second, especially in the wake of the
Supreme Court’s decision in Concepcion, critics argue that arbitration clauses may immunize
companies from a range of private civil liabilities, such as by reducing the availability of
discovery or by eliminating class proceedings.11 According to this argument, arbitration clauses
may undermine deterrence and leave widespread wrongdoing against consumers unaddressed.
Finally, critics assert that arbitration, which is almost always conducted in private, undermines

David S. Schwartz, The Federal Arbitration Act and the Power of Congress over State Courts, 83 OR. L. REV. 541, 541630 (2004); Christopher R. Drahozal, Federal Arbitration Act Preemption, 79 IND. L.J. 393, 393-440 (2004).
7

For example, a number of large credit card issuers moved to include arbitration clauses in their consumer credit
card contracts from the late 1990s. The reasons for this shift are contested and the focus of a court case, Ross v. Bank
of America et al., 05-Civ. 7116, which is pending in the United States District Court for the Southern District of New
York. We discuss the Ross case further in section 3. The detailed timeline for adoption by credit card issuers is
discussed, inter alia, in the report of the Ross defendants’ expert, Professor Kenneth G. Elzinga, dated September 21,
2010. The timeline of adoption for other products may differ, of course, and adoption has not been universal, as we
discuss in section 3.2.

8

AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011).

9

See, e.g., National Association of Consumer Advocates (“NACA”) and National Consumer Law Center (“NCLC”) RFI
Comment at 5.

10

See, e.g., Americans for Financial Reform (“AFR”) RFI Comment at 5.

11

See, e.g., Public Justice RFI Comment at 4.

7

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

benefits inherent in the public nature of the court system, such as transparency and the
development of clear precedents.12
In contrast, defenders of pre-dispute arbitration clauses take the view that arbitration offers “a
faster, more efficient and more cost-effective method of resolving disputes than court
litigation.”13 According to this point of view, arbitration “minimizes the disruption and loss of
good will that often results from litigation and … reduces litigation costs.”14 Arbitration
proponents also claim that these cost savings inure to the benefit of consumers through lower
prices and/or expanded access.15 Finally, while proponents of arbitration clauses may
acknowledge the potential impact on class proceedings, many take the view that such
proceedings typically are meritless, inefficient, and provide little or no benefit to consumers.16
They contend that the reduced cost of arbitration together with various provisions of arbitration
clauses (including the availability of small claims court as well as contingent minimum awards
in arbitration) provide ample opportunity for consumers to obtain redress for asserted wrongs
that involve relatively small amounts of money.

1.2

The Bureau’s mandate to study
consumer arbitration

Section 1028(a) of the Dodd-Frank Act requires the Bureau to “conduct a study of, and . . .
provide a report to Congress concerning, the use of agreements providing for arbitration of any
future dispute between covered persons and consumers in connection with the offering or
providing of consumer financial products or services.” “Covered persons” are defined as “any

12

See, e.g., NACA & NCLC RFI Comment at 5-6.

13

American Bankers Association (“ABA”), Consumer Bankers Association (“CBA”), and The Financial Services
Roundtable (“FSR”) RFI Comment at 2.
14

Id.

15

See, e.g., American Financial Services Association (“AFSA”), RFI Comment at n.30.

16

See, e.g., ABA/CBA/FSR RFI Comment at 9 n.16.

8

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

person that engages in offering or providing a consumer financial product or service.”17 Because
section 1028(a) specifically addresses the use of pre-dispute arbitration clauses in consumer
financial contracts, those “agreements”—and the consumer disputes that may be subject to their
terms—have been and will remain the focus of the Bureau’s study.
As a preliminary step in undertaking the study, the Bureau published a Request for Information
(the “RFI”) in 2012 that sought comments on the appropriate scope, methods, and data sources
for the required study.18 We received 60 comments in response to the RFI. We refer to a number
of those comments in this presentation.19 Most comments came from trade associations,
consumer groups, academics, or law firms. We received almost no comments from individual
financial institutions that include arbitration clauses in their standard-form contracts with
consumers.

1.3

What does this presentation cover?

This presentation includes preliminary results reached in the Bureau’s study to date. As our
study effort continues, we will refine and place this work into fuller context. The Bureau is
disclosing these preliminary results at this time to provide stakeholders with greater
transparency into the work the Bureau has undertaken to date and to provide further detail on
the work the Bureau is planning to undertake before issuing its report to Congress.
This presentation has five sections:


Section 2 is an Executive Summary of the results to date.



Section 3 addresses our preliminary findings with respect to the incidence of arbitration
provisions in the three markets which we have studied thus far: consumer credit card,

17

12 U.S.C. § 5481(6)(A).

18

See Consumer Financial Protection Bureau, Request for Information Regarding Scope, Methods, and Data Sources
for Conducting Study of Pre-Dispute Arbitration Agreements 4 (Apr. 2012) (Docket No. CFPB-2012-0017).
19

We cite to these using the name of the commenter and the title “RFI Comment.” All such RFI Comments are
available on www.regulations.gov, accessible from the Bureau’s website, www.consumerfinance.gov.

9

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

checking account, and general purpose reloadable prepaid card agreements. It identifies
the extent to which these clauses are included in consumer contracts in these three
markets. It also reviews the features of these clauses.


Section 4 provides a typology of consumer disputes filed with the AAA by consumers
and/or companies from January 1, 2010 through to the end of 2012 for two of the
markets for which we have studied the incidence of arbitration clauses—the credit card
and checking account markets—and also in the payday loan market.20 We review several
data points, such as the number of filings, the legal claims that they cover, claim
amounts, initial fee assessments, whether the parties are represented by counsel, and
demographic distributions of consumers in these arbitrations. We may expand our
analysis of AAA dispute filings to other consumer financial products in the next phase of
our work.



Section 5 reviews filings in small claims courts by consumers and companies in the credit
card marketplace. Many arbitration clauses contain small claims court “carve-outs”—
generally enabling either the consumer or the company to use small claims courts, rather
than arbitration, for claims resolution. This section reviews available data in the states
and largest counties that provide electronic access sufficient for these purposes to see
how much use consumers and companies make of small claims court.



Finally, section 6 describes areas that the Bureau expects to cover in the remainder of its
study of pre-dispute arbitration provisions.

Because the Bureau’s work on this study is ongoing, any of the findings presented here may be
refined or modified when we issue our report to Congress pursuant to section 1028(a).
Furthermore, that report will provide additional context to the preliminary results included here
in at least the following four respects.


This presentation focuses on the “front-end” of formal disputes involving consumers:
who files these disputes, in what numbers, against whom, and about what. In later work,

20

We have identified only four prepaid cases in the data we received from the AAA. We have not provided data
specific to that market because there are so few of these filed disputes.

10

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

we intend to address the “back-end” of formal disputes: what happens, in how long, and
at what cost.


This presentation also focuses on filings in arbitration and in small claims court. Later
work will address consumer cases filed in federal courts and in state courts other than
small claims courts.21 Numerous commenters have said that an informed assessment of
arbitration requires some consideration of available alternative forms of formal dispute
resolution, especially class and non-class actions in court. We agree.



This presentation also looks primarily at individual disputes, although it provides some
initial data points about the interrelationship between individual and class proceedings.
Later work will look further at consumer class actions.



Finally, this presentation is limited to certain aspects of four specific consumer financial
products: consumer credit cards, checking accounts, general purpose reloadable prepaid
cards, and payday loans. Later work will cover other aspects of those products, and
potentially other consumer financial products as well.

Readers should not interpret this presentation as our assessment, preliminary or otherwise, of
the relative importance of different areas to be covered in the statutory report to Congress.
Rather, the subjects addressed here are those as to which we already have been able to obtain
and analyze sufficient data in order to make some preliminary findings. When the Bureau issues
its report to Congress, we intend to cover arbitration and court proceedings, including class
actions—at least to the extent that meaningful comparisons can be made. Only once our work
reflects that fuller focus, will we submit the report called for by section 1028(a) of the DoddFrank Act.

21

As noted by several commenters, state court data are not available from electronic sources on a comprehensive
basis.

11

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

2. Summary of results to date
While our work remains in progress, we have preliminarily reached the following conclusions
based on the data that we have reviewed to date:


In the credit card market, larger bank issuers are more likely to include arbitration
clauses than smaller bank issuers and credit unions. As a result, while most issuers do
not include such clauses in their consumer credit card contracts, just over 50% of credit
card loans outstanding are subject to such clauses. (In 2009 and 2010 several issuers
entered into private settlements in which they agreed to remove the arbitration clauses
from their credit card consumer contracts for a defined period. If those issuers still
included such clauses, some 94% of credit card loans outstanding would now be subject
to arbitration.)



In the checking account market, larger banks tend to include arbitration clauses in their
consumer checking contracts, while mid-sized and smaller banks and credit unions do
not. We estimate that in the checking account market, which is less concentrated than
the credit card market, around 8% of banks, covering 44% of insured deposits, include
arbitration clauses in their checking account contracts.



In our GPR prepaid card sample, for which data are more limited than for our credit and
checking account samples, arbitration clauses are included across the market.22 Some
81% of the cards studied, and all of the cards for which market share data are available,
have arbitration clauses in their cardholder contracts.

22

12

“GPR” prepaid cards are general purpose reloadable prepaid cards. See infra note 26.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



Nearly all the arbitration clauses studied include provisions stating that arbitration may
not proceed on a class basis. Around 90% of the contracts with arbitration clauses—
covering close to 100% of credit card loans outstanding, insured deposits, or prepaid
card loads subject to arbitration—include such no-class arbitration provisions. Although
these terms effectively preclude all class proceedings, in court or in arbitration, some
arbitration clauses also expressly waive the consumer’s ability to participate in class
actions in court.



The AAA is the predominant administrator for consumer arbitration about credit cards,
checking accounts, and GPR prepaid cards.



From 2010 through 2012, there was an annual average of 415 individual AAA cases filed
for four product markets combined: credit card, checking account, payday loans, and
prepaid cards.23 The annual average was 344 credit card arbitration filings, 24 checking
account arbitration filings, 46 payday loan arbitration filings, and one prepaid
arbitration filing. These numbers do not indicate the number of cases in which the filing
was “perfected”24 and the matter proceeded to arbitration. They indicate only the
number of filings, deficient or otherwise.



Not all these arbitration filings were made by consumers. For the three product markets
combined, the standard AAA “claim form” records consumers filing an average of under
300 cases each year.25 The remaining filings are recorded as mutually submitted or made
by companies.



From 2010 through 2012, around half the credit card AAA arbitration filings were debt
collection disputes—proceedings initiated by companies to collect debt, initiated by

23

Over three years, only four of the filings were prepaid cases.

24

The AAA views an arbitration filing as “perfected” when all initial fees have been paid, the clause is deemed
sufficient under the Due Process Protocol, the submission complies with the AAA’s debt collection moratorium (if
applicable), and the matter is otherwise ready to proceed to arbitration. We discuss the debt collection moratorium in
section 4.2.2. The Due Process Protocol is discussed at several places, including sections 4.9, and 5.

25

With prepaid cases included, the average is still under 301 cases each year. The annual average was 235 consumerfiled credit card disputes, 20 consumer-filed checking account disputes, and 44 consumer-filed payday loan disputes.

13

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

consumers to challenge the company’s claims in court for debt collection, or mutual
submissions to the same effect. More than a quarter of these debt collection arbitrations
also included non-debt consumer claims. Many of the credit card disputes that we have
not formally defined as debt collection arbitrations also include a substantive dispute
about the amount of debt owed—or not owed—by the consumer to the company.


In contrast, very few of the checking account and payday loan AAA arbitration filings
from 2010 through 2012 were debt collection arbitrations.



From 2010 through 2012, a slight majority (53%) of consumers were represented by
counsel in the AAA arbitrations that we reviewed for these three product markets. For
non-debt collection disputes, 61% of consumers had a lawyer at some point in the
arbitration proceeding. For debt collection arbitrations, 42% of consumers had legal
representation at some point in the proceeding. Companies were almost always
represented by outside or in-house counsel in both debt collection and non-collection
arbitrations.



From 2010 through 2012, almost no AAA arbitration filings for these three product
markets had under $1,000 at issue. This was true for debt collection arbitrations and
other arbitrations as well. For these three product markets combined, during the period
2010 through 2012, there were an annual average of seven arbitrations per year filed
with the AAA that concerned disputed debt amounts that were at or below $1,000. For
the same products, there were an annual average of under eight AAA arbitrations in
which there was no disputed debt amount identified and the affirmative claim amount
was at or below $1,000.



From 2010 through 2012, for arbitration filings before the AAA involving these three
products, the average alleged debt amount in dispute was $13,418. The median alleged
debt amount in dispute was $8,641. Looking only at filings that did not identify a
disputed debt amount, and excluding one high-dollar outlier, the average amount at
issue was $38,726, and the median $11,805.



Most arbitration clauses that we reviewed contain small claims court carve-outs.
In 2012, consumers in jurisdictions with a combined total population of around 85
million filed fewer than 870 small claims court credit card claims—and most likely far

14

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

fewer than that—against issuers representing around 80% of credit card loans
outstanding.


Credit card issuers are significantly more likely to sue consumers in small claims court
than the other way around. In the two top-30 counties by population in which small
claims court complaints can be directly reviewed by electronic means, there were more
than 2,200 suits by issuers against consumers in small claims court and seven suits by
consumers against those issuers. (In one of these two counties, companies are not able to
or face severe limits to bring collection claims in small claims court, so these numbers
reflect company filings in only one of the two counties. The consumer filing numbers,
however, are across both counties.)

15

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3. Clause incidence and
features
A central aspect of the use of arbitration clauses is their incidence—that is, how often they
appear in contracts. This section examines how frequently arbitration clauses appear in three
types of consumer financial services contracts: cardholder agreements for credit cards, deposit
account agreements for checking accounts, and cardholder agreements for general purpose
reloadable (GPR) prepaid cards.26 Next we consider the length and complexity of arbitration
clauses relative to the rest of the cardholder agreement. We then look in more detail at
provisions included in arbitration clauses, such as whether the clauses permit the consumer to
reject the arbitration clause, what arbitration administrators are specified, and whether they
preclude class proceedings. Finally, we present data on changes over time in the use of
arbitration clauses and selected clause features. We look, in particular, at how clauses have
changed since the Supreme Court’s 2011 decision in Concepcion.

26

A general purpose reloadable (GPR) prepaid card is a card that “a consumer can use anywhere that accepts
payment from a retail electronic payments network, such as Visa, MasterCard, American Express, or Discover” and to
which the consumer can add funds after the card is issued. See Consumer Financial Protection Bureau, Advance
Notice of Proposed Rulemaking, 77 FED. REG. 30,923, 30,923 (May 24, 2012) (Docket No. CFPB-2012-0019). For
purposes of this presentation, we limit our analysis to GPR prepaid cards that consumers can purchase at retail
outlets or over the Internet. We do not cover payroll cards or electronic benefit transfer cards, which also can be used
over electronic payment networks and can be reloaded at least by the provider of the card.

16

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3.1

Prior studies

Several prior studies have examined the use of arbitration clauses in various types of consumer
financial services contracts. In 2004, Demaine and Hensler found that 69.2% of the consumer
financial contracts in their sample included arbitration clauses.27 Because they were seeking to
determine “the frequency with which the average consumer encounters arbitration clauses,”28
they included at most five contracts from a broad range of contract types in their sample, rather
than investigating any particular type of consumer contract in detail.29 Other studies focusing
specifically on the use of arbitration clauses in credit card contracts have also relied on small
samples, typically from the largest credit card issuers.30 One such study, by Eisenberg, Miller,
and Sherwin, found that 76.9% of the consumer contracts studied included arbitration clauses,31
and that “every consumer contract with an arbitration clause also included a waiver of classwide
arbitration.”32

27

Linda J. Demaine & Deborah R. Hensler, “Volunteering” to Arbitrate Through Predispute Arbitration Clauses:
The Average Consumer’s Experience, 67 LAW & CONTEMP. PROBS. 55, 64 (2004). They included tax preparation and
investment contracts, along with credit card and banking contracts, as consumer financial contracts. Limiting their
results to credit card and banking contracts, 12 of 17 (70.6%) included arbitration clauses. Id.
28

Id. at 57.

29

Their sample included two contracts for general credit cards and five each for airline credit cards, store credit cards,
and banking contracts. See id. at 64. Demaine and Hensler also examined a number of features of the arbitration
clauses they studied, but reported only aggregate findings for all consumer contracts.

30

E.g., Public Citizen, Forced Arbitration: Unfair and Everywhere 1 (Sept. 24, 2009),
http://www.citizen.org/documents/UnfairAndEverywhere.pdf (finding that eight of ten “major providers” of credit
cards and five of seven major banks used arbitration clauses); Amy J. Schmitz, Legislating in the Light: Considering
Empirical Data in Crafting Arbitration Reforms, 15 HARV. NEGOT. L. REV. 115, 145-46 (2010) (reporting that ten of 13
credit card contracts included arbitration clauses).
31

Theodore Eisenberg, Geoffrey P. Miller & Emily Sherwin, Arbitration's Summer Soldiers: An Empirical Study of
Arbitration Clauses in Consumer and Nonconsumer Contracts, 41 U. MICH. J.L. REFORM 871, 883 table 2 (2008).
Their sample consisted of “26 consumer agreements drafted by 21 companies,” several of which were consumer
financial services companies: three commercial banks (five consumer agreements), two credit card issuers (two
consumer agreements), and one financial credit company (one consumer agreement). Id. at 881.
32

17

Id. at 884.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Rutledge and Drahozal examined the incidence of arbitration clauses in credit card contracts
using a much larger sample made available under the Credit CARD Act of 2009.33 They found
that between 2009 and 2010, “the percentage of [credit card] issuers using arbitration clauses
declined from 17.4% … to 15.0%,” reflecting a net decrease of eight issuers, and that “the
percentage of credit card loans subject to arbitration clauses declined from 95.1% to only
48.0%.”34 This study attributed the decline to two events: (1) the National Arbitration Forum
ceasing to administer consumer arbitrations following its settlement of a consumer fraud
lawsuit filed by the Minnesota Attorney General; and (2) the settlement of an antitrust class
action, Ross v. Bank of America, by four large credit card issuers, under which they agreed to
remove the arbitration clauses from their credit card contracts for a three-and-one-half-year
period.35 This study also examined the use of various features of credit card arbitration clauses,
ranging from arbitration selection terms to class arbitration waivers.36
In November 2012, the Pew Charitable Trusts issued a study of the use of arbitration clauses in
the checking account contracts used by 100 large financial institutions.37 The study found that
43% of the institutions in the sample used arbitration clauses, with a “wide disparity” between

33

Peter B. Rutledge & Christopher R. Drahozal, Contract and Choice, 2013 BYU L. REV. 1; see also Christopher R.
Drahozal & Peter B. Rutledge, Arbitration Clauses in Credit Card Agreements: An Empirical Study, 9 J. EMPIRICAL
LEGAL STUD. 536 (2012). The relevant statutory provision is Section 204(a) of the Credit Card Accountability
Responsibility and Disclosure Act of 2009, Pub. L. No. 111-24, 123 Stat. 1746 (May 22, 2009).
34

Rutledge & Drahozal, Contract and Choice, supra note 33, at 19-20.

35

Id. at 18-19; see also infra note 51. The Ross case has continued against several other non-settling defendants. The
main allegations in Ross are that a group of large credit card issuers colluded to include arbitration clauses, including
class waivers, in credit card contracts. See First Am. Class Action Compl., Ross v. Bank of America, No. 05 CV 07116,
¶¶ 96-119 (S.D.N.Y. June 4, 2009), http://www.arbitration.ccfsettlement.com/documents/files/2009-06-04-1stamended-complaint.pdf. The case remains pending.

36

Rutledge & Drahozal, Contract and Choice, supra note 33, at 21-49.

37

Pew Charitable Trusts, Banking on Arbitration: Big Banks, Consumers, and Checking Account Dispute Resolution
(Nov. 2012), http://www.pewstates.org/uploadedFiles/PCS_Assets/2012/Pew_arbitration_report.pdf; see also
Public Citizen, supra note 30, at 10-11 (deposit account agreements); Pew Health Group, Hidden Risks: The Case for
Safe and Transparent Checking Accounts 18 (Apr. 2011),
http://www.pewtrusts.org/uploadedFiles/wwwpewtrustsorg/Reports/Safe_Checking_in_the_Electronic_Age/Pew_
Report_HiddenRisks.pdf (“For 189 of these [265] accounts (representing four out of 10 banks and 71 percent of all
accounts), the accountholder had to waive the right to a trial before a judge and agree to have the dispute resolved
before a private arbiter of the bank’s choice.”).

18

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

the 50 largest (with 56% providing for arbitration) and the remainder of the sample (with 30%
providing for arbitration).38 The study also reported on various other features of the arbitration
clauses, finding, for example, that “[o]f the institutions in the top 50 that have arbitration
clauses, 81 percent have class action bans,” while “[f]or the next 50 institutions, this number
drops to 62 percent.”39
No prior study of which we are aware has examined systematically the use of arbitration clauses
in GPR prepaid contracts.40

3.2

Clause incidence

This section presents data on how frequently arbitration clauses are used in credit card
contracts, checking account contracts, and GPR prepaid card contracts. The patterns are similar
for credit card and checking account contracts: most institutions do not use arbitration clauses,
and credit unions typically do not, but larger institutions are more likely to use arbitration
clauses than smaller institutions. As a result, a slight majority of credit card loans outstanding
within the scope of our contract sample are subject to arbitration. Similarly, we estimate that
just under half of insured deposits at banks are subject to arbitration.41 By comparison, most
GPR prepaid card agreements in our sample, covering the substantial majority of the amounts
loaded on such cards, use arbitration clauses. For credit cards, however, the impact of the Ross
settlement is potentially significant.42 While that settlement only removed four issuers from the

38

Pew Charitable Trusts, supra note 37, at 3-4.

39

Id. at 5. For other features studied, see id. at 4-6 (e.g., opt-outs, carve-outs for small claims court, discovery limits,
required qualifications for arbitrators, remedy limitations, and shortened statutes of limitations).
40

For anecdotal reports on the use of arbitration clauses in GPR prepaid contracts, see, e.g., Pew Charitable Trusts,
Loaded with Uncertainty: Are Prepaid Cards a Smart Alternative to Checking Accounts? 29 app. B (Sept. 2012),
http://www.pewstates.org/uploadedFiles/PCS_Assets/2012/Prepaid_Checking_report.pdf; Martin Eakes, Are
Prepaid Credit Cards Helping or Hurting Consumers? (May 25, 2012), http://www.responsiblelending.org/creditcards/policy-legislation/regulators/are-prepaid-credit-cards.html.
41

We use the term “insured deposits” to refer to the amount of deposits in accounts less than $250,000.

42

The claims at issue in Ross were directed at credit card contracts. Neither the claims in the case nor the settlements
reached checking account contracts or contracts for any other consumer financial product.

19

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

category of arbitration users, those four issuers were responsible for around 43% of credit card
loans outstanding as of the end of 2012.
We refer to credit card loans outstanding, insured deposits, and GPR prepaid card load volume
as “account values.” In this section, we report incidence both as a percentage of all companies in
the samples and as a percentage of all account values in the samples.43 As we explain more fully
in Appendix A, our measurements of account value for credit cards and checking accounts use
data from public “call reports” filed with regulators by banks and credit unions.44 Call reports do
not report consumer deposit volume separate from commercial deposit volume. We use
“insured deposits,” therefore, as a proxy for consumer deposits.45 Our account value results for
checking accounts should be viewed accordingly.
The scope of the arbitration clauses that are included in consumer contracts for these products
tends to be very broad. Typically, the arbitration clause will apply to all disputes arising out of or
relating to the contract and account or card, and sometimes it extends to other aspects of the
parties’ relationship. Exceptions to the arbitration clause take the form of carve-outs, such as the
small claims court carve-out discussed in section 3.4.2, which exclude certain types of claims
from arbitration.

3.2.1

Credit cards

The contracts in our sample cover almost all consumers in the credit card market. The sample
on which the findings reported here are based consists of 393 contracts filed by credit card
issuers with the Bureau as of December 31, 2012.46 Under applicable rules, all credit card issuers

43

By comparison, when we describe the features of arbitration clauses, we present our results as the percentage of
arbitration clauses (rather than all contracts) in the samples and as a percentage of account values subject to an
arbitration clause (rather than all account values) in the samples.
44

Call reports are available at the website for the Federal Financial Institutions Examination Council. See
www.ffiec.gov.

45

To the extent that our proxy includes commercial deposits that are not subject to arbitration, we will overstate the
amount of insured deposits subject to arbitration clauses. In general, however, we refer to the share of such deposits
subject to arbitration, which should minimize the impact of using this proxy. For more information, see Appendix A.

46

The agreements are available at Consumer Financial Protection Bureau, Credit Card Agreement Database,
http://www.consumerfinance.gov/credit-cards/agreements/.

20

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

are required to file agreements if they have more than 10,000 open credit card accounts.47
Although many issuers file more than one credit card contract, in almost every case the
contracts for a particular issuer contain the same dispute resolution clause. (In the rare cases of
inconsistency for a single issuer, we used the predominant form for that issuer as reflected in
their filings with the Bureau.) Accordingly, the sample includes one contract per issuer.48

FIGURE 1:

CLAUSE INCIDENCE BY NUMBER OF CREDIT CARD ISSUERS 2012

Arbitration
provisions,
17%

No
provisions,
83%

Of the 393 credit card issuers, 67 issuers (or 17%) included arbitration clauses in their credit
card contracts, while 326 issuers (or 83%) did not.49 Larger issuers (as measured by the dollar
value of credit card loans outstanding) and banks (as compared to credit unions) were more
likely to use arbitration clauses. Thirteen of the 20 largest bank issuers (or 65%) used arbitration
clauses. Of the 50 largest bank issuers, 29 (or 58%) used arbitration clauses, and of the
remaining bank issuers in the sample, 25 of 56 (or 44.6%) used them. However, only nine of 275

47

See 12 CFR § 1026.58(c)(5)(i). Issuers are not required to provide agreements for a private label credit card
program with less than 10,000 open accounts. See id. § 1026.58(c)(6)(i). (A private label credit card is a credit card
issued or managed by a financial institution on behalf of a merchant for use only to make purchases at that
merchant—for example, a department store credit card.)

48

For additional description of the sample, see Appendix A.

49

One issuer provided for arbitration only of disputes involving its credit card rewards program and another only for
disputes arising out of credit insurance for credit card loans. Because the agreements did not include a generally
applicable arbitration clause, they were coded as not providing for arbitration.

21

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

credit union issuers (or 3.3%) used arbitration clauses.50 These results are summarized below in
Table 1.
TABLE 1:

CLAUSE INCIDENCE IN CREDIT CARD CONTRACTS BY TYPE OF ISSUER 2012

Arbitration clause

No arbitration clause

# of contracts

% of credit card
loans outstanding

# of contracts

% of credit card
loans outstanding

50 largest bank
issuers

29
(58.0%)

52.2%

21
(42.0%)

47.8%

Other bank
issuers

25
(44.6%)

31.1%

31
(55.4%)

68.9%

Credit unions

9
(3.3%)

2.5%

266
(96.7%)

97.5%

Other issuers

4
(33.3%)

Not available

8
(66.7%)

Not available

Total

67
(17.0%)

50.2%

326
(83.0%)

49.8%

Overall, for the issuers in our sample, 50.2% of credit card loans outstanding were subject to
contracts with arbitration clauses; 49.8% of credit card loans outstanding were not. For the 50
largest bank issuers, 52.2% of credit card loans outstanding were subject to arbitration clauses.
By comparison, for the remaining (i.e., smaller) banks in the sample, 31.1% of credit card loans
outstanding were subject to arbitration, while for credit unions, only 2.5% were.
As noted, however, in late 2009, four of the ten largest issuers (Bank of America, Capital One,
Chase, and HSBC) settled an antitrust class action by agreeing to remove the arbitration clauses
from their credit card contracts for three-and-one-half years from a date specified in the

50

Four of 12 other issuers (e.g., retailers and finance companies) used arbitration clauses. Data on credit card loans
outstanding are missing for two bank issuers, one of which includes an arbitration clause in its credit card contract.
Another nine bank and credit union issuers reported zero credit card loans outstanding as of December 31, 2012.
Eight of those nine issuers used arbitration clauses.

22

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

settlement agreement.51 Those four companies include the three largest credit card issuers that
currently do not use arbitration clauses.52 Collectively, their credit card loans outstanding
constitute 86.8% of the outstandings that are not subject to arbitration clauses. Had the settling
defendants in Ross continued to use arbitration clauses, then nearly 94% of credit card loans
outstanding would be subject to arbitration clauses.

51

The dates differed among the settling defendants, and ranged from January 1, 2010, to May 1, 2010. The settlement
agreements also provided for the period to be extended in the event of a delay in mailing change-of-terms notices to
cardholders. See Stipulation and Agreement of Settlement with Bank of America, N.A. (USA) (N/K/A/ FIA Card
Services, N.A.) and Bank of America, N.A., ¶ 3(b), Ross v. Bank of America, N.A., (USA), No. 05-cv-7116 (S.D.N.Y.
Feb. 23, 2010), http://www.arbitration.ccfsettlement.com/documents/files/2010-02-23-stip-and-agreement-withbank-of-america.pdf; Stipulation and Settlement Agreement with Capital One Bank (USA), N.A. and Capital One,
N.A., ¶ 3(b), Ross v. Bank of America, N.A., (USA), No. 05-cv-7116 (S.D.N.Y. Feb. 23, 2010),
http://www.arbitration.ccfsettlement.com/documents/files/2010-02-23-stip-and-agreement-with-capital-one.pdf;
Stipulation and Agreement of Settlement with JPMorgan Chase & Co. and Chase Bank USA, N.A., ¶ 3(b), Ross v.
Bank of America, N.A., (USA), No. 05-cv-7116 (S.D.N.Y. Feb. 23, 2010),
http://www.arbitration.ccfsettlement.com/documents/files/2010-02-23-stip-and-agreement-with-chase.pdf;
Stipulation and Agreement of Settlement with HSBC Finance Corp. and HSBC Bank Nevada, N.A., ¶ 3(b), Ross. v.
Bank of America, N.A., (USA), No. 05-cv-7116 (S.D.N.Y. Feb. 24, 2010),
http://www.arbitration.ccfsettlement.com/documents/files/2010-02-24-stip-and-agreement-with-hsbc.pdf.
52

The other issuer, HSBC, sold a substantial portion of its credit card portfolio to Capital One in 2012. See Howard
Mustoe, HSBC Credit Card Sale to Capital One Yields $2.5 Billion Premium, Bloomberg.com (May 2, 2012),

http://www.bloomberg.com/news/2012-05-02/hsbc-gets-31-3-billion-in-cash-for-u-s-card-and-retailsale.html. As of December 31, 2011, HSBC filed agreements for private label credit cards that included arbitration
clauses. Capital One appears to have acquired those accounts as of December 31, 2012, and those cardholder
agreements no longer include arbitration clauses.

23

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 2:

CLAUSE INCIDENCE BY CREDIT CARD LOANS OUTSTANDING 2012

No
provisions,
7%

No
provisions
(Ross
settlement),
43%

3.2.2

Arbitration
provision,
50%

Checking accounts

Our sample of checking account contracts is also designed to provide an overall view of the
checking account market. It consists of available contracts from three sources: (1) the 100
largest banks, measured by insured deposits on a consolidated basis53 as of December 31, 2012;
(2) a random set of 150 mid-sized and small banks, identified using the same consolidated
measure; and (3) the 50 largest credit unions, also measured by insured deposits as of December
31, 2012.54 In a handful of cases, different banks in the same holding company used different
dispute resolution clauses; those banks were included separately in the sample, instead of on a

53

By consolidated basis, we mean that we calculated total insured deposits for all affiliated institutions, unless an
affiliate used a different dispute resolution clause. See Appendix A.
54

For companies we selected to be in our sample (either because of their size or because they were part of the random
selection), we primarily obtained agreements via one of two steps. First, for each credit union and depository
subsidiary within the sample, we sought to identify a current web-sourced agreement. (We took the most recent
agreement available for the institution’s website, but we did not assume that any web-sourced agreement with an
effective date earlier than mid-2011 was current.) When no such agreement was available via that first step, we sent a
request for a standard-form checking account contract, pursuant to the Bureau’s market monitoring authority under
Section 1022(c)(4) of the Dodd-Frank Act. The response rate to that 1022(c)(4) request was over 95% on a
consolidated basis at the time we finalized our data for this publication. We obtained agreements for four bank
holding companies via a third step of securing an agreement with no indication of a specific effective date from the
applicable website. Ultimately, we obtained agreements for all large banks, 141 out of 150 mid and small-sized banks,
and 49 of 50 credit unions. When we present the statutory report, we will include additional data we have secured by
that point.

24

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

consolidated basis.55 In addition, for a small number of banks and one credit union that were
selected to be in our sample, we were unable to obtain a standard-form contract by the time we
aggregated data for this presentation.56 Overall, our sample of checking account contracts
included in this presentation covers 49 credit unions and, on a consolidated basis, 103 large and
141 mid-sized and small banks.
Of the 103 largest banks, 47 (or 45.6%) used arbitration clauses while 56 (or 54.4%) did not.
Among the 50 largest banks, the use of arbitration clauses was higher—31 of the 50 largest
banks (or 62.0%) used arbitration clauses, while 16 of the remaining 53 large banks (or 30.2%)
used arbitration clauses. Overall, accounts representing 58.8% of insured deposits at large banks
were subject to arbitration clauses, while accounts representing 41.2% of insured deposits at
large banks were not.57 Taking just the largest 50 banks, accounts representing 61.5% of their
insured deposits were covered by arbitration clauses, while accounts representing 38.5% of their
insured deposits were not. By comparison, only 10 of the 141 mid-sized and small banks (or 7.1%
of the sample), with accounts representing 6.3% of the insured deposits in the sample, used
arbitration clauses in their checking account contracts. Finally, 4 of the 49 largest credit unions
(or 8.2%), with accounts representing 8.7% of insured deposits at the largest credit unions, used
arbitration clauses. These results are summarized in Table 2.
Extrapolating from our random sample and combining it with the large bank data, we can
provide an overall estimate of arbitration clause use by the number of banks and by bank
insured deposit volume.58 On that basis, while only 7.7% of banks use arbitration clauses for

55

This is why the sample of 100 large banks results in 103 contracts for these institutions.

56

One small institution responded that it did not offer consumer checking accounts.

57

Three of the large banks in the sample (with 0.5% of insured deposits) used jury trial waivers but provided for
arbitration in the event the jury trial waiver was unenforceable. Because arbitration was not the primary means of
dispute resolution, these banks were coded as not using arbitration.

58

For the largest banks there is no extrapolation since every bank reported. Out of the 6,320 small and mid-sized
banks, we used a randomly selected sample. Thus, our extrapolation provides an unbiased estimate. Moreover, with
the sample of 141 banks, out of which 10 reported using arbitration clauses, we can approximate the overall incidence
utilizing the Central Limit Theorem with the sample mean of 10/141=.071 and the sample variance of .071*(1.071)=.066. The sum of 6,320 variables has a mean of 448 and a standard deviation of 20.4. Thus, using the normal
approximation with a 95% confidence interval, we estimate that between 408 and 488 small and mid-sized banks use
arbitration clauses. That equates to a range of between 6.5% and 7.7% of small and mid-sized banks.

25

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

their checking account contracts, accounts representing some 44.4% of bank insured deposits
are subject to arbitration.59
TABLE 2:

CLAUSE INCIDENCE IN CHECKING ACCOUNT CONTRACTS BY TYPE OF FINANCIAL
INSTITUTION 2013

Arbitration clause

No arbitration clause

# of contracts

% of insured
deposits

# of contracts

% of insured
deposits

50 largest banks

31
(62.0%)

61.5%

19
(38.0%)

38.5%

Other large banks

16
(30.2%)

30.0%

37
(69.8%)

70.0%

Small to mid-sized
banks

10
(7.1%)

6.3%

131
(92.9%)

93.7%

50 largest credit
unions

4
(8.2%)

8.7%

45
(91.8%)

91.3%

Overall, therefore, we see for checking accounts the same phenomenon we observe for credit
cards: clause use is concentrated among large players with larger market shares. Conversely,
non-use is concentrated among small players with smaller market shares. And in both markets,
credit union use of arbitration clauses is low, by number of issuers and by market share.60

59

As of December 31, 2012, small and mid-sized banks accounted for $1,166,216,407,000 in insured deposits.
Extrapolating from our sample, for which institutions with arbitration clauses accounted for 6.3% of all insured
deposits in the sample, we estimate that $73,471,633,641 in insured deposits held by small and mid-sized banks are
subject to arbitration, and $1,092,744,773,359 are not so subject. For the largest banks, call report data show that
$1,820,588,262,000 of insured deposits are subject to arbitration, with $1,274,683,807,000 of insured deposits not
so subject. Combining the two sets of data (($73,471,633,641 +$1,820,588,262,000)/($ 73,471,633,641
+$1,820,588,262,000 + $1,092,744,773,359+ $1,274,683,807,000)) gives an estimate of 44.4% of insured bank
deposits subject to arbitration.

60

We cannot extrapolate from the sample of checking account contracts for the largest credit unions because it is not
random. That sample, however, represents more than 28.3% of all credit union insured deposit volume. Given the low
incidence of use among small banks, overall use by credit unions is likely to be even lower than the low share
indicated for the largest credit unions.

26

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 3:

3.2.3

CLAUSE INCIDENCE BY BANK AND BANK INSURED DEPOSIT VOLUME 2013

GPR prepaid cards

Our data on GPR prepaid card agreements is less complete than for the other types of contracts
studied. The sample here consists of 63 GPR prepaid cards that were listed on the Visa,
MasterCard, or NerdWallet web pages that list prepaid cards or that were included in several
recent studies of the terms of GPR prepaid cards.61 Three firms—Green Dot, H&R Block, and
NetSpend—dominated the market, collectively with over 68% of the dollar amount loaded on
cards.62
For the sample as a whole, 51 of 63 GPR prepaid card contracts (81.0%) included arbitration
clauses. All three of the leading firms, with 68.6% of the market, used arbitration clauses.
Indeed, all of the firms for which we have market share data, totaling 82.9% of the dollar

61

See Appendix A for more details.

62

See Aité Group, The Contenders: Prepaid Debit and Payroll Cards Reach Ubiquity 19 (Nov. 2012). The firms
identified in the text are formally not issuers of prepaid cards; the issuers are almost always depository institutions.
(GreenDot, however, now owns a bank issuer.) Firms like NetSpend (which has since been acquired by another
company, TSYS) are generally referred to as “program managers.” In the GPR prepaid market, the program manager
generally plays the dominant role with responsibility for most aspects of a program. Two depository institutions,
Bankcorp Bank and MetaBank, serve a large number of GPR program managers. There is no consistent pattern in the
cardholder agreements: some cards with the same program manager or the same issuing bank nonetheless have
different cardholder agreements.

27

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

amount loaded on cards, used arbitration clauses, and the substantial majority of the remaining
cards (40 of 52, or 76.9%) used arbitration clauses as well.63 Thus total load in the pre-paid
market subject to arbitration will be over—and likely substantially over—83%. The overall use of
arbitration clauses in GPR prepaid card contracts is greater than in either credit card or
checking account contracts, and much more uniform across larger and smaller players in the
market—i.e., smaller players are much more likely to use arbitration clauses in GPR prepaid
card contracts than in credit card or checking account contracts. Consistent with our results for
credit unions in the other two markets, the contracts in the sample from two credit union GPR
prepaid card programs did not use arbitration clauses, although a third program used by credit
unions and small banks does include an arbitration clause in its cardholder agreement.

3.3

Clause length and complexity

For credit card contracts with arbitration clauses, we examined various measures of the length
and complexity of the arbitration clause.64 We have not conducted the same analysis for the
other two product markets.
The word count for the credit card arbitration clauses ranged from 78 words to 2,410 words. The
mean was 1,098 words and the median was 1,074 words. On average, the arbitration clause
made up 14.1% of the words in the credit card contract (with the median at 13.1%), ranging from
1.2% to 27.5% of the words in the contract.
Credit card arbitration clauses almost always were more complex and written at a higher grade
level than the rest of the credit card contract. The mean Flesch readability score65 for credit card

63

Our market share data comes from the Aité Group report cited in note 62.

64

Three of the contracts (all from credit unions) incorporated by reference the arbitration clause in the credit union’s
membership agreement. For purposes of our analysis, we used the arbitration clause from the membership agreement
for the two membership agreements we were able to obtain rather than the incorporation-by-reference language in
the credit card contract. We treated the arbitration clause for the other credit union (which we were not able to
obtain) as missing. Accordingly, the results in this section are based on 66 credit card arbitration clauses.
65

The Flesch readability score is a widely used standard in plain language analysis. Scores range from 0.0 to 100.0,
with a higher number indicating greater readability. The calculation of the score takes into account total words, total
sentences, and total syllables.

28

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

arbitration clauses—with a higher score indicating greater readability—was 34.5 and the median
was 33.7.66 By comparison, the mean Flesch readability score for the remainder of the contract
(i.e., excluding the arbitration clause) was 52.2 and the median was 51.6.67 The readability score
for the remainder of the credit card contract exceeded the readability score for the arbitration
clause in every case.68
Similarly, the mean Flesch-Kincaid grade level69—with a lower grade level indicating greater
readability—for credit card arbitration clauses was 14.2 and the median was 14.7.70 By
comparison, the mean Flesch-Kincaid grade level for the remainder of the credit card contract
was 10.8 and the median was 11.71 Of the 66 contracts studied, only in three cases was the
Flesch-Kincaid grade level lower for the arbitration clause than for the remainder of the
contract.
Arbitration clauses from larger issuers tend to be longer than those from smaller issuers. On
average, the arbitration clause used by the 20 largest issuers contained 1,330 words, while for
the other issuers the arbitration clause contained 1,051 words. But the arbitration clauses used
by the larger issuers fare better on the readability metrics. The Flesch readability score for the
arbitration clause used by the 20 largest issuers was 37.1. For the remaining issuers it was 33.5.
The Flesch-Kincaid grade level for the arbitration clauses used by the 20 largest issuers was 13.5.
For the remaining issuers it was 14.6.72

66

The low was 18.2; the high, 51.1.

67

The low was 43.3; the high, 66.9.

68

On average, the readability score for the contract was 17.7 points higher than the score for the arbitration clause,
with the differences ranging from 2.3 to 38.9.
69

The Flesch-Kincaid Grade level translates readability to the level of education required to understand the text. A
lower grade level indicates greater readability.
70

The low was 9; the high, 20.3.

71

The low was 6.3; the high, 13.3.

72

Two of the three largest credit card issuers used two of the three most readable arbitration clauses (by both
measures) in their credit card contracts.

29

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3.4

Clause features

This section summarizes various provisions that appear in or with arbitration clauses in credit
card, checking account, and GPR prepaid card contracts. It examines: (1) whether the clause
allows the consumer to reject the arbitration clause for a limited period; (2) whether the clause
carves disputes eligible for small claims court out of the obligation to arbitrate; (3) which entity
or entities may administer the arbitration and how the arbitrator or arbitrators are to be
selected; (4) the extent to which enforceability decisions are delegated to the arbitrator;
(5) whether the clause precludes class proceedings; (6) whether the clause limits the recovery of
punitive or other damages; (7) whether the clause delimits the time period in which a claim
must be brought; (8) whether the clause precludes the disclosure of information about any
arbitration; (9) where any in-person hearing is to take place; (10) what the clause provides about
the costs of arbitration to the consumer and the company; (11) whether the clause provides for a
contingent minimum recovery; and (12) what the clauses disclose about various core
characteristics of the arbitration process. For all these variables, we present summary results
below.73
For several of the features studied, the contract provision can appear either in the arbitration
clause itself or elsewhere in the contract. For example, punitive damages or consequential
damages waivers sometimes are included in the arbitration clause, but more commonly are a
stand-alone provision of the contract. Indeed, such provisions appear not only in contracts with
arbitration clauses but also contracts without arbitration clauses. To obtain a meaningful
understanding of the incidence of these sorts of provisions requires examining not only the
features of arbitration clauses or even the features of consumer financial services contracts with
arbitration clauses, but also the features of consumer financial services contracts without
arbitration clauses. In the sections in which this type of comparison is appropriate, we report
the comparative numbers.
For each feature, we express the incidence of that feature as a percentage of the number of
arbitration clauses in the sample for that product. We also state incidence as a percentage of the

73

Some of the features described below may be inconsistent with the due process or fairness protocols applied by the
AAA and JAMS in administering consumer arbitrations. See infra text accompanying note 80. We describe the
features here as they appear in the arbitration clauses in the sample, without regard to the AAA’s or JAMS’s policies.

30

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

account values (credit card loans outstanding, insured deposits, amount loaded on prepaid
cards) of the market players in the sample subject to arbitration clauses (which, as a shorthand,
we refer to as “arbitration-subject” account values). For credit cards, we can infer that the data
reflect the overall incidence of these features for all market players who use arbitration clauses.74
For checking accounts, we can extrapolate from the two samples of banks to obtain an estimate
for overall incidence for banks that use arbitration clauses. However, this extrapolation is less
stable than it is for incidence overall given the very limited extent to which smaller banks use
arbitration clauses at all.75 For GPR prepaid cards, the market share data are limited as
discussed above.76 As a result, for checking accounts and prepaid cards we do not extrapolate
from the sample to estimate overall incidence of clause features.

3.4.1 Opt-outs
Some of the arbitration clauses in the sample permitted consumers a defined time period to optout of or reject the arbitration clause. To exercise the opt-out, a consumer must follow the stated
procedure—which generally requires a signed writing submitted by mail, and may include
requirement that all authorized users on the account consent in writing to the opt-out—within
the stated time limits.77
Just over a quarter of the credit card arbitration clauses in the sample (18 of 66, or 27.3%) and
of the checking account arbitration clauses in the sample (16 of 61, or 26.2%) included opt-outs,
as did 17.6% (9 of 51) of prepaid card arbitration clauses in the sample. The time allowed for
opting out was generally either 30 days or 60 days, typically from when the account was opened
or the agreement was mailed, depending on the arbitration clause. No agreements provided for
shorter or longer opt-out periods, and very few provided for periods (such as 45 days) between
the two ends of this range.

74

See supra text accompanying note 47.

75

Because the sample is randomly selected, the extrapolation provides an unbiased estimate. However, given that the
sample size for the small and mid-sized banks with arbitration clauses is only ten, it is arguably too small to apply the
normal approximation.
76

See supra text accompanying note 61; see also Appendix A.

77

We are not currently aware of prevailing opt-out rates.

31

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

For checking accounts and prepaid cards, larger players tended to be somewhat more likely than
smaller ones to permit consumers to opt out of the arbitration clause. Some 38.3% of
arbitration-subject insured deposits in the sample had an opt-out feature, as did 26.5% of dollar
amounts of arbitration-subject prepaid card loads. For credit cards, 26.0% of arbitration-subject
loans outstanding in the sample had an opt-out. These results are summarized in Table 3.
TABLE 3:

ARBITRATION CLAUSES PERMITTING OPT-OUTS FROM ARBITRATION 2012-13

Opt-out

No opt-out

# of contracts

% of arbitrationsubject account
values

# of contracts

% of arbitrationsubject account
values

Credit cards

18
(27.3%)

26.0%

48
(72.7%)

74.0%

Checking
accounts

16
(26.2%)

38.3%

45
(73.8%)

61.7%

Prepaid cards

9
(17.6%)

26.5%

42
(82.4%)

73.5%

3.4.2

Small claims

While arbitration clauses typically are drafted broadly to cover a wide range of claims and
disputes, most of the agreements studied “carved out,” or excluded, certain claims or disputes
from arbitration. The most common type of carve-out was for claims that either could be or had
been brought in small claims court. A small claims court carve-out is not necessary for parties to
use small claims court. What the carve-out typically provides, however, is a contractual right to
press a claim in small claims court even if the other side would prefer that the claim be resolved
in arbitration.
From 59% (checking) to 62.7% (prepaid card) to 66.7% (credit card) of arbitration clauses in the
samples included carve-outs for small claims court, with large institutions more likely to use
such carve-outs than small institutions. The 33.3% of credit card issuers that did not use a small
claims court carve-out in their arbitration clause made up only 1.0% of arbitrati0n-subject credit
card loans outstanding. Although the percentages were not as stark for checking account
contracts (41% without a carve-out, comprising 8.5% of arbitration-subject insured deposits in
the sample) and prepaid card contracts (37.3% without a carve-out, comprising between 5.6% to
15.3% of arbitration-subject loads in the sample), the pattern was the same.

32

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

The small claims court carve-outs in arbitration clauses in checking account contracts and credit
card contracts tended to apply to both companies and consumers.78 Carve-outs for consumer
small claims only, however, were the dominant form of carve-out for prepaid cards.79 These
results are summarized in Table 4.
TABLE 4:

ARBITRATION CLAUSES WITH SMALL CLAIMS COURT CARVE-OUTS 2012-13

Small claims court carve-out

No small claims court carve-out

# of contracts

% of arbitrationsubject account
values

# of contracts

% of arbitrationsubject account
values

Credit cards

44
(66.7%)

99.0%

22
(33.3%)

1.0%

Checking
accounts

36
(59.0%)

91.5%

25
(41.0%)

8.5%

Prepaid cards

32
(62.7%)

84.7%-94.4%

19
(37.3%)

5.6%-15.3%

3.4.3

Administrators and arbitrators

Arbitration clauses commonly specify a firm to administer the arbitration. The administrator is
not the arbitrator per se, although as discussed below the administrator may select the
arbitrator. The administrator generally sets out the procedural rules governing the arbitration.
In some cases, these rules may be modified by the terms of the applicable arbitration clause.
Some rules, however, may be deemed by the administrator to be not subject to contractual
modification. The two main administrators of consumer arbitration in the United States each
have due process or minimum procedural fairness protocols, and their respective rules state that

78

Some 34.4% of the checking account arbitration clauses in the sample had this feature, representing 74.2% of the
arbitration-subject insured deposits in that sample. For credit card arbitration clauses, the numbers were 39.4% of
the agreements and 62.2% of the arbitration-subject credit card loans outstanding.
79

Some 43.1% of prepaid card arbitration clauses in our sample had this feature, accounting for between 84.4% and
94.2% of arbitration-subject loads in our sample. One possible explanation for this difference is that few prepaid
cards offer a credit line, with the result that prepaid issuers generally have no small-dollar debts to collect.

33

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

they will not administer arbitrations except in accordance with those core provisions.80 Finally,
the administrator provides other administrative services, such as docketing or finding hearing
locations.
The arbitration administrator most commonly named in the clauses that we studied was the
AAA. Some 55.7% of checking account arbitration clauses in the sample listed AAA as the sole
option.81 One listed JAMS as the sole option. (One also listed the National Arbitration Forum
(“NAF”) as the sole option, even though NAF ceased administering consumer arbitrations more
than four years ago.82) Nearly half (48.5%) of credit card arbitration clauses in the sample listed
AAA as the sole option.83 Three listed JAMS and three listed NAF as sole options. More than a
third (37.3%) of prepaid arbitration clauses in the sample listed AAA as a sole option.84 For
prepaid, only one each listed JAMS or NAF as sole options.85
Counting clauses in which AAA is at least an option yields 91.8% of checking account arbitration
clauses, 83.3% for credit card arbitration clauses, and 94.1% of prepaid card arbitration clauses.
The comparable numbers for JAMS are 34.4% for checking accounts, 40.9% for credit cards,
and 52.9% for prepaid cards. By market share, the predominance of the AAA was even greater.
Over 98% of the relevant account value—whether insured deposits, credit card loans
outstanding, or prepaid load—subject to arbitration clauses in the samples listed the AAA as at
least one possible administrator. When an arbitration clause listed more than one
administrator, it typically permitted the party filing the claim to select among the

80

AAA, Consumer Due Process Protocol (Apr. 17, 1998); JAMS, Policy on Consumer Arbitrations Pursuant to PreDispute Clauses Minimum Standards of Procedural Fairness (July 15, 2009).
81

Counting clauses that listed AAA or NAF (which no longer administers consumer disputes), this share increases to
60.7%).
Consent Judgment, Minnesota v. National Arbitration Forum, Inc., No. 27-CV-09-18550 (Minn. Dist. Ct. July 17,
2009), http://pubcit.typepad.com/files/nafconsentdecree.pdf.
82

83

This increases to 50% counting clauses that list AAA or NAF.

84

This increases to 43.1% counting clauses that list AAA or NAF.

85

Other less well known administrators listed were National Arbitration and Mediation, Dispute Prevention and
Resolution, Inc., and the Arbitration Service of Portland, each included in a single arbitration clause from a small
institution.

34

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

administrators, except for prepaid card arbitration clauses, which commonly permitted the
consumer to override the company’s choice even when the company was the claimant.86
In AAA arbitrations, the AAA selects the arbitrator, subject to possible objections by the
parties.87 In JAMS arbitrations, JAMS may supply a list of arbitrators from which the parties
may choose.88 In either case, however, the default rule is that the administrator determines the
pool of prospective arbitrators, even though it does not arbitrate the dispute itself.89 Most of the
clauses within the scope of our review did not attempt to modify these default rules for
arbitrator selection. A minority specified that the arbitrator be a retired judge or an experienced
lawyer or a lawyer with expertise in the subject matter of the dispute.90 One clause for a small
bank, required “practical working experience in the banking industry.” That is the only clause
that we identified, however, to use selection criteria that might be seen as an explicit bias to one
side of the dispute.91

86

If prepaid issuers seldom sue consumers, see supra note 79, this consumer default may be equivalent to the
claimant default rule for checking and credit card.
87

See AAA, Supplementary Procedures for Consumer-Related Disputes, Rule C-4 (rules effective Sept. 15, 2005).

88

E.g., JAMS, Streamlined Arbitration Rules & Procedures, Rule 12(c) & (d) (effective July 15, 2009).

89

The administrator’s rules and applicable law typically require the arbitrator to disclose conflicts of interest, which
may provide a basis for a party to object to the arbitrator’s service. E.g., AAA, Commercial Arbitration Rules and
Mediation Procedures, Rules R-17 & R-18 (Oct. 1, 2013); JAMS, Streamlined Arbitration Rules & Procedures, supra
note 88, Rule 12(i) & (j).
90

The two most common formulations required either that the arbitrator be a lawyer with at least ten years’
experience or a retired judge (18.0% of checking account arbitration clauses, covering 16.8% of arbitration-subject
insured deposits; 30.3% of credit card arbitration clauses, covering 36.1% of arbitration-subject credit card loans
outstanding; and 3.9% of prepaid card arbitration clauses, covering 30.9% of arbitration-subject card loads), or that
the arbitrator be a practicing lawyer where the arbitration is held and have expertise in the applicable substantive law
(3.3% of checking account arbitration clauses, covering 26.7% of arbitration-subject insured deposits; 4.5% of credit
card arbitration clauses, covering 7.4% of arbitration-subject credit card loans outstanding; and 2.0% of prepaid card
arbitration clauses; no load data).
91

Compare Chavarria v. Ralphs Grocery Co., 733 F.3d 916, 924-25 (9th Cir. 2013) (clause providing for each party to
identify three candidates to serve as arbitrator; parties, beginning with respondent, take turns striking candidates;
last remaining candidate, who presumably had been initially identified by respondent, to serve as arbitrator) with
Hooters of Am., Inc. v. Phillips, 173 F.3d 933, 938 (4th Cir. 1999) (clause giving employer unilateral control of pool
from which arbitrators selected).

35

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3.4.4

Delegation

The Federal Arbitration Act allocates authority between courts and arbitrators to decide certain
legal challenges to the enforceability of arbitration clauses. As a general rule, only an arbitrator
can decide challenges to the legal validity of the overall contract that contains the clause. Courts,
however, may decide challenges to the enforceability of the arbitration clause itself, and they can
also decide whether a party assented to the contract that includes the clause.92 In Rent-A-Center
West, Inc. v. Jackson, however, the Supreme Court ruled that parties could delegate to the
arbitrator at least some issues that otherwise would be for the court to decide.93 In that case, the
Court concluded that based on the terms of the parties’ agreement, the arbitrator—not the
courts—should properly decide whether the arbitration clause was unconscionable.94 The effect
of such delegation clauses is to reduce substantially the role of courts in policing the fairness of
arbitration clauses when they are included in a contract.
Although none of the arbitration clauses in the samples directly track the language used by the
clause in Rent-A-Center, a majority of these clauses delegated to the arbitrator exclusive
authority to make decisions about the enforceability of the arbitration clause. 95 The share
ranged from 39.3% of arbitration clauses in the checking account sample, to 51.5% of credit card
clauses, to 60.8% of prepaid card clauses.96 A number of clauses, however, did the opposite: they
reserved such authority to the court through an “anti-delegation clause.” For checking accounts,
26.2% of the clauses in the sample have this feature. The numbers for credit cards and prepaid
cards are 13.6% and 7.8% respectively.

92

See Prima Paint Corp. v. Flood & Conklin Mfg. Co., 388 U.S. 395, 404 (1967) (fraudulent inducement of main
contract for arbitrator to decide); Buckeye Check Cashing, Inc. v. Cardegna, 546 U.S. 440, 449 (2006) (illegality of
main contract for arbitrator to decide).

93

130 S. Ct. 2772 (2010).

94

Id. at 2778-79.

95

Id. at 2777 (“[T]he Arbitrator, and not any federal, state, or local court or agency, shall have exclusive authority to
resolve any dispute relating to the ... enforceability ... of this Agreement including, but not limited to any claim that all
or any part of this Agreement is void or voidable.”).
96

For arbitration-subject account values, the shares are 51.6%, 46.0% and 42.6% for checking, credit cards, and
prepaid cards respectively.

36

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

These data points actually understate the extent of delegation to the arbitrator for two reasons.
First, an additional category of arbitration clauses delegated most enforceability issues to the
arbitrator, but expressly reserved to the court the exclusive authority to decide the enforceability
of any contractual limitations on class proceedings. This category appeared in 8.2% of the
checking account clauses, 25.8% of the credit card clauses, and 13.7% of the prepaid card
clauses. Second, most courts hold that language on arbitrator authority typically included in
arbitration rules promulgated by administrators has the same effect as a delegation clause
(although of course express delegation language in an arbitration clause overrides the
administrator rule).97 Because almost all of the arbitration clauses in the sample without
delegation clauses (ranging from 9.1% of credit card arbitration clauses to 17.6% of prepaid card
arbitration clauses to 26.2% of checking account arbitration clauses) nonetheless selected an
administrator, those clauses effectively also contained a delegation clause, at least under current
court decisions.98

3.4.5

Class action terms

Almost all of the arbitration clauses studied contained terms limiting class proceedings. Thus,
93.9% of the clauses in our credit card sample, 88.5% of arbitration clauses in our checking
account sample, and 96.1% of clauses in our prepaid sample did not allow arbitration to proceed
on a class basis. The handful of clauses that did not include such no-class-arbitration terms
tended to be from very small institutions. Thus, in our samples, class arbitration was unavailable
for 99.9% of arbitration-subject credit card loans outstanding, 97.1% of arbitration-subject
insured deposits, and essentially 100.0% of arbitration-subject dollar amounts loaded on
prepaid cards. An arbitration clause that does not allow class arbitration precludes any dispute
subject to arbitration from proceeding as a class action—either in court or in arbitration. These
results are summarized in Table 5.

97

See, e.g., Oracle Am., Inc. v. Myriad Group, A.G., 724 F.3d 1069, 1074 (9th Cir. 2013) (“Virtually every circuit to
have considered the issue has determined that incorporation of the American Arbitration Association’s (AAA)
arbitration rules constitutes clear and unmistakable evidence that the parties agreed to arbitrate arbitrability.”).
98

Fifteen of the 16 checking account arbitration clauses without delegation clauses specified an administrator (almost
always the AAA). Five of the six credit card arbitration clauses and eight of the nine prepaid card arbitration clauses
without delegation clauses likewise specified an administrator (again, most commonly the AAA, either by itself or
with JAMS).

37

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

TABLE 5:

ARBITRATION CLAUSES WITH NO-CLASS-ARBITRATION PROVISIONS 2012-13

No class arbitration

No provision on class arbitration

# of contracts

% of arbitrationsubject account
values

# of contracts

% of arbitrationsubject account
values

Credit cards

62
(93.9%)

99.9%

4
(6.1%)

0.1%

Checking
accounts

54
(88.5%)

97.1%

7
(11.5%)

2.9%

Prepaid cards

49
(96.1%)

100.0%

2
(3.9%)

0.0%

Some contracts with arbitration clauses also included provisions waiving the right to participate
in a class action in court, either as a named plaintiff or a member of the class, or otherwise
precluding the case from proceeding as a class action, for cases not subject to arbitration. Just
over 30% (19 of 61, covering 10.1% of arbitration-subject insured deposits) of checking account
arbitration clauses, 13.6% (9 of 66, covering 9.7% of arbitration-subject credit card loans
outstanding) of credit card arbitration clauses, and 5.9% (3 of 51; no load data) of prepaid
arbitration clauses included such provisions.99 By comparison, two checking account contracts
without arbitration clauses from the large bank sample and three from the mid-sized to small
bank sample included provisions directly waiving class actions in court. One credit card contract
without an arbitration clause and no prepaid card contracts without arbitration clauses included
such class action waivers.
Most of the checking and credit card arbitration clauses in the samples also contained an “antiseverability provision,” stating that if the no-class-arbitration terms are held unenforceable, the

99

A number of these class litigation provisions appeared outside the arbitration clause. This was the case for 13 of the
19 checking account contracts with arbitration clauses, and one of the nine credit card contracts with arbitration
clauses. The remainder appeared only within the arbitration clause. Class litigation provisions within the arbitration
clause are generally more ambiguous. They can be seen as waiving class actions in cases not subject to arbitration, but
they might instead be seen simply as stating the consequences of the arbitration clause.

38

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

entire arbitration clause is thereby rendered unenforceable as well.100 Absent that provision, a
court might hold a no-class-arbitration term unenforceable but the rest of the arbitration clause
enforceable, meaning that the dispute might then proceed as a class arbitration. With an antiseverability provision, however, if a court held the no-class-arbitration terms unenforceable the
arbitration clause would become unenforceable as well, and the case might proceed as a class
action in court rather than as a class arbitration. Close to a majority of the checking account
arbitration clauses (49.2%, covering 83.2% of arbitration-subject insured deposits) and a slight
majority of credit card arbitration clauses in the sample (54.5%, covering 66.3% of arbitrationsubject credit card loans outstanding) included such anti-severability provisions, with their use
being more likely by larger banks and issuers. But only 29.4% of prepaid card arbitration clauses
(covering 26.7% of arbitration-subject prepaid loads) had an anti-severability provision.

3.4.6

Relief limits

There is wide variation in the markets we have studied to date with respect to the inclusion in
contracts with arbitration clauses of provisions to limit damages—most commonly punitive and
consequential damages. Just over 15% of credit card contracts with arbitration clauses in the
sample, covering less than 9% of arbitration-subject credit card loans, included damage
limitations. A slight majority of the damage limitations in these credit card contracts precluded
the award of punitive damages or consequential damages or both. But many were not absolute
prohibitions, instead either requiring arbitrators to follow constitutional standards for the
award of punitive damages101 or setting out special procedures to be followed in the case of an
award of punitive damages.102

100

A severability clause generally states that if a contract provision is unenforceable that provision will be treated as
severable from the rest of the contract so that the rest of the contract remains enforceable. An anti-severability term
does the opposite—it makes one or more provisions not severable from the contract or, in this case, from the
arbitration clause.

101

Because courts usually hold that arbitration does not constitute state action, constitutional limitations on the
award of punitive damages might not otherwise apply. See, e.g., MedValUSA Health Programs, Inc. v.
MemberWorks, Inc., 872 A.2d 423 (Conn. 2005); Mave Enter., Inc. v. Travelers Indem. Co. of Conn., 2013 WL
5740159 (Cal. Ct. App. Oct. 23, 2013). For an example of such a constitutional limitation see BMW of N. Am., Inc. v.
Gore, 517 U.S. 559, 575 (1996) (holding that punitive damages award violated due process based on consideration of
“the degree of reprehensibility of the [conduct], the disparity between the harm or potential harm suffered by [the

39

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

By comparison, over 60% of checking account contracts with arbitration clauses in the sample,
covering almost 80% of arbitration-subject insured deposits, included some damages
limitation.103 (Prepaid card contracts were closer to credit card contracts in frequency.) In most
checking account contracts, the damages limitation was not in the arbitration clause but
elsewhere in the contract. Provisions precluding the award of punitive damages, consequential
damages, or both consequential damages and punitive damages, appeared in 52.5% of the
contracts (covering 69.0% of arbitration-subject insured deposits). Like the credit card
contracts, some checking account contracts (8 of 61, 0r 13.1%; 9.2% of arbitration-subject
insured deposits) made constitutional standards for the award of punitive damages applicable in
arbitration.104
A review of contracts without arbitration clauses reveals a roughly similar pattern, albeit with
damages limitations somewhat less common. Just over 35% of large bank checking account
contracts without arbitration clauses included either a consequential damages waiver or a
consequential damages waiver and a punitive damages waiver. For small to mid-sized banks,
6.1% of checking account contracts without arbitration clauses included such damages
limitations. A third of the prepaid card contracts without arbitration clauses included a
consequential damages waiver or a punitive damages waiver or both. Only one of the credit card
agreements without arbitration clauses limited the recovery of either punitive or consequential
damages.

plaintiff] and his punitive damages award, and the difference between the remedy and the civil penalties authorized
or imposed in comparable cases”).
102

The constitutional limit appears in three of 66 credit card arbitration clauses in our sample, representing 2.2% of
arbitration-subject loans outstanding. One clause, representing 4.7% of arbitration-subject credit card loans
outstanding, requires the arbitrator to follow specific procedures before making an award of punitive damages. The
required procedures include issuing a reasoned award and conducting a post-award review of punitive damages,
comparable to what would occur in court.
103

This share does not include provisions dealing with the award of consequential damages for specific types of
actions by banks, such as wrongful dishonor or errors in processing wire transfers, which are addressed specifically in
the Uniform Commercial Code. UCC §§ 4-402(b), 4A-305(c).
104

A handful of clauses purported to preclude the award of punitive damages while also authorizing the arbitrator to
award punitive damages subject to constitutional standards.

40

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3.4.7

Time limits

Few credit card and prepaid card contracts with arbitration clauses set time limits for
consumers to file claims in arbitration. Two prepaid card contracts with an arbitration clause set
a time limit of two years from when the consumer’s claim arose for the consumer to file a claim
in arbitration. (One of them applied the same time limit to company claims.) Four credit card
arbitration clauses, all from small issuers, specified time limits for consumer claims, most
commonly one year from when the claim arose. One of these issuers, however, required both the
issuer and the consumer to give the other notice of any claim within 90 days of the claim
arising.105
A greater number of checking account contracts with arbitration clauses set time limits on
consumers filing claims in arbitration, although the time limits themselves were not included in
the arbitration clause. Around 13% of the checking account contracts in the sample had such
provisions, representing 28.4% of arbitration-subject insured deposits. These generally ranged
from one to two years from when the consumer’s claim arose. Again, however, one included a
90-day notice of claim requirement for both the bank and the consumer.
Overall, the pattern was generally similar for contracts without arbitration clauses. For large
banks, 10.7% of checking account agreements without arbitration clauses had one-year time
limits for consumer claims; of the small and mid-sized banks, 1.5% had such limits. Of credit
card contracts without arbitration clauses, 2.5% had time limits, again requiring consumers to
bring claims within a year of the claim arising. Only one of the prepaid cards without arbitration
clauses had a time limit.

3.4.8

Confidentiality and nondisclosure

Unlike a judicial proceeding, arbitration as a general matter is a private process: filings are not
publicly available and hearings are not open to the public. Arbitration rules typically do not
impose express confidentiality or nondisclosure obligations on parties to the dispute, although

105

The provision added that the sending of a monthly billing statement by the issuer satisfied the issuer’s notice
obligations.

41

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

arbitrator ethics rules do impose confidentiality obligations on the arbitrator.106 Most
arbitration clauses in the sample were silent on confidentiality and did not impose any
nondisclosure obligation on the parties. Only one prepaid card arbitration clause (2.0% of
clauses; no load data) and two credit card arbitration clauses (3.0% of clauses, covering 7.3% of
arbitration-subject credit card loans outstanding) precluded the parties from making disclosures
about the arbitration proceeding, including its existence and outcome. Non-disclosure
provisions were more common in checking account arbitration clauses, where they appeared in
11.5% of clauses covering 28.0% of arbitration-subject insured deposits. By comparison, none of
the contracts without arbitration clauses imposed non-disclosure obligations on the parties to a
dispute.

3.4.9

Hearing location

The arbitration clauses in the samples generally addressed the location of in-person hearings.107
All but five prepaid card arbitration clauses (covering almost all arbitration-subject load
volume) and all but nine credit card arbitration clauses (covering 92.6% of arbitration-subject
credit card loans outstanding) addressed the issue. This feature was less common for checking
accounts, but even there 63.1% of the arbitration-subject insured deposits in the sample
addressed the hearing location.
The clauses specified a range of locations. The most common was that the hearing would be held
in the federal judicial district of the consumer’s residence. A common variation was for the
hearing to be held in the same city as the U.S. District Court closest to the consumer.108 Other

106

American Bar Association & American Arbitration Association, Code of Ethics for Arbitrators in Commercial
Disputes, Canon VI(B) (Feb. 9, 2004),
http://www.americanbar.org/content/dam/aba/migrated/dispute/commercial_disputes.authcheckdam.pdf (“The
arbitrator should keep confidential all matters relating to the arbitration proceedings and decision.”).
107

Under the AAA’s Supplementary Procedures, the default is for disputes under $10,000 to be resolved on the basis
of document submissions (which is known as a “desk hearing”), although either party may request an in-person or
telephonic hearing. AAA, Supplementary Procedures for Consumer-Related Disputes, Rule C-5 (Effective Sept. 15,
2005). For disputes over $10,000, “the arbitrator will conduct a hearing unless the parties agree not to have one.” Id.
Rule C-6.
108

A much less common variation, used by four small credit card issuers and one small bank in its checking account
contract called for the hearing to be held in the “federal judicial circuit” of the consumer’s residence.

42

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

clauses identified the consumer’s state or county as the site of the hearing. Depending on the
product market, between 3.9% to 18.0% of the clauses in the samples provided that any
arbitration hearing would be at a location “reasonably convenient” for the customer. A handful
of arbitration clauses—all associated with small institutions—identified specific cities or states in
which hearings were to be held. Similarly, two arbitration clauses in checking account contracts,
both from small banks, identified the place where the account was opened as the location of the
hearing.
Contracts without arbitration clauses also specified hearing locations using choice-of-court
clauses mandating an exclusive forum for any court case, but did so less frequently than
contracts with arbitration clauses. Of the large banks using checking agreements without
arbitration clauses, 21.4% specified the location of any court proceeding (most commonly, the
state where the account is located); 4.6% of checking account agreements without arbitration
clauses for small and mid-sized banks specified the location of any court proceeding (most
commonly the city where the contract was signed or a specific state and federal court). Only
3.7% of the credit card contracts without arbitration clauses specified the location of any court
proceeding, while 33.3% of prepaid card contracts without arbitration clauses did so.

3.4.10 Costs
In court systems, the government pays the salaries of judges and much of the cost of
administering cases, although the filing fees required when initiating a case may defray a part of
these costs. In arbitration, by contrast, the parties pay all the costs of arbitrating the dispute. As
a result, the total up-front cost of filing a claim in arbitration—at least for the parties
combined—is typically higher than the up-front cost of filing a lawsuit in court. Under the rules
adopted by the AAA and JAMS, procedural costs are generally allocated to the parties at one or
both of two stages. First, the rules set fees to be paid at the time a claim or counterclaim is filed
(and sometimes at later points in the process, such as for a hearing). Second, the rules provide
that the arbitrator may reallocate the fees between the parties in the arbitral award. The fees
covered by the arbitration rules include both fees to be paid to the administrator and fees to be
paid to the arbitrator. The parties also may incur attorneys’ fees.
In addition, the terms of an arbitration clause may address—beyond simply incorporating the
cost rules of the administrator—how these different costs will be allocated between the parties.
The arbitration clauses we studied contained three different types of cost provisions: first,
provisions addressing the initial allocation of arbitration fees; second, provisions addressing the

43

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

allocation of arbitration fees in the award; and third, provisions addressing the award of
attorneys’ fees.
Some contractual allocation of costs—beyond the default rules of the administrator—was the
norm in the clauses we studied. Only seven credit card clauses (10.6%, all from small issuers and
covering a negligible market share), 14 checking account arbitration clauses (23.0%, covering
2.4% of arbitration-subject insured deposits), and five prepaid card arbitration clauses (9.8%, all
from cards for which load data are not available) did not contain provisions altering the default
arbitration cost provisions in the administrator’s governing rules.
Many of the contracts, and in particular the checking account contracts, included general
provisions on the allocation of costs and expenses that were not specific to arbitration costs.
This document does not address such provisions or how they may interact with provisions
specifically addressing arbitration costs.
PROVISIONS ADDRESSING THE INITIAL PAYMENT OF ARBITRATION FEES
In consumer arbitration, administrative and arbitrator fees are first assessed to the parties at
filing. We refer to this as the “initial fee” allocation. Under the consumer arbitration rules of the
AAA and JAMS, initial fees are predominantly allocated to the business rather than the
consumer.109 (We discuss the AAA’s allocation in more detail in section 4.9.1.) In addition, the
administrator’s rules may bar the parties from contractually allocating a greater share of fees to
the consumer. The AAA’s rules, for example, do not permit it to administer a case in which the
consumer is required by the applicable arbitration clause to pay more at filing than the
maximum amounts stated in the AAA’s consumer fee schedule.
Some credit card arbitration clauses provided that the issuer would pay at least some of the
initial fees otherwise allocated to the consumer under the governing rules. This was true for 22
clauses (33.3%) representing 46.4% of arbitration-subject credit card loans outstanding. These
clauses provided that the issuer would pay the fees either unconditionally, for good cause, or
only if the administrator did not waive the fees, with the amount of the payment varying and
sometimes limited to amounts in excess of court fees. A slightly smaller proportion of the credit

109

E.g., AAA, Supplementary Procedures for Consumer-Related Disputes, Rule C-8 (“Costs of Arbitration”) (Rules
Effective Sept. 15, 2005; Fees Effective March 1, 2013); JAMS Policy on Consumer Arbitrations Pursuant to PreDispute Clauses, Minimum Standards of Procedural Fairness ¶ 7 (effective July 15, 2009).

44

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

card clauses (15 clauses, or 22.7%, covering 43.2% of arbitration-subject outstandings) stated
that the issuer would advance at least some portion of the consumer’s arbitration fees under
specified circumstances. Finally, eleven clauses used by small issuers (16.7% of clauses, covering
a negligible share of outstandings) indicated that the issuer would consider paying or advancing
the consumer’s arbitration fees, either on request or if the administrator does not waive the fees.
These results are summarized in Table 6.
TABLE 6:

ARBITRATION CLAUSE PROVISIONS ADDRESSING THE INITIAL PAYMENT OF ARBITRATION
FEES 2012-13

# of contracts

% of arbitration-subject
account values

Company will pay some or all fees
Credit cards

22
(33.3%)

46.4%

Checking accounts

27
(44.3%)

43.7%

Prepaid cards

18
(35.3%)

32.0%-41.8%

Credit cards

15
(22.7%)

43.2%

Checking accounts

8
(13.1%)

16.0%

Prepaid cards

7
(13.7%)

31.2%

Company will advance some or all fees

Company will consider or advancing paying some or all fees
Credit cards

11
(16.7%)

0.2%

Checking accounts

2
(3.3%)

0.4%

Prepaid cards

17
(33.3%)

27.0%-36.8%

Similarly, 44.3% of checking account arbitration clauses (43.7% of arbitration-subject insured
deposits) provided that the institution would pay or reimburse some portion of the consumer’s
share of the initial arbitration fees. Again, the prerequisites and amounts varied, with some
contracts requiring good cause or that the administrator not waive the fees, and some only
paying the amount in excess of court filing fees. A smaller number (8 clauses, or 13.1%, covering

45

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

16.0% of arbitration-subject insured deposits) provided that the institution would advance the
arbitration fees under the specified circumstances. Two clauses (3.3%; 0.4% of arbitrationsubject insured deposits) stated that the institution would consider paying the fees. These
results are summarized above in Table 6.
Prepaid card arbitration clauses most commonly provided that the institution would consider
advancing the consumer’s share of arbitration fees (17 clauses, or 33.3%, covering between 27%
and 36.8% of card loads); would advance the consumer’s arbitration fees (7 clauses, or 13.7%,
covering 31.2% of card loads ); or would simply pay the consumer’s arbitration fees, either in
their entirety (14 clauses, or 27.5%, covering between 5.6% and 15.3% of card loads), to the
extent the fees exceed filing fees in court (1 clause, 2.0%; no data on loads), up to $500 (1 clause,
2.0%; no data on loads), or for claims under $50,000 to $75,000 (2 clauses, or 3.9%, covering
26.5% of card loads). These results are also summarized above in Table 6.
PROVISIONS ADDRESSING THE ALLOCATION OF ARBITRATION FEES IN THE AWARD
The rules of arbitration administrators may permit the arbitrator to allocate all arbitration fees
between the parties in the award, including reallocating initial fees from one party to the other.
As we explain further in section 4.9.1, prior to March 1, 2013, the default AAA rules provided for
such reallocation. (From that date, however, the AAA rules restrict reallocation.110) The JAMS
Streamlined Arbitration Rules also allow for such reallocation, and the JAMS Minimum
Standards of Procedural Fairness for consumer arbitrations do not appear to restrict the
practice, except for cases in which California law so requires.111 Our review in this section uses
the situation under the JAMS rules and the AAA rules in force from 2010 through 2012 as the
baseline for comparison.

110

The new fee schedule, effective March 1, 2013, provides that “[a]rbitrator compensation … and administrative fees
(which include Filing and Hearing Fees) are not subject to reallocation by the arbitrator(s) except pursuant to
applicable law or upon the arbitrator’s determination that a claim or counterclaim was filed for purposes of
harassment or is patently frivolous.” AAA, Supplementary Procedures for Consumer-Related Disputes, Rule C-8. The
credit card contracts studied all predate the revision to the AAA rules. The checking account and prepaid card
contracts also may not reflect the revision to the rules.
111

JAMS Streamlined Arbitration Rules & Procedures, supra note 88, Rule 19(e); JAMS Minimum Standards of
Procedural Fairness, supra note 109, ¶ 8 (“In California, the arbitration provision may not require the consumer to
pay the fees and costs incurred by the opposing party if the consumer does not prevail.”).

46

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Arbitration clauses took noticeably different approaches to the allocation of arbitration fees in
the arbitrator’s award.112 First, a number of credit card arbitration clauses (23 clauses, or 34.8%,
covering 21.8% of arbitration-subject credit card loans outstanding) expressly permitted the
arbitrator to shift the payment of arbitration fees from the issuer to the consumer, as the default
AAA and JAMS rules already allowed.113 All but one of the clauses permitted the converse as
well—shifting fees from the consumer to the issuer. Second, a smaller number (5 clauses, or
7.6%; 17.0% of arbitration-subject credit card loans outstanding) expressly precluded such
shifting arbitration fees from the issuer to the consumer. Third, others (18 clauses, or 27.3%;
51.5% of arbitration-subject credit card loans outstanding) affirmatively permitted the consumer
to recover arbitration fees from the issuer. Five of these clauses (covering 4.3% of arbitration
subject credit card loans outstanding) also precluded cost-shifting to the consumer, while the
rest were silent on reallocation to the consumer. While clauses in all three categories would
allow the arbitrator to shift fees from the consumer to the company, only clauses in the second
category (and five of the clauses in the third category) offer the consumer any contractual
protection against the possibility of an adverse reallocation of costs at the award stage.
Checking account arbitration clauses contained similar provisions. Almost 25% of the clauses
(15 of 61, covering 16.9% of arbitration-subject insured deposits) expressly permitted the
arbitrator to shift arbitration costs to the consumer (and the converse as well), in line with the
default rules. Just over 11% of the clauses (7 of 61, covering 7.9% of arbitration-subject insured
deposits) precluded cost-shifting back to the consumer. Finally, 14.8% of clauses (9 of 61, but
covering 28.0% of arbitration-subject insured deposits) expressly permitted the consumer to
recover arbitration fees from the institution, but only one such clause (covering 0.4% of
arbitration-subject insured deposits) precluded the arbitrator from shifting costs to the
consumer.
The pattern also was similar for prepaid cards. Nine prepaid card arbitration clauses (17.6%; no
data on card loads) permitted fees to be shifted to consumers (with all but one permitting the
converse as well), while three clauses (5.9%; 26.5% of arbitration-subject prepaid card loads)

112

Some clauses clearly covered both the administrator’s fees and the arbitrator’s fees, but most were ambiguous
about whether both types of fees are covered or only the administrator’s fee.
113

Only one such clause (for a small issuer) requires that a losing consumer pay the issuer’s arbitration costs. The rest
permitted the arbitrator to so decide but do not require the arbitrator to do so.

47

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

precluded such fee shifting. Many more prepaid arbitration clauses (22 clauses, or 43.1%;
57.9%-67.7% of arbitration-subject prepaid card loads) affirmatively permitted prevailing
consumers to recover their arbitration fees, although without precluding cost-shifting back to
the consumer. These results are summarized in Table 7.
TABLE 7:

ARBITRATION CLAUSE PROVISIONS ADDRESSING ALLOCATION OF ARBITRATION FEES IN THE
AWARD 2012-13

# of contracts

% of arbitration-subject
account values

Permits shifting issuer fees to consumer
Credit cards

23
(34.8%)

21.8%

Checking accounts

15
(24.6%)

16.9%

Prepaid cards

9
(17.6%)

No data

Credit cards

5
(7.6%)

17.0%

Checking accounts

7
(11.5%)

7.9%

Prepaid cards

3
(5.9%)

26.5%

Bars shifting issuer fees to consumer

Permits shifting consumer fees to issuer
Credit cards114

18
(27.3%)

51.5%

Checking accounts115

9
(14.8%)

28.0%

Prepaid cards116

22
(43.1%)

57.9%-67.7%

114

Five of these credit card arbitration clauses, covering 4.3% of arbitration-subject credit card loans outstanding,
also precluded shifting the issuer’s arbitration fees to the consumer.
115

One of these checking account arbitration clauses, covering 0.4% of arbitration-subject insured deposits, also
precluded shifting the issuer’s arbitration fees to the consumer.

116

48

None of these prepaid card arbitration clauses precluded shifting the issuer’s arbitration fees to the consumer.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

PROVISIONS ADDRESSING THE AWARD OF ATTORNEYS’ FEES
A significant share of credit card arbitration clauses directed that the parties bear their own
attorneys’ fees unless the law requires otherwise (22 clauses, or 33.3%; 45.5% of credit card
loans outstanding). This was also true for a smaller share of checking account contracts (9
clauses, or 14.8%; but 37.6% of arbitration-subject insured deposits) and prepaid card contracts
(3 clauses, or 5.9%; no load data). One prepaid card (which covers 26.5% of arbitration-subject
card loads in our sample) waived any right of the company to recover attorneys’ fees from the
consumer.
Significant shares of arbitration clauses across all three markets, however, did not address
attorneys’ fees. This was true for 18 credit card clauses (27.3%, covering 21.4% of arbitrationsubject credit card loans outstanding), 22 checking account clauses (36.1%, covering 27.2% of
arbitration-subject insured deposits), and 35 prepaid clauses (68.6%, covering 73.3% of
arbitration-subject card loads). When the arbitration clause does not address the issue, the
arbitrator may award attorney’s fees when permitted by the agreement or applicable law.117
Five credit card arbitration clauses (7.6%, from small issuers representing a negligible market
share) directed or permitted the arbitrator to award attorneys’ fees to the prevailing party, which
presumably would have permitted the issuer to recover its attorneys’ fees from the consumer
when it prevailed, and also would have permitted a prevailing consumer to recover his or her
attorneys’ fees. Five prepaid arbitration clauses (9.8%; no load data) permitted an award to a
prevailing party, either the consumer or the company. Three checking account clauses (4.9%;
1.0% of arbitration-subject insured deposits) would have permitted an award to the prevailing
party, consumer or company.
A number of the remaining clauses permitted or directed the award of attorneys’ fees to a
prevailing consumer. Five credit card clauses (7.6%, covering 10.2% of arbitration-subject credit
card loans outstanding) directed the issuer to pay the consumer’s attorneys’ fees if the consumer
prevails. Other credit card clauses expressly authorized (but did not require) the arbitrator to
award attorneys’ fees to consumers, either if the consumer prevailed (1 clause, or 1.5%; 15.1% of
arbitration-subject credit card loans outstanding), if the amount awarded was greater than the

117

AAA, Commercial Arbitration Rules, Rule R-43(d)(ii) (effective June 1, 2009); JAMS Streamlined Arbitration
Rules, supra note 88, Rule 19(f).

49

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

issuer’s last settlement offer (1 clause, or 1.5%; 0.0% of arbitration-subject credit card loans
outstanding), or if the arbitrator so determined (1 clause, or 1.5%; 0.2% of arbitration-subject
credit card loans outstanding).
Eleven checking account clauses (18.0% of clauses; 18.1% of arbitration-subject insured
deposits) provided that the arbitrator would award, and another two clauses (3.3% of clauses;
2.4% of arbitration-subject insured deposits) might award attorneys’ fees to a prevailing
consumer. An additional three checking account clauses directed (4.9%; 2.5% of arbitrationsubject insured deposits), and another permitted (1.6%; 0.5% of arbitration-subject insured
deposits), the arbitrator to award the consumer attorneys’ fees if the award exceeded the
institution’s last written settlement offer, while another directed the award of double attorneys’
fees under those circumstances (1.6%; 0.2% of arbitration-subject insured deposits). One
prepaid clause (no load data) permitted the arbitrator to award attorneys’ fees to the consumer.
TABLE 8:

ARBITRATION CLAUSE PROVISIONS ADDRESSING THE AWARD OF ATTORNEYS’ FEES IN THE
AWARD 2012-13

# of contracts

% of arbitration-subject
account values

Parties bear own attorneys’ fees
Credit cards

22
(33.3%)

45.5%

Checking accounts

9
(14.8%)

37.6%

Prepaid cards

3
(5.9%)

No load data

Credit cards

5
(7.6%)

0.0%

Checking accounts

3
(4.9%)

1.0%

Prepaid cards

5
(9.8%)

No load data

Credit cards

8
(12.1%)

25.5%

Checking accounts

18
(29.5%)

23.7%

Prepaid cards

1
(2.0%)

No load data

Attorneys’ fees to prevailing party

Attorneys’ fees to prevailing consumer

50

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

No provision
Credit cards

18
(27.3%)

21.4%

Checking accounts

22
(36.1%)

27.2%

Prepaid cards

35
(68.6%)

73.3%

3.4.11 Contingent minimum award recoveries
The AT&T arbitration clause at issue in Concepcion provided that a customer would receive a
minimum recovery of $10,000 if the customer was awarded more in arbitration than the
amount of the last written settlement offer made by AT&T.118 Contingent minimum recovery
provisions were uncommon in the arbitration clauses that we studied, although they occurred
more often in checking account arbitration clauses than in credit card or prepaid card
arbitration clauses. We did not identify any such terms in contracts without arbitration clauses.
None of the arbitration clauses in the prepaid card contracts in the sample included a contingent
minimum recovery provision. Only three (4.5%) of the credit card arbitration clauses studied—
representing 2.3% of arbitration-subject credit card loans outstanding in the sample—included
such a provision, with the contingent amount ranging from $5,100 to $7,500. By comparison,
ten arbitration clauses (16.4%) in the checking account sample—representing 10.5% of
arbitration-subject insured deposits in the sample—included such a provision. For these ten
checking account contracts, the contingent minimum recoveries generally ranged from $2,500
to $10,000.119

118

See 131 S. Ct. at 1744 & n.3.

119

Most of the provisions that we identified make the minimum recovery contingent on the arbitrator awarding the
consumer the relief sought, or greater relief, after the institution refuses to provide such relief. A smaller share use a
different contingency: whether the arbitrator awards relief equal or in excess of the value of the company’s last
settlement offer.

51

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3.4.12 Disclosures
Most of the arbitration clauses described certain differences between arbitration and litigation
in court. They typically highlighted some combination of four differences. First, no jury trial is
available in arbitration. Second, discovery typically is more limited in arbitration than in civil
court litigation. Third, appeal rights are more limited in arbitration than for decisions of trial
court judges in civil court cases. Fourth, when parties have agreed to arbitrate, they cannot
participate in class actions in court.120 Often this descriptive language was capitalized or in
boldfaced type.
Of the credit card arbitration clauses studied, 49.3% (covering 40.8% of arbitration-subject
credit card loans outstanding) identified all four procedural differences; only a number of small
issuers identified none. Almost every credit card arbitration clause indicated that the consumer
would not have a right to a jury trial in arbitration (92.5% of clauses, covering 99.7% of
arbitration-subject credit card loans), and slightly more (94.0% of clauses, covering 99.9% of
arbitration-subject credit card loans) stated that for claims subject to arbitration the consumer
could not be party to a class action in court. Over half of the clauses mentioned more limited
rights to discovery and appeal.
The checking account arbitration clauses studied contained fewer disclosures. Only 27.9% of
clauses (covering 19.0% of arbitration-subject insured deposits) identified all four procedural
differences, while 24.6% of clauses (although again covering smaller issuers) identified none.
The most common difference disclosed was the lack of a jury trial (75.4% of clauses, covering
98.1% of arbitration-subject secured deposits). The majority (60.7%, covering 67.5% of
arbitration-subject insured deposits) stated that for claims subject to arbitration, the consumer
could not be part of a class action in court. Around 40% noted limited discovery and appeal
rights.
The disclosures in the prepaid card arbitration clauses were more like those in credit card
clauses, with 41.2% (covering from 27.0% to 36.8% of arbitration-subject card loads) disclosing

120

The type of disclosure provisions discussed here are not the same as the no-class-arbitration provisions examined
earlier. See section 3.4.5. The disclosure provision explains to the consumer that by agreeing to arbitration, the
consumer will not be able to participate in a class action in court. The no-class-arbitration provision provides that any
arbitration proceeding will be conducted on an individual basis and not a class basis. Most contracts included both,
but a few contracts with no-class-arbitration provisions did not make the type of disclosure considered here.

52

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

all four procedural differences and only 7.8% (with no card load data available) disclosing none.
Almost all (92.2% of clauses; 100% of arbitration-subject card loads) stated that no jury trial was
available, most (88.2% of clauses; 100% of arbitration-subject card loads) noted the inability to
participate in a class action in court, and many highlighted limited discovery (66.7% of clauses;
42.3% of arbitration-subject card loads) and limited appeal rights (47.1% of clauses; 57.9%67.7% of arbitration-subject card loads).

3.5

Clause changes

The data presented so far are for arbitration clauses as of the end of 2012 (for credit card
contracts) or as of the third quarter of 2013 (for checking account and GPR prepaid card
contracts). This section examines whether the inclusion of arbitration clauses in consumer
financial services contracts has changed since December 31, 2010, shortly before the Supreme
Court’s April 2011 decision in Concepcion.121 In Concepcion, the Court upheld a no-classarbitration provision in an arbitration clause against a state law unconscionability challenge,
overriding a number of state and federal court decisions upholding state laws that had
invalidated such provisions.122 Following Concepcion, several scholars and commenters
suggested that companies would inevitably include arbitration clauses with no-class-arbitration
provisions in all their consumer contracts123 or would revise their arbitration clauses in line with

121

See 131 S. Ct. at 1753.

122

See, e.g., Skirchak v. Dynamics Research Corp., 508 F.3d 49 (1st Cir. 2007); Shroyer v. New Cingular Wireless
Services, Inc., 498 F.3d 976 (9th Cir. 2007); Dale v. Comcast Corp., 498 F.3d 1216 (11th Cir. 2007); Muhammad v.
County Bank of Rehoboth Beach, Delaware, 912 A.2d 88 (N.J. 2006).
123

E.g., Ian Millhiser, Supreme Court Nukes Consumers’ Rights In Most Pro-Corporate Decision Since Citizens
United, ThinkProgress: Justice (Apr. 27, 2011) (“After Concepcion, it is only a matter of time before nearly every
credit card provider, cell phone company, mail-order business or even every potential employer requires anyone who
wants to do business with them to first give up their right to file a class action.”)
http://thinkprogress.org/justice/2011/04/27/176997/scotus-nukes-consumers. But see Peter B. Rutledge &
Christopher R. Drahozal, “Sticky” Arbitration Clauses?: The Use of Arbitration Clauses After Concepcion and Amex,
67 VAND. L. REV. ___ (forthcoming 2014) (finding that most franchisors have not switched to arbitration after
Concepcion and explaining that “[b]y using an arbitration clause, businesses do more than simply contract out of
class actions: they contract for a bundle of dispute resolution services, including, for example, a very limited right to
appeal. For businesses that perceive themselves as unlikely to be sued in a class action (and hence to receive little
benefit from an arbitral class waiver), the other services bundled with the waiver of class actions … may discourage

53

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

the case.124 We focus on changes in the incidence in contracts of arbitration clauses or clause
features; we do not address changes in the manner and frequency with which clauses are applied
that may result from intervening legal decisions such as Concepcion or Rent-A-Center.
Because of data availability, we focus here on changes in the inclusion of arbitration clauses in
credit card contracts. Only limited data on changes in checking account contracts since
Concepcion are available, but those data reveal a noticeable increase in the inclusion of
arbitration clauses among large banks since mid-2012. We have no data on changes in the
inclusion of arbitration clauses in prepaid card contracts since Concepcion.
We also note that many companies—particularly smaller entities—use standard forms acquired
from a form provider, rather than preparing their own customized form. At least 83 of the 141
small to mid-sized banks (58.9%) in the checking account sample used some version of a
standard form prepared by a single form provider. At present, that standard form does not
include an arbitration clause, although the form company does offer an optional free-standing
arbitration clause.

3.5.1

Credit cards

The incidence of arbitration clauses in credit card contracts has increased since Concepcion, but
only slightly. Examining the issuers in our sample that have agreements available for the entire
period from 2010-2012, the number of issuers using arbitration clauses increased from 57 as of
year-end 2010, to 58 as of year-end 2011, to 59 as of year-end 2012.125 A total of five credit card
issuers in our sample have adopted arbitration clauses since Concepcion, while three issuers
that previously used arbitration clauses stopped using them, for a net increase of two. The dollar
amount of credit card loans outstanding subject to arbitration clauses has increased somewhat
more―from 47.2% of credit card loans outstanding as of year-end 2010 to 50.1% as of year-end

them from using an arbitration clause. In addition, even standard form contracts might be sticky―i.e., resistant to
change even if change might be in the business’s best interest.”).
124

See Myriam Gilles, Killing Them with Kindness: Examining “Consumer-Friendly” Arbitration Clauses After
AT&T Mobility v. Concepcion, 88 NOTRE DAME L. REV. 825, 844-45 (2012).

125

A review of the credit card contracts available on the CFPB web page reveals no additional issuers switching to
arbitration between December 31, 2012, and June 30, 2013. We have not yet reviewed the clauses for changes in
features over this period.

54

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

2012.126 Of course, 86.8% of credit card loans outstanding without arbitration clauses were
subject to the Ross settlement during this period, which limited the extent of any shift toward
arbitration after Concepcion.127 Of the five issuers that have switched to arbitration since
Concepcion, all included no-class-arbitration provisions, four of five included delegation clauses
(the other excepted class issues from its delegation clause), and none had a minimum recovery
provision.
Of the issuers that have used arbitration clauses for the entire period from 2010-2012, only a
small number have changed provisions in their clauses since Concepcion. The only change in the
inclusion of no-class-arbitration terms by those issuers was the deletion by one issuer of
language that excluded California accounts from its class waiver. The inclusion of antidelegation clauses by those issuers increased from 33.6% of arbitration-subject credit card loans
outstanding as of December 31, 2010, to 42.6% as of December 31, 2012.128 Finally, in 2012, two
of these issuers added contingent minimum recovery provisions to their existing arbitration
clauses that were similar to those in the clause at issue in Concepcion. Even so, the market share
covered by such provisions—some 2.3% of arbitration-subject credit cards loans outstanding—
remains small.

126

The five issuers switching to arbitration contributed only a marginal amount to the increase, accounting for 0.2%
of credit card loans outstanding. Changes in the amount of credit card loans outstanding of existing users of
arbitration clauses explain almost all of the increase.

127

See supra text accompanying notes 51 to 52. It is possible, of course, that one or more of the settling defendants in
Ross might have removed their credit card arbitration clauses even absent Ross. The Ross allegations did not apply to
checking accounts, but Bank of America removed its arbitration clause from checking accounts in August 2009. See
Jonathan Stempel, Bank of America ends arbitration of card disputes, REUTERS.COM (Aug. 14, 2009),
http://www.reuters.com/article/2009/08/14/us-bankofamerica-arbitration-sb-idUSTRE57D03E20090814. On the
other hand, two issuer defendants did not agree to settle Ross, maintained their clauses, and together with another
issuer in a parallel proceeding, took the matter through trial. As this presentation was being finalized, the case
remained pending against all three issuers.
128

Two issuers replaced their delegation clauses with anti-delegation clauses, two replaced class exceptions with antidelegation clauses, one removed the class exception from its delegation clause, while one added a new delegation
clause, though it was still subject to a class exception.

55

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

3.5.2

Checking accounts

By comparison to credit card contracts, only limited data are available on changes in checking
account contracts since Concepcion. We compare the contracts used in the Pew Charitable
Trusts study of checking account contracts―collected from June to August 2012―to the
contracts we collected just over a year later ― from August to September 2013.129 Thus, the data
cover only a portion of the period since Concepcion. Any changes to checking account contracts
prior to summer 2012 will not be reflected in the findings described below.130
The sample consists of 88 large financial institutions (banks and credit unions) as to which the
sample in the Pew study overlaps with the sample used by the Bureau.131 As of summer 2012, 35
of the 88 institutions (39.8%) used arbitration clauses in their checking account contracts. As of
summer 2013, one year later, 42 of 88 (47.7%) used arbitration clauses. Eight banks and one
credit union switched to arbitration during that one-year period, while two banks switched away
from arbitration. The eight banks switching to arbitration constituted 9.2% of all arbitrationsubject insured deposits in the large bank sample. The credit union became the largest credit
union using an arbitration clause, accounting for over one-third of all arbitration-subject
insured deposits at credit unions in the sample as of year-end 2012. Of the institutions that
switched to arbitration within the past year, six included delegation clauses, two had antidelegation provisions, and one had neither an anti-delegation nor delegation provision; all but
one had a no-class-arbitration provision; and none had a contingent minimum recovery
provision.
As with credit card contracts, only a small number of institutions have changed the terms of
their existing arbitration clause within the past year. Of the institutions that had arbitration
clauses in both 2012 and 2013, one adopted a delegation clause, one adopted an anti-delegation
clause, and a third added language that delegates most issues of enforceability to the arbitrators;

129

Pew Charitable Trusts, supra note 37, at 14. We describe our sourcing of checking account contracts in Appendix A.

130

For example, press reports indicate that Umpqua Bank adopted an arbitration clause in January 2012, prior to the
period for which we have data here. See Brent Hunsberger, Umpqua Bank Joins Wells Fargo and Chase in Requiring
Customers to Arbitrate Disputes, Barring Class Actions, THE OREGONIAN (Jan. 23, 2012),
http://www.oregonlive.com/business/index.ssf/2012/01/umpqua_bank_joins_wells_fargo_1.html.
131

Pew was unable to obtain agreements for eight banks in its sample, while four of the banks in the Pew sample
(based on total deposits) are not in the sample used here (based on total insured deposits).

56

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

its clause previously stated only that class issues were to be decided by a court and was silent on
the other issues. Two of those institutions adopted no-class-arbitration provisions, and one
added a contingent minimum recovery provision to its arbitration clause.

57

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4. Incidence and typology of
consumer arbitration filings
This section presents preliminary results from our review of 1,241 credit card, checking and
payday loan consumer disputes filed with the American Arbitration Association (“AAA”) from
the start of 2010 through the end of 2012.132 Based on our review of AAA consumer arbitration
files, these were the only consumer disputes about any of these three products filed with the
AAA during that period.

4.1

Data sources

4.1.1

AAA case data

Our review used electronic case records that the AAA voluntarily provided to the Bureau (the
“AAA Data”). The AAA began using its electronic case record system from the beginning of 2010
for all filings administered as consumer arbitrations. Under the AAA rules—in this case the AAA
Supplementary Procedures for Resolution of Consumer-Related Disputes—a dispute between a
consumer and a company is administered as a consumer arbitration if it meets a number of
criteria. First, the business must have a “standardized, systematic application of arbitration
clauses with customers and where the terms and conditions of the purchase of standardized,
consumable goods or services are non-negotiable or primarily non-negotiable in most or all of
its terms, conditions, features, or choices.” Second, the product at issue must be for personal or

132

We identified four filings relating to prepaid or gift cards. They are not included in the 1,241 because we
considered that number of cases too small to include as a separate product market in this presentation.

58

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

household use.”133 (If there is no applicable arbitration clause in the contract between the
consumer and company, the consumer arbitration rules may still apply if the parties agree to
submit the dispute to AAA.134) As a result, an AAA “consumer arbitration” is not necessarily one
filed by a consumer. So long as the criteria described above apply, companies can file consumer
arbitrations against consumers.
The AAA agreed to provide to the Bureau all case files from its electronic case record system for
consumer arbitrations filed from January 1, 2010 through the end of 2012. The material
volunteered by the AAA covers class arbitrations and non-class arbitrations; filings that went no
further than an initial filing, as well as those that did advance to formal initiation by the AAA
and beyond; and all arbitration filings that were submitted as consumer arbitrations regardless
of subject matter.
To date, our review of the AAA Data has: (1) identified consumer arbitration filings that cover
credit card, checking account, payday loan, or prepaid disputes; and (2) extracted data points for
the first three of these categories. This presentation, therefore, does not cover data about AAA
consumer arbitrations outside of these particular product markets. We refer to the AAA Data
that covers these three product markets for filings between January 1, 2010 and December 31,
2012, as the “AAA Case Data.”
Our review to this point was limited to AAA files. This is because we have not secured data from
JAMS and, indeed, JAMS may not have data of the same duration as the AAA. Although we
intend to continue to explore with JAMS the possibility of obtaining its data, we do not view the
absence of such data as materially impacting our analysis because, during the 2010-12 period
under review, the AAA was and remains the largest administrator of consumer arbitration in the
United States.135 JAMS, the other leading consumer arbitration administrator, has reported that

133

AAA, Consumer-Related Disputes Supplementary Procedures C-1(a) (effective Sept. 15, 2005). The AAA reserves
discretion whether or not to apply the consumer arbitration rules, with the parties being free to raise disputes about
their application directly to the arbitrator. See id.
134

See id. Rule C-3 (effective Sept. 15, 2005).

135

That was not true until July 2009, when the National Arbitration Forum (“NAF”) settled a consumer fraud suit
brought by Minnesota Attorney General Lori Swanson. Until that point, NAF was almost certainly administered the
largest caseload of consumer arbitrations in the U.S. In the settlement, however, it agreed to stop administering new
consumer arbitrations and exited the consumer arbitration business altogether.

59

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

it handles “at most” a few hundred consumer cases every year.136 In contrast, the AAA Data
contain approximately 7,000 consumer arbitration filings over three years. (This number is not
limited to consumer financial matters.137) Although many of these cases were outside the scope
of our review, these numbers provide some indication of the relative roles of the AAA and JAMS.
Moreover, during the time period under study, the AAA was the largest provider of consumer
arbitration services with respect to consumer financial disputes within the scope of our current
review. As discussed in section 3.4.3 above, almost all the credit card arbitration clauses studied
provide for arbitration before the AAA alone (48.5%) or before a choice of the AAA and other
administrators (a further 34.8%). For checking accounts, the picture is the same. Banks with an
arbitration clause either specify (91.8%) or require (55.7%) the AAA to be used if arbitration
were elected by either party. While JAMS is a permitted option for 40.1% of credit card
agreements and for 34.4% of checking account agreements, it is only once the sole option for
checking account agreements and only three times the sole option for credit card agreements.138

4.1.2

Existing empirical studies and data

Although a relatively large number of empirical studies have examined employment and
securities arbitration, relatively few such studies have examined consumer arbitration in
detail.139 Drahozal and Zyontz reviewed AAA consumer arbitrations that resulted in an

136

See Jean Sternlight, Mandatory Binding Arbitration Clauses Prevent Consumers from Presenting Procedurally
Difficult Claims, 42 SW. U. L. REV. 87, 99 n.68 (2012) (“Mandatory Binding Arbitration Clauses”) (citing Jay Welsh,
JAMS Executive Vice President and General Counsel.) There is general agreement that the AAA and JAMS are the
leading consumer arbitration administrators. See, e.g., ABA, CBA, & FSR RFI Comment at 3.

137

It covers all consumer arbitration filings, regardless of subject matter. It also includes filings that were not
perfected for one or more of the reasons discussed at section 4.2.1 below. Approximately one-third of the filings were
not perfected.
138

By market volume subject to arbitration, the AAA is specified still more frequently. See section 3.4.3.

139

Part of the reason may be that “all arbitration awards in the securities industry are publicly available.” David B.
Lipsky et al., The Effect of Gender on Awards in Employment Arbitration Cases: The Experience in the Securities
Industry, 52 INDUSTRIAL RELATIONS: A JOURNAL OF ECONOMY AND SOCIETY 314, 317 (2013). Securities industry
arbitration encompasses claims involving customers and brokers as well as employment claims involving employees
and brokerage firms. These arbitrations are administered by the Financial Industry Regulatory Authority (“FINRA”).
See id. at 322.

60

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

arbitrator’s “award” during nine months of 2007.140 The AAA Case Data available to us is
significantly broader than the data covered by this earlier study. It covers a much longer period,
and it is not limited to case files for awarded cases. A number of summary reviews published by
the AAA covered all administered AAA consumer arbitrations from 2006 and a sample from
2007.141
Empirical review of non-AAA consumer arbitrations has generally used NAF data. A number of
studies have used public data in California for NAF consumer arbitrations from 2003 through
early 2007.142 Public Citizen identified a set of roughly 34,000 NAF arbitrations for this period,
all but 118 of which companies filed against consumers, rather than the other way around.143
Almost all the cases were described by NAF as “collection” cases. Ernst & Young carried out a
review of a sample of NAF consumer-filed arbitrations from 2000 through 2003.144 NAF also

140

See Christopher R. Drahozal & Samantha Zyontz, An Empirical Study of AAA Consumer Arbitrations, 25:4 OHIO
ST. J. ON DISP. RES. 843, 845 (2010) (“Empirical Study”) (noting relative lack of studies of consumer arbitration). This
study presented results primarily from analyzing 301 consumer arbitrations closed by an award between April 2007
and December 2007. See id. at 867-71 (describing data and methodology). This study was part of a broader research
project supported by the Searle Civil Justice Institute, which was then associated with Northwestern University
School of Law. Drahozal and Zyontz used the same AAA data for a follow-on study that compared debt collection
claims by companies in AAA consumer arbitrations with debt collection claims in federal court and in state court
proceedings in certain Virginia and Oklahoma jurisdictions. See Christopher R. Drahozal & Samantha Zyontz,
Creditor Claims in Arbitration and in Court, 7 HASTINGS Bus. L. J. 77 (2011) (“Creditor Claims”).
141

See AAA, Statement of the AAA Before the Constitution Subcomm. of the Senate Comm. on the Judiciary (2007)
and Analysis of the AAA’s Consumer Arbitration Caseload (2007), both available at the AAA’s website, www.adr.org.

142

Pursuant to section 1281.96 of the California Code of Civil Procedure law, arbitration providers are required to
disclose a number of data points about consumer and employment arbitrations in California. See Cal. Civ. Proc. Code
§ 1281.96. A number of other jurisdictions now have similar requirements. The AAA does not limit its public
disclosures to California arbitrations, but discloses information on consumer arbitrations throughout the United
States. Because we have access to AAA electronic case records, however, we have not relied for our review on these
public data. In addition, as discussed in section 4.2 below, our review covers all consumer arbitration filings with the
AAA, whereas the publicly available data only reaches disputes that the AAA actually processes.
143

Public Citizen, The Arbitration Trap: How Credit Card Companies Ensnare Consumers (Sept. 2007),
http://www.citizen.org/documents/ArbitrationTrap.pdf.
144

Ernst & Young, Outcomes of Arbitration: An Empirical Study of Consumer Lending Cases (2004),
http://www.adrforum.com/rcontrol/documents/ResearchStudiesAndStatistics/2005ErnstAndYoung.pdf. This study
also briefly reviewed certain public data on AAA arbitrations available in California.

61

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

published at least one review of its own case data.145 Finally, data about consumer arbitration
can sometimes be obtained from public litigation proceedings. In a number of court cases,
companies or other parties have proffered evidence about arbitration.

4.2

Case incidence

4.2.1

Overall

For 2010 through 2012, excluding class cases, we identified 1,241 AAA non-class consumer
arbitration filings about credit cards, checking account, and payday loans.146 The cases covered
the entire range of arbitral procedure, from cases filed with the AAA but never processed (for
one of the reasons noted below) to cases decided by the arbitrator in a written award.147
There are several reasons for an AAA consumer arbitration to proceed no further than filing.
First, there may be no pre-dispute arbitration clause between the parties, and one or other party
may not agree to arbitrate post-dispute. Second, the parties may settle the dispute before the
matter proceeds further. Third, the filing party may abandon the claim before the matter
proceeds further. Fourth, the AAA may refuse to accept the dispute for one of a number of
reasons: if the case is inconsistent with the terms of its moratorium on certain debt collection
disputes, which we describe more fully below; if the company fails to comply with the AAA’s Due

145

Mark Fellows, The Same Result as in Court, More Efficiently: Comparing Arbitration and Court Litigation
Outcomes, METRO. CORP. COUNSEL (July 2006) (covering NAF cases from 2003 and 2004),
http://www.metrocorpcounsel.com/articles/6988/same-result-court-more-efficiently-comparing-arbitration-andcourt-litigation-outcomes.
146

In this time period, there were only two class arbitration filings before the AAA for these three product markets.
(Claimants in these two cases sought to represent “thousands” of consumers.) One reason for the lack of class
arbitration filings may be that the applicable credit card and checking account arbitration clauses almost universally
prohibit class arbitration, as described in section 3.4.5. (That may also be the case for payday lenders, but we do not
yet have significant data on the point. The Bureau is not aware, however, of any arbitration clause in the consumer
financial area that expressly permits class arbitration.) This presentation only reviews non-class arbitration filings.
147

While our focus in this presentation is on the number and type of consumer disputes filed with the AAA, we will
necessarily be looking at a smaller set of disputes when we later examine awards in arbitrated disputes. We will also
look to see whether particular types of cases tend to end at specific stages, and comparing that to disposition in the
court system.

62

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Protocol by failing to pay the required filling and administrative fees; if the arbitration clause
violates the terms of the Due Process Protocol, and the company is not willing to waive the
violative terms for purposes of the arbitration; if the company has previously refused to comply
with the Due Process Protocol; or for a claimant’s failure to meet other filing requirements.
The Bureau recognizes that the number of filed arbitrations is not a metric that can be looked at
in isolation. We do not know the numbers of credit cardholders, checking account holders, or
payday borrowers with potentially viable legal claims. With the exceptions noted below, we do
not yet know the number of claims filed in court proceedings.148 Plainly, the number of
arbitrations was low relative to the total populations using these products. Using the Bureau’s
Consumer Credit Panel, which provides a nationally representative, random sample of deidentified credit records procured from a large, national consumer reporting agency, we
estimate that, as of the end of 2012, there were approximately 160 million credit cardholders in
the United States. On the assumption that the number of cardholders and the volume of credit
card loans outstanding are proportionate, when combined with our incidence data from section
3.2.1, this data indicate that around 80 million cardholders were subject to arbitration clauses as
of the end of 2012. Data from the Census and from the Survey of Consumer Finance collectively
suggest that in 2010 about 108 million U.S. households held one or more transaction accounts, a
category that includes consumer checking accounts.149 Combining data from the Census and the
FDIC’s National Survey of Unbanked and Underbanked Households suggests that in 2011,
around 105.1 million households had at least one checking or savings account.150 Again on the
assumption that the number of checking accounts is proportionate to the volume of insured

148

Of course, the litigation numbers may be impacted up or down by the presence of arbitration clauses, making a
direct comparison hard. In addition, there are limits on the availability of state court case records.
149

According to the Survey of Consumer Finances, 92.5% of families had at least one transaction account in 2010. See
Jesse Bricker, Arthur B. Kennickell, Kevin B. Moore, and John Sabelhaus, Changes in U.S. Family Finances from
2007 to 2010: Evidence from the Survey of Consumer Finances, Federal Reserve Bulletin, vol. 98, no 2 (Feb. 2012),
Table 6B at page 28. In this study, “family” is comparable to the Census definition of “household.” Id. at 3. According
to the Census, there were 116,716,331 U.S. households in 2010, which means that approximately 108 million
households had at least one transaction account in that year. See U.S. Census Bureau; 2010 Demographic Profile Data,
Profile of General Population and Housing Characteristics: 2010, Table DP-1.
150

In 2011, according to the FDIC’s survey, there were 9.9 million households without either a checking or a savings
account. See Federal Deposit Insurance Corporation, 2011 FDIC National Survey of Unbanked and Underbanked
Households at 510 (2012). The Census identifies 115 million households in that year. See U.S. Census Bureau; 2011
American Community Survey 1 Year Estimates, Population and Housing Narrative Profile: 2011, Table NP01.

63

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

deposits, it appears that tens of millions of households are subject to arbitration on one or more
checking accounts.151 Finally, the FDIC also estimates that around 2 million households use
payday loans annually.152

FIGURE 4:

NUMBER OF AAA CONSUMER ARBITRATIONS BY PRODUCT 2010-12

Figure 4 summarizes the breakdown of the 1,241 filings by product market: 1,033 were credit
card disputes; 71 were checking account disputes; and 137 were payday loan disputes.153 The

151

We do not offer a more specific estimate for checking accounts because: (a) these account data cover savings and
other transaction accounts; (b) our overall incidence estimate from section 3.2.2 is specific to bank use of arbitration
clauses for checking accounts, and thus does not reflect the incidence of credit union use; and (c) the section 3.2.2
incidence estimate uses 2012 and 2013 data, whereas the account data offered here is for 2010 and 2011.

152

Federal Deposit Insurance Corporation, Addendum to the 2011 FDIC National Survey of Unbanked and
Underbanked Households Use of Alternative Financial Services (2013) 48. The Pew Charitable Trusts recently
estimated that around 12 million individuals use payday loans every year. See Pew Charitable Trusts, Who Borrows,
Where They Borrow, and Why: Payday Lending in America (2012) 4.
153

For purposes of this analysis, we have included in the category of “payday loan” cases a number of arbitration
filings involving credit service organizations (“CSOs”). Approximately 23 states allow credit service organizations
(CSOs) to broker loans. See http://www.responsiblelending.org/payday-lending/policy-legislation/states/CRL-CSOIssue-Brief-FINAL.pdf. In some of these states, CSOs can charge unlimited fees to broker loans. See
http://forabettertexas.org/images/EO_2013_09_StateofPayday.pdf. Although CSO statutes generally contemplate
that a CSO will broker a loan with a third party lender, there are claims that payday lenders pose as CSOs to evade
state interest rate caps and make a larger profit on each loan. See http://www.responsiblelending.org/paydaylending/policy-legislation/states/CRL-CSO-Issue-Brief-FINAL.pdf. To account for this possibility, arbitration

64

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

average annual number of disputes, therefore, was 344 for credit card, 24 for checking account,
and 46 for payday loans.

4.2.2

Types of consumer arbitration filings

Prior analyses of arbitration cases have distinguished company-filed and consumer-filed cases.
Company-filed cases were generally debt collections arbitrations in which the company sought
to collect debt or alleged debt from consumers. At least for some creditors for some period of
time, debt collection arbitration took the place of debt collection litigation. The available data
indicates that company-filed debt collection actions were the predominant form of consumer
arbitration cases prior to 2010, with a number of specific credit card issuers making significant
use of debt collection arbitration before NAF.154
Our analysis of the AAA records persuades us that it is important to identify debt collection
arbitrations, but that the category of debt collection arbitrations now encompasses more than
company-initiated arbitrations.155 In 2009, prior to the period for which we have studied AAA
case records, the AAA adopted a moratorium on the filing of debt collection actions. The
moratorium continued throughout the study period, and continues to this day.156 It generally
bars consumer financial companies from filing arbitrations with the AAA to collect debt from
demands filed against CSOs were included in the study when consumers alleged that the CSO originated any loan
except for an auto-title loan.
154

NAF reported that in 2006 alone it administered 214,000 debt collection cases. See Drahozal & Zyontz, Creditor
Claims, supra note 140, at 77 (noting that it used to be generally accepted that “the vast majority of disputes between
businesses and consumers, both in arbitration and in litigation, involve claims by creditors seeking to recover
amounts allegedly owed by the consumers” (emphasis added)). In July 2009, however, as a result of its settlement
with the Minnesota Attorney General, NAF agreed permanently to cease administering most consumer arbitrations,
including debt collection arbitrations. The Attorney General based the suit, in part, on allegations that NAF shared
common ownership with a number of firms that filed debt collection claims before NAF. The demise of NAF has been
extensively covered. See, e.g., Nancy A. Welsh, What is “(Im)partial Enough” in a World of Embedded Neutrals?, 52
ARIZ. L. REV. 391, 427-31 (2010).
155

Our review procedures for AAA case records are discussed in detail in Appendix B.

156

The AAA announced its moratorium on debt collection arbitration on July 23, 2009. See AAA Press Release,
www.adr.org. Its terms are also stated by the AAA in a Notice on Consumer Debt Collection Arbitrations (Oct. 19,
2010), available on the same website. The AAA put the moratorium in place after it identified, by its own account, a
number of potential due process and fairness concerns with debt collection arbitration. See Arbitration or Arbitrary:
The Misuse of Mandatory Arbitration to Collect Consumer Debts: Hearing Before the Subcomm. on Domestic Policy
of the H. Comm. on Oversight, 111th Cong. (July 22, 2009) (testimony of Richard W. Naimark on behalf of the
American Arbitration Association).

65

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

consumers, but it allows for two exceptions to this general rule. First, such companies may file
such debt collections if a court orders the case to arbitration. Second, companies can file debt
collection arbitrations if the consumer affirmatively consents in writing to the arbitration after
the dispute arises, in which case the arbitration is a “mutual submission.” Furthermore, the
moratorium does not apply at all to arbitrations filed by the consumer alone.157
Following the moratorium, there are still company-filed AAA arbitrations that seek to collect
debt. As noted, these must now follow a court proceeding sufficient to generate a court order in
favor of arbitration. We included these cases in our debt collection arbitration category. But in
addition to company-filed debt collection cases there also can be mutual submissions or
consumer filings.158 To capture all of these categories, we defined “debt collection arbitrations”
to mean not only collection actions filed by the company, but also those arbitrations filed
mutually or by the consumer in which the parties are contesting an alleged debt claimed by the
company in a preceding action in court.
More specifically, we counted mutual submissions and consumer-filed disputes as debt
collection arbitrations when the case included a substantive debt dispute and the arbitration
record shows that there was a prior court proceeding as to which the consumer invoked
arbitration.159 We recognize that an invocation of arbitration can take several forms. In some
cases, the record may show that the consumer moved to compel arbitration in a prior court
collection action. In others, the invocation may consist of filing a demand for arbitration in lieu
of filing an answer in court to the company’s collection litigation in court. In others, the
consumer may simply inform the company that he or she is invoking the arbitration clause in
order to end the collection action in court. Although in some of these cases the consumer may
also raise non-debt, affirmative claims, we think it is appropriate to describe such disputes as

157

We have not reviewed the AAA filings for compliance with the AAA moratorium. Based on our review to date,
however, we have no reason to believe the AAA is not following its stated terms.
158

For purposes of this presentation, we categorize a claim as consumer-filed, company-filed, or mutually submitted
based on what the AAA claim form records. Further details are in Appendix B.
159

We do not count a procedural challenge to debt collection, without more, as a substantive debt dispute. So a claim
under the Fair Debt Collection Practices Act that covers only the manner in which debt is being collected would not be
considered a substantive debt dispute. We use that term for disputes that cover claims about what amount of money
is owed (or not owed) to the company by the consumer.

66

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

debt collection arbitrations.160 Collection claims were previously at issue in court, and the
amount of debt subject to collection is at issue in the arbitration.
We recognize that this approach works an expansion of the traditional understanding of debt
collections as company-filed disputes.161 There were a number of pleading formats in consumerfiled and mutually submitted AAA arbitrations, however, that persuaded us that the debt
collection category needs to be broadened in this way. In some cases, for example, the consumer
may affirmatively state that he or she has no claims but wants the arbitrator to resolve the
merits of the company’s underlying debt collection claim. In others, the consumer states that he
or she is filing the arbitration demand instead of filing an answer to a collection claim in court,
or the consumer may file an arbitration for a declaratory judgment that he or she does not owe
the amount claimed by the company. At least where there is evidence of prior proceedings in
court, we think it is appropriate to recognize these as debt collection arbitrations, even though in
some cases they may contain other consumer claims as well.162
In some respects, however, we may have undercounted debt collection arbitrations in our total
pool of cases. Our definition relies on an indication in the arbitration record of prior court
proceedings. The arbitration record may not contain that indication, even when there was, in
fact, a prior collection action in court. In addition, even when a company has not yet sued in
court to collect debt, it is possible that some consumers are preemptively filing arbitrations to

160

We provide information about the affirmative claims made by consumers in debt collection arbitrations in section
4.6.2 below.
161

There may be a precedent in some of the NAF data. In 2007, Public Citizen published results from its review of the
nearly 34,000 consumer arbitrations that NAF disclosed as occurring in California between the beginning of 2003
and March 2007. Consumers filed 118 of these arbitrations. All but 15 of the 118 consumer-filed cases, however, were
described by NAF as “collection” cases. See Public Citizen, supra note 143, at 1-2. Drahozal & Zyontz did not identify
any consumer-filed cases as debt collection arbitrations in their review of 2007 AAA consumer case records. It is
possible, however, that their study would not have identified such cases because their sample of case records was
limited to awarded cases. To the extent that consumer-filed debt collection cases tend to settle out or are withdrawn
or otherwise closed pre-award, they will not show up in awarded cases. Alternatively, these cases may have become
more common since 2007.
162

Overall, most of the consumer- and mutually-filed debt collection submissions encompass only a substantive
dispute about the amount of debt owed. Fewer than half include non-debt claims as well. See section 4.6.1.

67

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

challenge the company’s pre-judicial assertion of a debt.163 In fact, a substantial number of the
“non-debt collection” credit card arbitrations in our review appeared to involve only a
substantive debt dispute and no non-debt claims at all, even though the arbitration may be filed
by a consumer. Ultimately, though, we opted to use objective rules to define debt collection
arbitrations, rather than trying to assess whether the weight of the arbitration record indicated
that collection activity was already underway before a substantive debt dispute reached
arbitration. But it is important to bear in mind that our “non-debt collection category” included
a large number of cases in which debt was at issue.164

FIGURE 5:

ARBITRATION FILINGS BY TYPE AND PRODUCT

2010

2011

2012

163

Cf. Sternlight, Mandatory Binding Arbitration Clauses, supra note 136, at 99 & n.69 (noting statement by JAMS
General Counsel that most consumer arbitrations filed before JAMS “are from claims alleging that certain credit card
companies violated federal collection statutes, asserted to fend off potential collections actions by those companies.”).
164

We considered relying only on the subject matter of the claims at issue in arbitration to differentiate “debt
arbitrations” from “non-debt arbitrations.” Our subject-matter typology for disputes, which we discuss in sections
4.5.2 and 4.6.2, identifies when arbitrations include substantive claims about the amount of debt owed. We
determined, however, not to rely on that debt dispute tag alone.

68

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Figure 5 shows the breakdown between debt collection arbitrations and non-debt collection
arbitrations by product market for the three years covered by the AAA Case Data. Overall, there
were 522 debt collection arbitrations and 719 non-debt collection arbitrations for all three years
combined for our product markets. For checking account and payday loan arbitrations, the
norm was non-collection. For credit card, by contrast, debt collection arbitrations were a slight
majority of the disputes. All but four debt collection arbitrations were credit card disputes.
At least in the period of our review, parties initiated AAA consumer arbitrations by filing a
standard one-page “claim form.” The claim form asks whether the arbitration is submitted at the
request of the consumer, at the request of the company, or by “mutual agreement,” meaning that
the arbitration is submitted by both parties.165 We recorded the identity of the filing party based
on the indicated response to this question.166 Claim forms recorded 146 (or 28%) of the debt
collection arbitrations in the AAA Case Data as company-filed. Another 47 (or 9%) of debt
collection arbitrations were mutual submissions. Finally, 329 (or 63%) were consumer
submissions. On the non-collection side, 569 (or 79%) of the filings were by consumers. There
were smaller shares of mutual submissions (132 filings or 18%) and company-filed (18 filings or
3%) non-collection claims.
We discuss debt collection arbitrations and non-collection arbitrations in more detail in sections
4.5 and 4.6.

4.2.3

Credit card federal court filings

Although we have not completed our litigation review of court filings raising disputes about
these same three products, we have obtained some results. For comparative purposes, we
provide those results here. Our methodology to identify these cases is provided in Appendix C.
For the 2010 through 2012 period, we have preliminarily identified 3,054 credit card cases filed
in federal court. This breaks down into 1,033 cases in 2010, 883 cases in 2011; and 1,138 cases in

165

The claim form first asks: “How is this claim being filed?” Parties are instructed to check only one option. In
addition to “By request of the consumer,” and “By request of the company,” the form offers “By mutual agreement
(“submission”).”
166

We have not attempted to verify whether that recording was accurate. At least in some cases, we have noted that
there is no apparent signature from the identified claimant in the record we have reviewed.

69

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

2012. All but four cases were filed by consumers, which means that consumers filed more than
four times as many federal court credit card disputes as AAA credit card arbitrations over this
time period.167 There were jury demands in 2,739 of these federal filings. Some 418 of the cases
were filed as purported class actions. Of these, 156 were filed in 2010, 121 were filed in 2011, and
141 were filed in 2012.

4.3

Representation

4.3.1

Prior research

Prior research has studied representation rates, and their potential impact, in AAA consumer
arbitrations. In their study of 301 AAA consumer arbitrations that resulted in an award in 2007,
Drahozal & Zyontz found that consumers proceeded without an attorney in almost half the
cases.168 Consumers were significantly more likely to have counsel when they were the claimant
in arbitration (55.4%) than when they were responding to company claims in arbitration
(29.5%). Some studies also report high rates of employee representation by counsel in AAA
employment arbitration.169

4.3.2

Data

The following four figures show representation rates in the AAA Case Data, for consumers and
companies, for all three product markets, and by debt collection and non-collection arbitrations.
These data may understate company representation rates. It was straightforward from the AAA
Case Data to identify the presence of outside counsel acting for the consumer or the company. It

167

Over three years, some 705 of the AAA filings were credit card disputes that the claim form recorded as being filed
by consumers. None of the class arbitration filings were credit card disputes.
168

Drahozal & Zyontz, Empirical Study, supra note 140, at 903.

169

See Alexander Colvin, An Empirical Study of Employment Arbitration: Case Outcomes and Processes, 8:1 J. EMP.
LEG. STUD. 1 (March 2011) at 16-17; see also Elizabeth Hill, Due Process at Low Cost: An Empirical Study of
Employment Arbitration Under the Auspices of the American Arbitration Association, 18 OHIO ST. J. ON DISP. RES.
777, 800 n.93 (2003) (for a sample of awarded AAA employment arbitrations, employees subject to a mandatory
arbitration clause proceeded pro se in one-third of the cases).

70

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

was hard to tell, however, whether an internal company representative was a lawyer. Except
where we could reliably identify outside or in-house counsel, therefore, we assumed that
companies were unrepresented.

FIGURE 6:

71

OVERALL REPRESENTATION RATES BY ARBITRATION TYPE 2010-12

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 7:

CREDIT CARD ARBITRATION REPRESENTATION RATES 2010-12

600

Arbitration filings

500
400
300
200
100
0

No lawyer
Lawyer

FIGURE 8:

Consumers in
credit cardrelated debt
collection filings
301
217

Companies in
credit cardrelated debt
collection filings
4
514

Consumers in
credit cardrelated non-debt
collection filings
247
268

Companies in
credit cardrelated non-debt
collection filings
58
457

CHECKING ACCOUNT ARBITRATION REPRESENTATION RATES 2010-12

80
70

Arbitration filings

60
50
40
30
20
10
0

No lawyer
Lawyer

72

Consumers in
checking-related
debt collection
filings
1
1

Companies in
checking-related
debt collection
filings
0
2

Consumers in
checking-related
non-debt
collection filings
31
38

Companies in
checking-related
non-debt
collection filings
14
55

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 9:

PAYDAY ARBITRATION REPRESENTATION RATES 2010-12

The AAA Case Data show that companies rarely participated in consumer arbitration about
these three products without counsel. It should be noted, however, that at least some
jurisdictions have unauthorized practice rules that may require companies to be represented in
arbitration.170 Whatever the reason, companies were overwhelmingly represented by counsel,
both in debt collection and non-debt collection arbitrations, and across all three product
markets we examined.
Furthermore, the data are consistent with the notion that a significant share of the consumers
who are in arbitration see it as being in their interest to have counsel as well. Overall, nearly 53%
of consumers in these arbitrations had counsel. In non-collection cases, consumers were
generally more likely (61%) to be represented by counsel than to proceed without counsel

170

73

See, e.g., Nisha, LLC v. TriBuilt Const. Group, LLC, 388 S.W.3d 444 (Ark. 2012).

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

(39%).171 (The rate of representation, moreover, did not vary to any extent between noncollection cases that included debt disputes and those that did not.) In debt collection
arbitrations, consumer representation was somewhat lower at 42% of the disputes filed.

4.3.3

Repeat counsel

Another feature of representation in these consumer arbitrations was the presence of “repeat
counsel,” meaning counsel who appeared in more than one case in our dataset. There were
several firms that showed up repeatedly across the cases in which there was representation.
This phenomenon was similar for consumer representation in debt collection and non-collection
cases. Of consumers represented in debt collection arbitrations, 71% were represented by repeat
counsel, and 59% were represented by one of only five repeat firms.172 Of consumers represented
in non-collection disputes, 77% of represented consumers had a repeat counsel. In this case,
however, one firm accounted for 29% of the non-collection cases in which consumers had repeat
counsel. Overall representation data, including repeat representation data, are depicted in
Figures 10 and 11. (The shares depicted in Figures 10 and 11 are for consumers overall. In
contrast, the numbers above use represented consumers as the denominator.)

171

Absent one counsel, who represented consumers in 18 effectively identical payday arbitrations in 2010 alone, this
number would fall. These cases were all follow-on filings from a single court case in which the same counsel
represented a total of 19 consumers.
172

Scholars who have attempted to study whether there is a “repeat player” impact on arbitration outcomes debate
whether or not to count the first instance of a “play” by a counsel or party that then becomes a repeat player. See, e.g.,
Colvin, supra note 169, at 13. For present purposes, we have counted the first play visible to us in the AAA Case Data.

74

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 10:

CONSUMER REPRESENTATION IN DEBT COLLECTION ARBITRATIONS 2010-12

FIGURE 11:

CONSUMER REPRESENTATION IN NON-COLLECTION ARBITRATIONS 2010-12

Repeat counsel players also featured on the company side of these disputes, as might be
expected given that companies were themselves repeat players. Repeat outside counsel figures
in 78% of non-collection cases in which companies were represented. Five repeat players
accounted for 22% of the non-collection cases in which companies were represented. Repeat
outside counsel were more prevalent in debt collection cases. Some 90% of the debt collection
cases in which companies were represented featured repeat counsel. Five repeat outside counsel
accounted for 45% of the debt collection cases in which companies were represented.

75

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4.4

Amounts at issue

4.4.1

Prior research

Some earlier empirical work makes claims for the effectiveness of consumer arbitration in
handling individual small-dollar disputes. Ernst & Young reported in a 2004 study of consumer
arbitration that 73% of the 226 “lending-related” arbitration claims that consumers filed before
NAF between January 2000 and January 2004 were for “small claims.” Ernst & Young used this
term to describe consumer claims up to $15,000. The study distinguished “small” claims from
“medium” claims—ranging from $15,000 to $75,000—and from “large” claims—those above
$75,000.173
Other studies have described amounts at issue in consumer arbitration, though without claiming
that consumer arbitration is effective at handling small claims.174 Using case records from a
nine-month sample of AAA consumer arbitrations that reached awards in 2007, Drahozal and
Zyontz found that 91.5% of consumer claimants brought compensatory claims for $75,000 or
less, and 39.1% of consumer claimants brought claims seeking less than $10,000.175 The average
consumer claim was for $46,131, with most consumer claims under $70,756, although there
were a small number of very high consumer claims.176 They also noted that the proportion of
business claims under $75,000 was higher than the proportion of consumer claims, although
the number of consumer claims under $75,000 was much greater.177

173

Ernst & Young, supra note 144, at 8.

174

See Drahozal & Zyontz, Empirical Research, supra note 140, at 875.

175

See id.

176

See id.; see also id. at 876 (graph showing distribution of amounts claimed by consumers and businesses). For
arbitrations that closed without reaching an award (for example, because they were settled, withdrawn, or otherwise
closed), the study noted that the average consumer claim amounts for those cases increased to $66,367. See id. at 877.
177

Some 94.5% of company claims were for under that threshold, and company claims averaged $22,037. See id. In a
related study, Drahozal and Zyontz also reviewed submitted individual AAA debt collection cases closed between
April and December of 2007, and found the average amount sought by creditors was $20,445. See Drahozal & Zyontz,
Creditor Claims, supra note 140, at 84. The sample for this later study consisted of 61 awarded cases and 406 nonawarded cases. The later study also noted that average claim amounts fell to $1,172 in cases awarded under an AAA-

76

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4.4.2

Metrics

To identify amounts at issue, we have thus far captured data on amounts at issue at two main
points. The standard one-page AAA consumer arbitration “claim form” includes brief
instructions to consumer and company claimants about how to submit a dispute using the
form.178 It asks the submitting party to “briefly explain the dispute.” It then asks: “Do you believe
there is any money owing to you? If yes, how much?” Our first claim amount data point is the
amount stated on the claim form.179 We refer to this as the “claim form amount.”
Second, if the filing revealed a substantive debt dispute, we separately recorded the disputed
debt amount, where that information was available.180 Some such arbitrations involve not only
the disputed debt amount, but other affirmative, non-debt claims. In cases that contain a
disputed debt amount and such affirmative consumer claims, the amount at issue is open to
some interpretation. The consumer may mean to dispute the debt and seek damages for the full
amount of the affirmative claims, but it is also possible that the affirmative claims may not fully
represent a separate amount at issue.181 For these cases, at least for the time being, we provide
administered program of debt collection arbitrations filed by a single debt purchaser. The AAA administered this
program between September 2007 and June 2009. See id. at 85. (To the Bureau’s knowledge, the AAA has not
restarted the program.)
178

The claim form was in consistent use by the AAA for the consumer arbitrations that we studied. In some cases, it
represents the full initial statement of the claims at issue. In others, there may be further pleadings attached, such as
a demand for arbitration or a copy of an underlying court complaint.

179

Sometimes claim amounts are expressed as a range (“$10,000 to $75,000”) or as a bounded inequality (“up to
$75,000” or “at least 30,000”). There are a number of different ways to treat these. Our approach here is the same
method used by Drahozal and Zyontz in their 2010 study. See Drahozal & Zyontz, Empirical Study, supra note 140, at
874. We take the mid-point of ranges, and the base amount for any bounded inequality. Thus, a range of “$10,000 to
$75,000” is treated as a claim for $42,500.” A claim for “up to $10,000” or for “at least $10,000” would be treated the
same: as a claim for $10,000.

180

We include in the disputed debt amount all components of the alleged debt—principal, fees, and other costs. In
some cases, there is a specific indication in the record that the consumer disputes the full amount of the debt. In
others, there is no such qualifier or an explicit statement that the full amount of the debt is at issue; in these cases, we
record the full amount of the alleged debt as the disputed debt amount. If the consumer only takes issue with a
specific portion of the debt, however, we use that stated amount as the debt in issue amount.

181

We recognize that at least in some arbitrations involving disputed debt amounts, the additional consumer claims
may primarily reflect the consumer’s effort to defeat the company’s effort to collect an alleged debt, which raises the
possibility that at least in some such cases the disputed debt amount alone may most accurately reflect the “true”
amount at issue. An alternative view, however, is that even if the debt dispute and the affirmative claims are related,
the sum of the disputed debt amount and the affirmative claim amounts reflects the size of the range between the

77

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

data only on the debt amount at issue, but acknowledge that this may under-report the full
amount at issue for this set of cases.
We have also begun to record the latest statement of the affirmative claim amount that we were
able to find prior to claim resolution, but have not completed that review. (Parties may change
or clarify a claim amount for a number of reasons. They may develop a different valuation based
on evidence. Their awareness of fees may also change claim amounts.182) We also attempted to
record punitive damage claims, but we found such claims so seldom quantified that we do not
think we will be able to offer any assessment of punitive damage claims. (Where punitive
damage claims were separately identified, we did not include them in the claim form amounts.)

4.4.3

Data

Figure 12 shows the distribution of discernible disputed debt amounts in debt collection
arbitrations in the AAA Case Data. Across 492 disputes, the average was $15,725, with a median
of $10,266.183 Across 141 disputes, company-filed debt collection arbitrations had an average
disputed debt amount of $19,619 with a median of $11,303.184 Across 351 disputes, consumerand mutually-filed debt collection arbitrations had an average disputed debt amount of $14,161,
with a median of $10,000.185

outcomes requested by each side. For example, if a consumer sues for $12,000 and a company counterclaims for
$5,000, the range of possible outcomes is between the consumer losing $5,000 (and not winning the affirmative
claims) and the consumer winning $12,000 in damages on the affirmative claims (and not losing on the debt claim).
The difference between these two outcomes—and therefore the amount in dispute—is $17,000, which is the sum of
the disputed debt and the affirmative claim amounts.
182

As explained in section 4.9 below, until March of 2013, claim amounts had a direct impact on fee levels for
consumers and companies in arbitrations covered by the AAA Supplementary Procedures for Consumer Arbitration.
From March 2013, that is no longer the case. See section 4.9.

183

This excludes 30 debt collection cases in which the disputed debt amount was not apparent from the arbitration
record.

184

This excludes five cases in which the disputed debt amount was not identified.

185

This excludes 25 cases in which the disputed debt amount was not identified.

78

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 12:

DEBT AMOUNTS IN DISPUTE IN DEBT COLLECTION ARBITRATIONS 2010-12

FIGURE 13:

DEBT AMOUNTS IN DISPUTE IN ALL ARBITRATIONS 2010-12

79

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

As noted, disputed debt amounts were also at issue in non-debt collection arbitrations. These
were arbitrations in which there was a substantive debt dispute, but where we lacked one or
more of the objective criteria we used to identify debt collection arbitrations. Figure 13,
therefore, shows the distribution of discernible disputed debt amounts across debt collection
and non-collection cases. For the 787 total disputes in which we could identify a disputed debt
amount, the average debt amount was $13,418 and the median was $8,641. Over three years,
there were 21 cases in which the disputed debt amount was at or below $1,000, or an average of
seven such disputes a year. In total, we saw approximately $10.6 million in debt amounts
disputed in AAA arbitrations for these product markets from 2010 through 2012.
Figure 14 shows the consumer claim form amount in all cases in which we could identify a claim
form amount but could not identify a disputed debt amount.186 Across these 326 cases, the
average consumer claim amount was $38,726, and the median was $11,805. Overall, we
identified just under $15 million in claim form amounts in these cases for these product markets
over this period. The breakdown by product was as follows: 146 credit card cases187 had an
average claim amount of $26,187 and a median of $8,945; 61 checking account cases188 had an
average of $66,577 and a median of $15,000; and 119 payday loan cases189 had an average of
$39,834 and a median of $42,500.

186

We have excluded one claim for $100,000,000, both from the figure and from the calculation of the mean and the
median. We have included in the calculation of the mean and median two higher claims, for $500,000 and $1,314,965,
but we have not shown them in the depicted figure. The data showed some clumping of claims at $10,000, $42,500
and $75,000, presumably because consumers may align claim amounts to the thresholds at which filing fees change.
See section 4.9 for more details on how the fee schedules applicable during the period of our review varied by claim
amount. Finally, this calculation excludes cases in which no specific claim amount was identifiable.
187

This excludes 64 cases for which no specific claim amount was identifiable.

188

This excludes 4 cases for which no specific claim amount was identifiable.

189

This excludes 3 cases for which no specific claim amount was identifiable.

80

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 14:

CLAIM AMOUNTS IN CASES WITHOUT DISPUTED DEBT AMOUNTS 2010-12

Figure 15 shows a more detailed breakdown for the claims in Figure 14 at the lower end of this
range. Over three years, therefore, there were 23 claims seeking damages of $1,000 or less, for
an annual average of just over eight such claims. Of these 23, some 14 were for exactly $1,000,
meaning that there were nine such cases over three years seeking under $1,000.

FIGURE 15:

81

SMALLER CLAIM AMOUNTS IN CASES WITHOUT DISPUTED DEBT AMOUNTS 2010-12

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

The Chamber of Commerce, citing Justice Breyer, offers one possible basis for a “small-dollar”
threshold: an example of a “small damage claim” would be “the value of only a defective
refrigerator or television set.”190 The average price of a TV set purchased in Q2 2012 was
$1,224.191 Using that number as a threshold for “small-dollar” claims, there were under 19 cases
on average each year in which there was a “small dollar” debt dispute. Looking only at cases
without disputed debt amounts, but only affirmative claim amounts, there were just over 8 cases
each year on average that were “small dollar.”

4.5

Non-collection arbitrations

4.5.1

Incidence by company

The figures below show the total and annual number of non-collection cases for the most
frequently named companies in these arbitrations. For each of the three markets, we have
included the ten most frequently named companies.192 In some cases, the same company shows
up in more than one figure. These data reflect arbitration filings only; they do not represent
resolved arbitrations or arbitration awards. Given the overall count of these disputes relative to
the customer populations of the various companies, the differences among the companies in
terms of numbers of arbitrations may not be statistically significant. We provide the data in
order to show the number of disputes faced by the companies that are most often parties to
these disputes.
Figure 16 provides data that covers 469 individual credit card arbitration filings over three
years. Company 1 was named in 183 of these; at least one of companies 2 through 10 was named
in the remaining 286. The same ten companies were parties in more than 350 federal court

190

Allied-Bruce Terminix Cos. v. Dobson, 513 U.S. 265, 281 (1995), cited in Chamber RFI Comment at 11-12.

191

Time reported that in Q2 2012, the average price paid for a new TV in the US was $1,224. See
http://business.time.com/2012/08/01/tv-prices-shrink-yet-average-tv-purchase-costs-more.

192

82

We have counted all disputes naming a member of the relevant corporate family.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

credit card filings for the same time period, of which more than 50 were class action filings.193
Company 1 accounted for less than 15% of the federal court filings, although it faced a number of
class filings in court and no class filings in arbitration. The other nine companies, however, were
parties in more federal court filings than in AAA non-debt collection arbitration filings for this
product market.

FIGURE 16:

193

83

NON-COLLECTION CREDIT CARD ARBITRATIONS BY COMPANY 2010-12

See section 4.2.3.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 17:

NON-COLLECTION CHECKING ACCOUNT ARBITRATIONS BY COMPANY 2010-12

FIGURE 18:

NON-COLLECTION PAYDAY ARBITRATIONS BY COMPANY 2010-12

4.5.2

Claim types and claim subject matters

To date, there has been relatively little data available about the kinds of claims that consumers
bring in arbitrations. Some public data give some disaggregation by subject matter of consumer
arbitrations generally, but there is nothing available that would let a reviewer isolate cases in

84

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

particular consumer financial product markets, let alone disaggregate those cases into different
types.
Within each product market, we used a number of different data fields to capture the type of
claims involved and their subject matter (or matters). For all arbitrations, we allowed for an
identification of the basic claim type(s) involved, such as contract law claims, federal statutory
claims, state statutory claims, fraud, and tort.194 This data is shown in Figures 19 through 21. If a
case involved claims under a federal statute, we also recorded the specific statute involved.195
This data, too, is shown in Figures 19 through 21. Each case may have raised more than one type
of claim, but no single case was counted twice in any specific claim category. (Thus, a case that
raised two FDCPA claims and one FCRA claim would show up once in the federal statutory
count, once in the FDCPA count, and once in the FCRA count.) Thus, the percentages shown for
the different claim types do not sum to 100% for any given product.

194

We do not double-count common law fraud under tort as well. “Tort” as used here excludes fraud claims.

195

The federal statutes listed in the credit card cases are the Fair Debt Collection Practices Act (“FDCPA”), the Fair
Credit Reporting Act (“FCRA”), the Fair Credit Billing Act (“FCBA”), the Fair and Accurate Transactions Act
(“FACTA”), the Equal Credit Opportunity Act (“ECOA”), the Communications Act, the Alternative Mortgage
Transactions Parity Act (“AMTPA”), the Telephone Communications Privacy Act (“TCPA”), the Truth in Lending Act
(“TILA”) and the Electronic Financial Transactions Act (“EFTA”). The FCBA is actually codified as part of TILA, but
for clarity we listed FCBA claims separately from other TILA claims. The checking account cases do not add any new
statutes. The payday cases add the Credit Repair Organizations Act (“CROA”).

85

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 19:

CONSUMER CLAIM TYPES IN NON-COLLECTION CREDIT CARD ARBITRATIONS 2010-12

Note: 0% entries in Figure 19 reflect a single instance of the claim.

86

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 20:

CONSUMER CLAIM TYPES IN NON-COLLECTION CHECKING ACCOUNT ARBITRATIONS
2010-12

FIGURE 21:

CONSUMER CLAIM TYPES IN NON-COLLECTION PAYDAY ARBITRATIONS 2010-12

87

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Finally, we also attempted to characterize each dispute by the core subject matter of its claims.
We used a maximum of three such subject matter fields for each dispute. Our intention was to
capture the essence of the dispute, within an objectively applied dispute typology.196 The
typologies were not identical for each product market. This data is shown in Figures 22 through
24. (In addition to the subject matter fields we defined, we also used two residual categories for
each product market: “other” and “not enough information.”)
For credit card disputes, the listed subject matters in Figure 22 are: substantive debt dispute;
debt collection process (harassment); debt collection process (other); payment allocation;
payment posting; interest rate promotions; balance transfer promotions; add-on products;
interest rates/charges; late fees; account opening issues; account closing issues; credit line
issues; unauthorized account use; merchant-related error; credit reporting; disclosure of private
information; discrimination; improper set off; and other fee issues. As discussed in Appendix B,
our subject matter typology for credit card cases is broader than this, but these were the core
subject areas within that typology that were raised by these cases. It is notable that while these
cases did not meet our formal ‘debt collection’ definition, the clear majority (346 out of 515, or
67%) still included substantive disputes over the amount of debt owing. In fact, although almost
all non-collection credit card disputes were filed by consumers or as mutual submissions, a
substantial share (131 out of 515, or more than 25%) of these disputes raised only substantive
debt issues—and no non-debt claims at all. (As noted above, therefore, our formal definition of
debt collection arbitrations may have resulted in our undercounting the true number of such
arbitrations.)

196

Our subject matter categories are to some extent subjective. Clearly, it is possible to devise other typologies that
are more or less disaggregated. We derived ours from a review of complaints filed on electronic litigation databases
and by consultation with Bureau experts in specific fields. Our methodology is explained in greater detail in Appendix
B.

88

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 22:

CREDIT CARD NON-COLLECTION ARBITRATION SUBJECT MATTERS 2010-12

Notes on Figure 22:
1.
2.

Entries showing 0% reflect either one or two cases.
In 17 of the 32 “Debt collection process (harassment)” disputes, the consumer also brought claims refuting the
amount of debt claimed, but that information is not reflected in the substantive debt dispute entry. Similarly, in
37 of 42 “Debt collection process (other)” disputes, the consumer also brought claims refuting the amount of debt
claimed, and that information is not reflected in the substantive debt dispute entry. These records were not
characterized as substantive debt disputes due to process limitations.

For checking account disputes, the listed subject matters in Figure 23 are: debt collection
process (harassment); overdraft ordering/timing; other overdraft issues; account closing issues;
account opening/reopening issues; deposits—process issues; deposits—other errors; authorized
payments—process issues; authorized payments—other errors; improper set-off; unauthorized
recurring payments; disclosure of private information; other unauthorized use; credit reporting;
and discrimination. As with credit card, our typology was broader than this, but these were the
core subject areas raised by these cases.

89

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 23:

CHECKING ACCOUNT NON-COLLECTION ARBITRATION SUBJECT MATTERS 2010-12

For payday loan disputes, the listed subjects in Figure 24 are: substantive debt dispute; debt
collection process (harassment); rollover issues; improper loan duration; failure to
disclose/follow core terms; 197 failure to disclose/follow other terms; unlicensed lending activity;
interest/fees above state cap; loan amount above state cap; improper access to DDA (ACH);
improper access to DDA (check); product mischaracterization; other unfair loan practices;
Credit Service Organizations (CSO)—state law issues; and CSO—federal law issues. Once again,
our typology was broader than this, but these were the core subject areas raised by these cases.
For non-collection arbitrations that did not raise substantive debt disputes, we have not
generally identified recurring pleading forms across cases. We have identified 18 payday
disputes that were essentially the same, each involving the same consumer counsel in 2010.
These cases were previously all joined in a single court action that was dismissed in favor of
arbitration upon the defendant’s motion. They were then refiled in arbitration.

197

90

We defined core terms to be finance charge and loan duration.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 24:

PAYDAY LOAN NON-COLLECTION ARBITRATION SUBJECT MATTERS 2010-12

Note: In six of the seven “debt collection process (harassment)” disputes, the consumer also brought claims refuting
the amount of debt claimed, but that information is not reflected in the substantive debt dispute entry because of
limitations in our coding hierarchy.

For non-collection arbitrations that included substantive debt disputes (a category
overwhelmingly made up of credit card cases198), we have not yet identified recurring patterns in
the cases that raise only debt issues. In the cases that raised debt disputes and non-debt
affirmative claims, however, we did see some recurring pleading forms used by consumer
counsel across some of those cases.

198

There are 359 non-collection arbitrations that include substantive debt disputes. Thirteen of these are payday
cases. The remaining 346 are credit card cases.

91

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4.6

Debt collection arbitrations

4.6.1

Incidence

In light of NAF’s demise and the AAA’s continuing moratorium on company-filed debt collection
arbitrations, the Bureau noted in its 2012 RFI that we were unaware of recent practice by
covered persons to bring claims—and in particular debt collection claims—against consumers in
arbitration. No RFI commenter specifically disputed this observation.199 The number of
company-filed debt collection claims—an annual average of 49—in the AAA Case Data was
consistent with the Bureau’s view.200 Adding consumer-filed and mutually-submitted debt
collection arbitrations to the total count, we identified 522 debt collection arbitrations across
three years, or an annual average of 174 such filings. Across three years, all but four of these
were credit card disputes, meaning that there were an annual average of 173 credit card debt
collection disputes.
Although the 518 credit card debt collection disputes were the biggest single category in the AAA
Case Data, they represented a negligible share of credit card debt collection activity overall.
Although comprehensive national data on credit card collection litigation is not generally
available, our review of small claims court cases offers some indication of how rare debt
collection arbitration is at this point. In Philadelphia County, in 2012, in small claims court, and
focusing only on a group of 10 issuers (rather than all issuers and debt buyers of credit card
debt), we identified more than 2,200 collection cases, more than four times as many credit card
debt collection disputes as there were in AAA debt arbitration for the entire country over three

199

The American Bar Association, the Consumer Bankers’ Association, and the Financial Services Roundtable
confirmed that “since July 2009 arbitration has not been used on a large scale for consumer debt collections initiated
by companies.” ABA/CBA/FSR RFI Comment at 7, 16.
200

Almost all company-filed claims were debt collection arbitrations. For credit card disputes, 145 out of 153
company-filed cases were debt collection arbitrations. This was consistent with prior research concluding that
company claims in arbitration seek payment for goods delivered or services rendered and for “usually very little else.”
Drahozal & Zyontz, Empirical Study, supra note 140, at 872; see also Drahozal & Zyontz, Creditor Claims, supra note
140, at 84 (noting that of the 61 company-filed consumer arbitrations to reach an award in their sample, only 3 were
likely not classifiable as debt collection cases).

92

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

years.201 Despite the relative rarity of debt collection arbitration, however, a slight majority of
the credit card disputes in the AAA Case Data were debt collection arbitrations.
Figure 25 shows the number of credit card debt collection arbitrations, broken out by the ten
companies that recur most frequently in the AAA debt cases that we reviewed.202 As with the
company data on non-collection arbitration filings, the variation here may not have been
significant in light of the size of the overall product markets. One company, however, accounted
for more than two and a half times the arbitration proceedings of the other nine companies,
averaging 118 debt collection arbitrations per year. The drop-off to the remaining companies
was precipitous: the fifth company on the list averaged only two debt collection arbitrations per
year, and every company after that averaged only one.

FIGURE 25:

CREDIT CARD DEBT COLLECTION ARBITRATIONS BY COMPANY 2010-12

Company 1
Company 2
Company 4
Company 3
Company 7
Company 6
Company 26
Company 6

2010

Company 5

2011

Company 27

2012
0

201

100

200

300

400

See section 5.3 for detailed data on Philadelphia County small claims court cases.

202

We have not included checking account or payday debt collection cases because there are only four across the
three years.

93

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

We also note that a relatively high share of the debt collection cases were recorded on claim
forms as being filed by consumers or mutually submitted. Of the 518 credit card debt collection
arbitrations, more than 72% showed as mutually submitted or consumer-filed. (Most of these
were consumer-filings, but nearly 9% were recorded as mutual submissions.203) In addition,
more than 64% of the credit card debt collection cases recorded as filed by consumers or
mutually submitted (237 out of 373) raised only substantive debt collection issues, and no nondebt claims at all.
To our knowledge, prior studies covering consumer-filed AAA arbitrations have not previously
noted consumers bringing debt collection disputes to arbitration. We may look further into
possible explanations for this apparent phenomenon. How and when such cases are resolved
may provide us with additional information. We may also confer with some of the counsel
involved. As we note below, some firms that represent consumers in these cases show a pattern
of using common pleading forms across cases.

4.6.2

Subject matter and claims

Figure 26 shows the distribution of claim types made by consumers in the debt-collection cases.
It follows the same format as Figures 19 through 21. Any one case may have given rise to more
than one claim, but no case recorded the same type of claim twice. We limited this to credit card
cases only because there were so few debt collection cases for the other two products.
As noted above, some debt collection arbitrations contained consumer claims or counterclaims
that went beyond disputing the substance of the alleged debt. Over a quarter of the credit card
debt collection arbitrations included non-debt affirmative consumer claims. Almost all of these
non-debt claims were made in debt collection disputes filed by consumers or that were mutual
submissions.204

203

Consumers were recorded as filing 67% of the credit card non-c0llection cases that include substantive debt
disputes. The remaining 33% of this same category were recorded as mutual submissions.
204

Consumers made non-debt counterclaims in only around 3% of company-filed debt collection arbitrations. By
contrast, they made non-debt claims in 36% of the consumer-filed debt collection arbitrations and in 42% of mutually
submitted debt collection arbitrations.

94

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 26:

CONSUMER CLAIM TYPES IN CREDIT CARD DEBT COLLECTION ARBITRATIONS 2010-12

Note: 0% entries in Figure 26 reflect a single instance of the claim.

Figure 27 shows the subject matter of non-debt claims for credit cards.205 (We excluded
substantive debt disputes from Figure 27 because, by definition, they were present in all these
cases. In addition, we do not provide a separate break-out for payday and checking account
disputes because almost all the debt collection arbitrations concerned credit cards.) In collection
cases that raised affirmative non-debt claims as well as substantive debt disputes, we saw some
recurring pleading forms used by consumer counsel across a share of those cases, just as we did
for non-collection cases raising debt and non-debt disputes. There were more limited
indications of common patterns across the debt collection cases in which the only claims
concern the amount of debt owed.

205

We use “non-debt claims” expansively to include all affirmative claims made by the consumer. That includes
FDCPA claims, unless the only basis of the FDCPA claim is that the amount of debt claimed is incorrect. If we
identified a consumer FDCPA claim that lacked any further specification of the nature of the claim, we assumed for
purposes of the subject matter typology that the claim related to the amount of debt asserted and did not otherwise
challenge the process of collection. Absent that assumption, the debt collection subject categories in Figure 27 (and
the debt collection process categories in Figure 22) would be larger.

95

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 27:

CREDIT CARD DEBT COLLECTION ARBITRATION SUBJECT MATTERS 2010-12

Note: Entries showing 0% reflect either one or two cases.

4.7

Which consumers bring arbitrations?

We are not aware of scholarship attempting to characterize the demographics of consumers
involved in consumer arbitrations. In the area of employment arbitration, however, scholars
have addressed demographic issues in a number of ways. There has been some work to try and
determine the income level of employee claimants in arbitration.206 There has also been work to
address outcome variations across arbitrations filed by highly paid employees and lower paid
employees.207 A number of studies have found significant variations in employment arbitration

206

See Hill, Due Process at Low Cost, supra note 169, at 794 (estimating that 43.5 % of employee claimants in an
AAA sample of 200 awarded cases earned between $14,000 and $60,000).

207

See Theodore Eisenberg & Elizabeth Hill, Arbitration and Litigation of Employment Claims: An Empirical
Comparison, DISP. RES. J., Nov. 2003-Jan. 2004, at 48 (noting an employee “win rate” of 64.9% in AAA employment
arbitrations that involve higher paid employees, and 39.6% in those involving lower paid employees). Building on this
work, Choi and Eisenberg note that securities arbitration claimants have win rates between these levels. As an issue
for future research, they comment that “this pattern may suggest that increased resources are associated with
claimant arbitration success.” Stephen Choi & Theodore Eisenberg, PUNITIVE DAMAGES IN SECURITIES ARBITRATION: AN
EMPIRICAL STUDY 535 (2010). Alexander Colvin has identified a 19.7% employee win rate in a data set of AAA
employment arbitrations for which the “large majority” did not involve highly paid employees. See Alexander Colvin,

96

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

outcome according to whether the employee claimants were relying on arbitration clauses in
individually negotiated contracts—which was likely to be the case when the claimant was a
manager or executive—or instead on clauses in personnel manuals—which was likely to be the
case for other employees.208 More recently, some work has asserted that a claimant’s gender
makes a difference to employment arbitration outcomes.209 These various studies suggest that
there may be value in trying to understand which consumers bring arbitrations in the consumer
financial area.
To address this issue, we constructed a location-based income profile of U.S. credit
cardholders.210 To do this, we identified census tract data from a representative sample of U.S.
credit cardholders using data from the Bureau’s Consumer Credit Panel, which provides a
nationally representative, random sample of de-identified credit records procured from a large,
national consumer reporting agency. Using the Census, we then obtained median income data
for all the census tracts involved.211 (We do not have actual income data for any individual from
the credit panel.) This provides an estimate for the percentage of U.S. credit cardholders that
live in census tracts within certain median income ranges. This background income profile is
shown in the darker green columns in Figure 28.

Empirical Research on Employment Arbitration: Clarity Amidst the Sound and Fury?, 11 EMPLOYEE RIGHTS &
EMPLOYMENT POL’Y J. 418-19 (2007). A later Colvin study shows a similar win rate for analogous AAA employment
arbitrations at 21.4%. See Colvin, supra note 169, at 6. This later study also shows win rates increasing with salary
levels. For employees with salaries below $100,000, win rates were 22.7%. They rose to 31.4% for employees with
salaries between $100,000 and $150,000. For employees salaried at above $150,000, win rates were 42.9%. See id. at
10-11.
208

See Lisa B. Bingham, An Overview of Employment Arbitration in the United States, N. Z. J. INDUS. RELS. at 13, 16
(June 1998). The AAA classifies employee arbitrations according to the type of underlying agreement: individually
negotiated or in employer promulgated agreements. See Colvin, supra note 207, at 415. Bingham’s results were that
employees won 68.8% of the former type of arbitrations, but only 21.3% of the latter type. See id.

209

See Lipsky et al., supra note 139. This study reports that the gender of the claimant and the claimant’s counsel had
a “significant effect” on award size. The gender of the arbitrator did not. See id. at 314. The authors acknowledge that
numerous explanations are possible for this observed effect, including discrepant settlement rates, lower seniority of
female claimants, and greater seniority of male counsel. See id. at 320-22.

210

Appendix B provides more detail about our methodology.

211

For 309 Census tracts, covering around 0.12% of the data relating to the overall credit card population, we had no
reported median income.

97

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

We then constructed a location-based income profile of 420 consumers in non-debt collection
credit card arbitrations.212 (We focused, first, on non-debt collection cases, but, as we discuss
below, we also covered debt collection cases.) To build this profile, we determined a census tract
location for filing consumers. Once again, we then used the Census to determine median income
for that census tract. As a result, we do not have any actual income data for any consumer that
filed an AAA arbitration. We show this information in the lighter green columns in Figure 28.
By plotting one distribution against the other, we were able to compare the median income of
the locations involved in consumer arbitrations in the AAA Case Data with the median income of
the locations of credit cardholders generally. Figure 29 shows the difference between the two
distributions. Overall, we saw that locations with incomes below $50,000 were less likely to
account for credit card arbitrations than for credit cardholding overall. Locations with incomes
above that level sometimes accounted for more credit card arbitrations than for credit
cardholding overall, but not consistently so.

212

We were not able to associate census tract-based median income information for all of the 515 credit card nondebt collection cases. In some cases, the AAA case records did not provide sufficient location information. The
analysis here, therefore, covers the 420 cases for which we were able to obtain such information.

98

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 28:

MEDIAN INCOME OF CREDIT CARDHOLDER LOCATIONS AND LOCATIONS OF CONSUMERS
INVOLVED IN NON-COLLECTION CREDIT CARD CONSUMER ARBITRATIONS 2010-12

FIGURE 29:

PERCENTAGE DIFFERENCES BETWEEN DISTRIBUTIONS IN FIGURE 28

99

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

We also reviewed all credit card arbitrations, including debt collection arbitrations. Such
disputes might be expected to originate in poorer locations relative to the distribution of credit
cardholders overall. But even including all credit card arbitration filings in this analysis—which
are primarily debt collection disputes—we see a similar pattern. Consumers in such
arbitrations—even when the disputes involve debt claims—were, relative to the cardholding
population overall, disproportionally located in areas above $35,000 in average income.
Locations with incomes above that level sometimes account for more credit card arbitrations
than for credit cardholding overall, but not consistently so. We show this analysis in
Appendix D.
We were not able to follow this same approach for checking account and payday loan
arbitrations because we do not have data that provides a proxy for the income level of users of
these products by geographic distribution. We can construct distributions for the income level of
areas that originate such arbitrations, but we can only compare these to the background
distribution of income levels for all areas. As a result, we cannot control for the income level of
the consumers that use these two products in the manner that we can for credit cards.213
The AAA Case Data, therefore, may suggest that credit cardholders who arbitrate will tend to
exceed a certain income threshold more than credit cardholders generally exceed that same
threshold. It is important to note, however, that we cannot tell from our data whether or not the
same was true for consumers in litigation.214 Our data were not sufficient to address the issue
comparatively. In addition, as noted in section 4.2.1, we worked from a relatively small set of
arbitrations when compared to the overall credit cardholding population, which underscores the
need for caution in interpreting this demographic data.

213

Appendix D also contains the checking account and payday distributions. As might be expected, payday
arbitrations are disproportionately filed from lower-income zip codes, which may reflect that lower-income
Americans are the primary users of this product. Conversely, checking account arbitrations are disproportionately
filed from zip codes with a median income of $50,000 to $74,999 in average income, which may reflect the higher
income level of checking account users. (As discussed in the appendix, however, the distribution regarding checking
account arbitrations contained the least number of data points—56.)
214

Class litigation that results in an automatic distribution of relief to affected consumers, however, stands to provide
some benefit to all income and demographic segments represented in the affected customer base.

100

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4.8

Prior litigation

In some cases, the case record for a dispute in the AAA Case Data provides a clear indication of
prior court proceedings. We discuss the available record on prior court litigation in this section.

4.8.1

Individual disputes

In credit card debt collection disputes, we were able to identify evidence of prior court
proceedings in 97% of such cases. This is unsurprising given how we have defined debt
collection arbitrations. In these cases, prior collection proceedings were the norm; in effect, the
debt collection arbitration took the place of the preceding debt collection litigation.
Furthermore, leaving aside the mutually submitted debt collection cases, we have identified only
eight debt collection cases in which there was a clear record that the company was the party to
invoke arbitration in the prior litigation, indicating that in many of the debt collection cases the
consumer was the party ending the litigation in favor of arbitration.215 Although this may be a
common pattern in the AAA Case Data, our Philadelphia County small claims court results gave
some indication of how rare this consumer response to collection litigation may be overall.216
There were more credit card collection cases filed by just two issuers in that county for 2012
alone than for all three years of credit card data in the AAA case records from 2010 through
2012.
In non-collection cases, by contrast, an indication of prior litigation is the exception, not the
rule. We were able to identify from the AAA case records only 78 such cases, or 11% of the 718
non-collection arbitrations. In 19 of these 78 cases, the consumer appeared to have been the
party to invoke arbitration as to the prior court proceeding. In the remaining 59 cases, the
available indications were that the company was the party invoking arbitration as to the court
proceeding.
It is important to note, however, that these results were based only on a review of the AAA
electronic case record. We have not attempted to identify a preceding federal or state court case

215

It is possible that in the absence of the moratorium, the company might have filed such disputes in arbitration
directly.
216

See section 5.3.

101

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

using electronic litigation case data.217 Our data, therefore, may show false negatives on this
point: there could be additional court proceedings that were not viewable from the arbitration
record alone. In addition, we do not know how frequently companies move in court to invoke
arbitration in individual proceedings and thus cannot assess the frequency with which such
actions by companies lead to consumer-initiated arbitration cases.

4.8.2

Class disputes

A number of RFI commenters suggested that the Bureau compare the benefits to consumers
from arbitration to the benefits from class action litigation.218 In fact, the value of such a
comparison is an area of relative agreement across RFI commenters, including consumer groups
and members of the plaintiffs’ bar. Some commenters noted that there is limited empirical data
in this area.219
The comparison can potentially be made along a number of different metrics, although the lack
of small-dollar disputes in the AAA arbitration caseload may mean that at least some “apple to
apple” outcome comparisons may not be available. The presence of prior class litigation
proceedings, however, provides one point of potential comparison.
In damages class actions, applicable rules generally entitle potential class members to “opt out”
of any proposed class.220 Unless a potential class member opts out, he or she will be bound by
the resolution of the dispute. Potential class members, therefore, generally receive individual
notice of their potential inclusion in a certified class, and are thereby given the opportunity to
opt out.221

217

As several commenters pointed out, state court records are not consistently available across all state court
jurisdictions via electronic means. There is no analog to PACER—the federal court record database—that covers all
state courts.
218

ABA/CBA/FSR RFI Comment at 3-4, 8 (stating that the Bureau should study whether “class actions provide
meaningful benefit to individual consumers as compared with individual arbitration”).

219

ABA/CBA/FSR RFI Comment at 8.

220

See FED. R. CIV. P. 23(c)(2)-(3).

221

See id.

102

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

In many cases, however, notice will not be provided to potential class members until the parties
have reached a proposed settlement. As a result, many class action notices will also include
notice of a proposed settlement, laying out two broad options. A potential class member can take
the offered class settlement terms, which will extinguish his or her right to bring individual
proceedings on the settled claims. Alternatively, he or she can opt-out of the class action and the
proposed settlement, and thus be free to pursue an individual case if he or she wishes to do so.
Such “opt outs” are not bound by the terms of the class settlement releases. They receive no
relief from the compensatory terms of the damages settlement. As they are not bound by the
releases, they are free to press their own claim individually against the settling company. To
preserve that right, courts that approve a class action settlement will typically publish a list of
individual opt-outs by name.
Some class action settlements will occur in situations in which the individual consumer who
opts out can bring a claim in arbitration.222 Class proceedings will toll the applicable statute of
limitations for putative class members to bring individual actions outside the class litigation.223
The class proceeding may have brought forward certain evidence bearing on the claims in the
case. It may also have identified and tested certain legal theories, and it may have identified
certain conduct as potentially actionable. In short, a certain amount of work will already have
been done for any consumer who may have a preference for a potential arbitrated outcome over
the proposed class settlement outcome.
Using electronic case databases, as well as blogs and websites that track class action settlements,
we have identified eight class action damages settlements that meet the following criteria: first,
final approval of the settlement took place from the latter half of 2009 or later; second, the
contract at issue in the class action contained an arbitration clause that offered AAA as an
arbitral forum; and third, the class action involved one of the three products covered by our
existing AAA Case Data review. (Other class action settlements may also meet these criteria, but
these are the ones we have identified to date.) Our criteria were intended to locate cases for

222

This will generally only be the case where, for whatever reason, the defendant has either opted against invoking
the arbitration clause in the class action or tried to do so and been unsuccessful in that effort. Had the defendant
successfully invoked the arbitration clause as to the class action, there would be no class action settlement in court. In
all eight of the cases identified below, one or more defendants attempted to invoke an arbitration clause but were not
fully successful in that effort.
223

103

See, e.g., Crown Cork & Seal v. Parker, 462 U.S. 345, 352 (1983).

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

which we would be able to tell if consumers opted out and brought a AAA arbitration instead of
applying for or receiving benefits under the class action settlement. In addition to the AAA Case
Data for 2010 through 2012, we also had some AAA data for 2008, 2009 and 2013 that allowed
us to check for potential overlap.224 Finally, none of the cases we identified involved presettlement notice to the class. Had they done so, opt-outs might have filed arbitrations earlier
than the period of our review.
For all the cases below, we have looked through all three years of the AAA Case Data as well as
the 2013 AAA data to count the number of consumers who opted out and pressed the same
claim against one or more of the relevant defendants in arbitration before the AAA. For one
case, for which notice was sent in November 2007, we also check AAA data for 2008 and 2009.
Thus far, we have identified three arbitrations in which an opt-out from one of these cases may
have made the same claim in AAA arbitration against a party within the scope of the applicable
settlement.225 No other opt-out from one of these cases appears to have filed the same dispute
before the AAA. A total of 3,605 individuals opted out of these settlements. More than 13 million
class members made claims or received payments under these settlements. Total payments or
debt relief to the classes are in excess of $350 million, exclusive of attorneys’ fees and the value
of injunctive relief. Further details on each of these class cases are provided below.

224

Because the detailed AAA Case Data were available from the beginning of 2010, we set late 2009 as the cut-off for
final approval. For cases settled on that schedule, opt-outs bringing AAA arbitrations after final approval should
generally be visible in the detailed 2010-12 case records.
225

We cannot be sure that these three opt-outs, in fact, pressed the same claim in arbitration. For two of these optouts, the potentially overlapping AAA arbitration was filed before 2010. For that period, we lack detailed case records,
but the available AAA data record arbitrating parties matching two opt-outs. In addition, the relevant AAA
arbitrations involved companies covered by the scope of the applicable class settlement. For these two opt-outs,
therefore, it is possible that they were involved in follow-on AAA arbitration, but we lack the information to say
whether that was the case. For the third opt-out, there is a match in the AAA case records and that arbitration
involves the company that reached the relevant class settlement. However, even though we have the case record for
this arbitration, the consumer’s claims are stated only in very general terms. It is possible that these general terms
were intended to capture one or more of the claims covered by the prior settlement, but we do not have enough
information to say whether that was the case.

104

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

IN RE CURRENCY CONVERSION FEE ANTITRUST LITIGATION
The plaintiffs challenged foreign currency and other foreign transaction action fees imposed on
payment card transactions.226 The defendants were Visa, MasterCard, and a number of large
credit card issuers. Notice of the proposed settlement was sent to potential class members in
November 2007.227 The settlement was “claims-made,” meaning that consumers had to submit
claims to secure payments under the settlement.228 More than 10 million consumers submitted
claims under the settlement. (Of these, more than 7 million consumers requested a flat fee
payment of $25, which was subsequently reduced to $18 because of claim volume. Another 3
million consumers submitted more detailed claim materials, which entitled them to a
significantly larger recovery.229) Overall, the settlement distributed around $263 million to
consumer members of the settlement class.230 The court entered final approval of the settlement
in 2009, at which time the court excluded 2,878 individuals from the damages class.231 Several
of the defendants did not have AAA arbitration clauses during some portion of the relevant
period, but most of the defendants had clauses up until the end of 2009, and one had a clause
through the entire period. In addition, the settlement released all U.S. Visa and MasterCard
members, not just the defendants, so the potential for arbitration was widespread.232
HOOPER V. ADVANCE AMERICA
The plaintiffs alleged violations of Missouri’s Merchandising Practices Act and Missouri’s
payday loan statute.233 An arbitration clause accompanied the payday loan contracts provided by

226

In Re Currency Conversion Fee Antitrust Litig., MDL Docket No. 1409 (S.D.N.Y., filed before J.P.M.L. Apr. 13,
2001).

227

See Decl. of Edward J. Radetich, Jr., CPA Regarding the Dissemination of Notice to the Class ¶ 7 (Jan. 28, 2008).

228

See Aff. of Edward J. Radetich, Jr., CPA Regarding Claims Administration at 5 (Sept. 14, 2011).

229

See id. at 4, 6.

230

See id. at 21.

231

See Final Judgment & Order of Dismissal, Ex. A (Nov. 4, 2009).

232

See id. ¶ 9.

233

Hooper v. Advance America, Cash Advance Centers of Missouri, Inc., No. 08-cv-04045-NKL (W.D. Mo. filed Mar.
10, 2008).

105

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

the defendant, and AAA was an arbitral forum offered in the contract. In July 2010, notice of a
proposed settlement was mailed to potential class members.234 The court approved the
settlement in November 2010, identifying 316 consumers who opted out of the settlement
class.235 There were approximately 10,400 timely claims submitted. Claimants shared
approximately $520,000 in cash payments, and between $3.8 million and $9 million in debt
forgiveness.236
HOFFMAN V. CITIBANK
The plaintiffs alleged that Citibank retroactively increased interest rates on certain customers’
outstanding credit card balances in violation of several California laws.237 Citibank administered
the settlement, and provided notice of the proposed settlement via statements.238 The court
approved the proposed settlement in December 2010, identifying 140 opt-outs.239 At that point,
the claims period was still open, so we have not been able to identify the final number of claims
or the amount paid to class members. Class members were entitled to payments of $18, subject
to pro rata decrease if there was sufficient claims volume.240 At the time of final approval, when
the claims period had several months to run, there were 12,500 claims, which suggests that the

234

See Decl. of Jennifer M. Keough Regarding Notice Dissemination in Support of Plaintiffs’ Unopposed Motion for
Final Approval of Class Action Settlement ¶ 3 (Oct. 27, 2010), ECF No. 65.

235

See Order and Final Judgment, Ex.1 (Oct. 27, 2010), ECF No. 68.

236

The $520,000 is an estimate based on the following amounts: a $2 million cash fund, less $950,000 in attorney
fees and costs, $2,000 in service fees to class representatives, $398,800 to cy pres, and the fund was offset by
$127,810 for dollar-to-dollar debt forgiveness for one subclass. See id. at 8; Memorandum in Support of Motion for
Cy Pres Disbursement and Notice of Final Accounting at 2 (Dec. 29, 2010), ECF No. 70.

237

Hoffman v. Citibank (South Dakota), N.A., No. 06-cv-00571-AG-MLG (C.D. Cal. filed Jun. 21, 2006).

238

See Supp. Decl. of Deborah L. Thompson re Claims Administration in Support of Final Approval of Settlement ¶ 2
(Dec. 15, 2010), ECF No. 86.

239

See Settlement Order and Final Judgment ¶ 6 (Dec. 22, 2010), ECF No. 95.

240

See Settlement Agreement at 10 (Apr. 22, 2010), ECF No. 61.

106

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

class received over $225,000.241 Citibank has maintained an arbitration clause providing for
AAA arbitration throughout the relevant period.
KUCAN V. ADVANCE AMERICA
Plaintiffs alleged various violations of North Carolina law by a payday lender. A proposed state
class settlement was reached in September 2010.242 Notice was mailed in November 2010.243 In
early 2011, the court entered final approval of the settlement, identifying 19 opt-outs.244 Advance
America’s loan contract included an arbitration clause that permitted AAA-administered
arbitration. This was not a claims-made settlement—payments were made automatically to all
class members. Approximately 135,000 class members shared approximately $11.5 million.245
HAGER V. CHECK INTO CASH
Paralleling the claims in Kucan, plaintiffs alleged various violations of North Carolina law by a
payday lender.246 The arbitration clause at issue allowed AAA to serve as the arbitral forum. A
proposed settlement was reached in December 2010.247 Notice was sent in February 2011.248
Final approval was entered in April 2011.249 Ten consumers opted out of the settlement class.250

241

See Supp. Decl. of Deborah L. Thompson re Claims Administration in Support of Final Approval of Settlement ¶ 3
(Dec. 15, 2010), ECF No. 86.

242

Kucan v. Advance America, Cash Advance Centers of North Carolina, Inc.; Advance America, Cash Advance
Centers, Inc.; and William M. Webster, IV, No. 04-CVS-2860 (N.C. Super. filed Jul. 26, 2004).
243

See Report by Settlement Administrator ¶ 3 (Jan. 18, 2011).

244

See Status Report & Mot. Requesting Entry of Final Judgment ¶ 2 (Jan. 18 2011).

245

See Aff. of Settlement Administrator ¶¶ 2-3 (Jun. 26, 2013).

246

Hager v. Check Into Cash of North Carolina, Inc. et al., No. 04-CVS-2859 (N.C. Super. filed Jul. 27, 2004).

247

See Settlement Agreement (Dec. 10 2010).

248

Interim Status Report & Mot. Requesting Entry of Final Judgment ¶ 1 (Mar. 10, 2011).

249

See Settlement Order and Final Judgment (Apr. 15, 2011).

250

See id., Ex. 1.

107

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

As in Kucan, there was no claims process and payments were made automatically to class
members. Approximately 104,000 consumers shared more than $7.6 million.251
IN RE CHECKING ACCOUNT OVERDRAFT LITIGATION
In a number of cases consolidated in this multidistrict litigation in the Southern District of
Florida, plaintiffs have challenged overdraft fees that were allegedly incurred as a result of the
order in which the banks processed certain debit transactions.252 Several banks have separately
settled overdraft class actions. Of the defendants to settle so far, Chase, M&I, and Compass each
have checking account agreements that provided for AAA arbitration.
Chase reached a proposed settlement in May 2012.253 Notice was mailed to class members in
August 2012.254 The court gave final approval in December 2012, identifying 173 consumers who
opted out.255 Most class members received automatic payments under the settlement, but class
members were also entitled to submit claim forms for periods for which Chase lacked relevant
records.256 The class had more than five million members, who received approximately $61

251

See Aff. of Settlement Administrator ¶ 2 (Jul. 1, 2013).

252

In Re: Checking Account Overdraft Litig., MDL No. 2036 (S.D. Fla., am. consolidated compl. filed Nov. 9, 2009).

253

See Pls.’ Unopposed Mot. for Prelim. Approval of Class Settlement and for Certification of Settlement Class, Ex. A
(May 22, 2012), ECF No. 2710-1.
254

See Decl. of Shannon R. Wheatman, Ph.D. on Implementation and Adequacy of Notices and Notice Plan ¶ 7 (Oct.
15, 2012), ECF No. 3010-5.

255

See Final Judgment, Ex. A (Dec. 19, 2012), ECF No. 3135.

256

See Order of Final Approval of Settlement, Authorizing Service Awards, and Granting Application for Attorneys’
Fees at 9-11 (Dec. 19, 2012), ECF. No. 3134.

108

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

million in payments.257 As part of the settlement, Chase also agreed to restrictions on fees for
transactions under $5. The court valued this additional relief at $52 million.258
M&I reached a proposed settlement in September 2012.259 Notice was mailed in January
2013.260 The court gave final approval in August 2013, identifying 34 opt-outs.261 Class members,
who numbered around 190,000, did not have to submit a claim form to share in $2.7 million in
cash relief.262 The settlement also contained limits on charging fees for small overdrafts.263
Compass Bank reached a proposed settlement in March 2013.264 Notice was mailed in April
2013.265 Notice was sent to approximately 826,000 potential class members.266 The court

257

The $61 million cash distribution is estimated as follows. The total cash to be distributed was $110 million. The
court awarded class counsel $48.6 million, which was 30% of the total $162 million value of the settlement, including
injunctive relief. The settlement fund also covered $309,326.50 for litigation expenses, and $5,000 or $2,500 service
awards to each named plaintiff. See Order of Final Approval of Settlement, Authorizing Service Awards, and Granting
Application for Attorneys’ Fees at 39 (Dec. 19, 2012), ECF No. 3134.
258

See id. at 12.

259

Settlement Agreement (Oct. 1, 2012), ECF No.2981-1.

260

Decl. of Cameron R. Azari on Implementation and Adequacy of Settlement Notice Program ¶ 16 (Oct. 1, 2012),
ECF No. 3279-5.

261

Final Judgment, Ex. A (Aug. 2, 2013), ECF No. 3571.

262

The $2.7 million cash distribution is estimated as follows. The court awarded class counsel 30% of the total value
of the settlement, which was $4 million. It also awarded $67,362.50 for reimbursement of litigation expenses, and
$2,500 service awards to each named plaintiff. See Order of Final Approval of Settlement, Authorizing Service
Awards, and Granting Application for Attorneys’ Fees (Aug. 2, 2013), ECF No. 3570.

263

See Settlement Agreement ¶ 50 (Aug. 2, 2013), ECF No. 3570.

264

Settlement Agreement, No. 09-md-02036-JLK (S.D. Fla. Mar.12, 2013), ECF No. 3330-1.

265

See Decl. of Cameron R. Azari on Implementation and Adequacy of Settlement Notice Program ¶ 15 (May 16,
2013), ECF No. 3469-5.

266

109

See id.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

entered final approval in August 2013, identifying 35 opt-outs.267 Class members did not have to
submit claim forms to share in approximately $8 million in cash relief.268
With respect to the Compass and M&I settlements, we do not have significant AAA data to
review because these settlements were approved so recently. We do, however, have some
months of AAA data for the period immediately following class notice and can report that that
period at least was insufficient to generate any pertinent AAA arbitrations.

4.9

Initial fee allocation

The consumer arbitrations reviewed by the Bureau are subject to the AAA’s Due Process
Protocol, which requires that arbitration be available to consumers at a reasonable cost.269 The
AAA’s Supplementary Procedures for Consumer-Related Disputes are intended to implement
this and other requirements of the Protocol. Below we briefly describe the fee schedule that
applied to consumer arbitrations filed with the AAA during the period of our review. The AAA
changed its fee schedule effective March 1, 2013.270

267

Final Judgment, Ex. A (Aug. 7, 2013), ECF No. 3586.

268

The $8 million cash distribution is estimated as follows. The court awarded class counsel 30% of the total value of
the settlement, which was $11.5 million. It also awarded $106,300.20 for reimbursement of litigation expenses, and
$5,000 service awards to each named plaintiff. See Order of Final Approval of Settlement, Authorizing Service
Awards, and Granting Application for Attorneys’ Fees (Aug. 7, 2013), ECF No. 3585.

269

The AAA’s Protocol states that: “Providers of goods and services should develop ADR programs which entail
reasonable cost to Consumers based on the circumstances of the dispute, including, among other things, the size and
nature of the claim, the nature of goods or services provided, and the ability of the Consumer to pay.” Principle 6,
AAA Consumer Due Process Protocol (Apr. 17, 1998) at 19, www.adr.org.
270

The revised fee schedule is relatively simple. At present, for cases subject to the Supplementary Procedures,
consumers pay a non-refundable $200 filing fee, regardless of who files the claim or of the amount at issue. In cases
before one arbitrator, the company pays a partially refundable filing fee of $1,500, which increases to $2,000 in cases
involving three arbitrators. The company also pays the arbitrator’s compensation, which will be $750 per arbitrator
for a desk arbitration, and $1,500 per arbitrator per day for arbitrations involving an in-person or telephonic hearing.
The arbitrator can reallocate initial filing fees, arbitrator compensation, or other expenses to the consumer (or, with
respect to the $200 consumer filing fee, to the company), but only pursuant to applicable law or the arbitrator’s
determination that a claim or counterclaim was filed for purposes of harassment or was patently frivolous. The
business pays a number of other fees, including a hearing fee and any hearing room rental fees. The consumer and

110

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

4.9.1

Initial fee schedule

Aside from attorneys’ fees, there were generally two kinds of fees associated with AAA consumer
arbitrations filed in the period 2010 through 2012. First, there were administrative fees payable
to the AAA itself. Second, there were arbitrator’s fees payable to the arbitrator.271 Some portion
of each of these fees was assessed at filing. (We refer to these as the “filing fees.”) During the
period of our review, all components of filing fees were either administrative fees or deposits
towards arbitrator fees.
For the AAA to administer an arbitration case filed in the period of our review, each side was
required to meet its initial fee requirements. The amounts and allocation of the filing fees during
the applicable period are set out below. These filing fees were based on the amount of
compensatory damages claimed or counterclaimed, not the amount of punitive damages,
attorney fees, or any other amounts. Note that these are not the AAA’s current filing fees for
consumer arbitrations, which were revised effective March 1, 2013, and therefore did not apply
to the cases that we reviewed.272
INITIAL FEES IN AAA CONSUMER ARBITRATIONS DURING RELEVANT PERIOD
Claims up to $75,000

Regardless of who filed the arbitration, the consumer’s initial filing requirements were up to
$125 for claims under $10,000 and up to $375 for claims between $10,000 and $75,000.273 This
payment was made as a deposit to cover the consumer’s share of the arbitrator fee; none of the
payment was for administrative fees. The consumer could be refunded some or all of the

company must each pay a $300 abeyance fee if a case is held as inactive for more than a year. Full details of the
current fee schedule are available on the AAA’s website, www.adr.org.
271

California requirements call for publication of data about the amount and allocation of arbitrator fees. See Cal.
Code Civ. Pro. 1281.96(a).
272

See supra note 270.

273

Counterclaims that were below $75,000 had no impact on the filing fees. Above $75,000, however, they would be
subject to the AAA Commercial Rules, as described below.

111

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

arbitrator’s deposit if it were not to be used—for example, if the case were to settle before an
arbitrator was appointed.274
The company incurred all administrative fees, which were $775 for cases up to $10,000 and
$975 for cases between $10,000 and $75,000. If a hearing were held, the company paid an
additional administrative fee of $200 for disputes up to $10,000, or $300 for disputes between
$10,000 and $75,000.275 The company also paid all remaining arbitrator deposits, which ranged
from $125 for a “desk arbitration” (meaning an arbitration that is resolved on the basis of
written submissions alone) or telephonic hearing to $625 for an in-person hearing.276 Again, for
claims up to $75,000, these fees did not change with the identity of the filing party.
The company could receive some refund of its administrative fees and arbitrator deposits.
Arbitrator deposits were refunded if not used. Administrative fees were refundable pursuant to
the Commercial Fee Schedule.277 This required that $350 in administrative fees were not
refundable to the business, but the remaining administrative fees would be refunded in full if the
case were settled or withdrawn within 5 days of filing. For cases settled or withdrawn between 6
and 30 days of filing, the refund was 50%. If the case settled or was withdrawn between 31 and
60 days of filing, the refund was 25%.278 If a consumer-filed case was withdrawn or settled
before the consumer paid his or her initial fee, the company might not incur any fees at all.279

274

See AAA, Supplementary Procedures C-8 (effective Jan. 1, 2010), “Fees and Deposits to be Paid by the Consumer.”

275

For claims under $10,000, the default rule is for the dispute to be resolved without a hearing. However, either
party may ask for a hearing, or the arbitrator may independently decide that one is necessary. See AAA
Supplementary Procedures C-5. Conversely, if the claim is for over $10,000, the default is for the dispute to be
resolved with a hearing. If both parties request no hearing, however, the arbitrator may waive the hearing
requirement. See id. C-6.
276

AAA, Supplementary Procedures (effective Jan. 1, 2010) C-8 “Fees and Deposits to be Paid by the Business.”

277

See id.

278

AAA, Commercial Arbitration Rules and Mediation Procedures (effective Jan. 1, 2010) Administrative Fee
Schedules.

279

One possible implication of this fee schedule is that it may have provided the consumer with certain forms of
leverage. By filing an arbitration in this time period, a consumer that was willing to incur $125 as an initial—though
potentially refundable—fee could cause the company to incur a contingent obligation of up to $1,725 for claims below

112

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Claims above $75,000

For consumer-filed claims above $75,000, the consumer paid all the administrative fees,
pursuant to the regular fee schedule in the AAA Commercial Rules, plus half the arbitrator’s fee
based on the arbitrator’s usual rates, with a deposit of half the arbitrator’s fee due on filing.280
(For a $100,000 claim, for example, the consumer’s initial administrative fees would be $1,850
with an additional administrative fee of $750 in the event a hearing is scheduled.281 Arbitrator
fees would be in addition to this amount.) The administrative fees were subject to refund per the
terms of the Commercial Fee Schedule discussed above.
Conversely, for company-filed claims above $75,000, the business paid the administrative fees
based on the regular fee schedule in the AAA Commercial Rules, plus half the arbitrator’s fee
based on the arbitrator’s usual rates, with a deposit of half the arbitrator’s fee due on filing. For
mutually submitted claims above $75,000, the parties were required to allocate the applicable
administrative fees between them.
Three-Arbitrator Panels

For the period in our review, the AAA charged a minimum fee of $2,800 when the arbitration
clause provided for a 3-arbitrator panel.282
WAIVERS AND ADJUSTMENTS
Although the applicable contract could reallocate initial fee burdens away from the consumer, it
could not add to the consumer’s initial fee burden without violating the applicable AAA rules,

$10,000, and up to $2,025 for claims between $10,000 and $75,000—albeit with the potentially significant caveat
that, depending on the terms of the arbitration clause, some or all these company fees ultimately could be reallocated
to the consumer at the discretion of the arbitrator. In addition, the consumer had some degree of control over the
applicable contingencies. Once he or she paid the initial deposit, the company would incur a partially refundable
administrative fee of either $775 or $975 and a refundable arbitrator fee of $125. Up until that point, however, the
company would not incur any fee obligation. Once an arbitrator had been appointed, the company would no longer be
able to recoup the arbitrator fee and it would face additional costs of between $825 and $925 if the consumer
successfully pressed for an in-person hearing.
280

AAA, Supplementary Procedures (effective Jan. 1, 2010) C-8 “Fees and Deposits to be Paid by the Consumer.”

281

See id.

282

See AAA, Commercial Rules (fees effective June 1, 2010) at 39. We identified two cases in which the AAA made an
initial fee assessment of $2,800 for a 3-arbitrator panel. In one case, the AAA assigned the company the full
obligation. In the other, the AAA did not assign the obligation, but sought payment from the parties collectively.

113

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

including the Due Process Protocol. Some contracts, therefore, allowed for an advancement of
the consumers’ fees.283
The consumer could apply for a waiver of otherwise applicable fees.284 In the period of our
review, the AAA allowed consumers to apply for a hardship waiver or deferral of any applicable
administration fee if their gross annual income was below 200% of the federal poverty
guidelines. The AAA rules also recognize California law providing for a waiver of arbitration of
fees and costs, exclusive of arbitrator fees, for consumers with a gross monthly income of less
than 300% of the federal poverty guidelines.285 For claims of any value, the consumer was also
able to request an arbitrator willing to serve pro bono, although the rules did not guarantee that
one would be provided.
In addition, as noted above, the arbitrator had discretion to reallocate administrative or
arbitrator fees in the award, as he or she might deem appropriate.286 Unless the contract or
applicable law required otherwise, the arbitrator might also reallocate attorneys’ fees. The AAA
fee schedule did not require any specific allocation of attorneys’ fees.

4.9.2

Data

For the three product markets covered, Figures 30 and 31 show the fee amounts assessed at the
front end of the AAA arbitrations from 2010 through 2012. These initial assessments are
typically set out in AAA correspondence to the parties.
We have not determined if the parties paid the initial assessments reflected below. For example,
it is possible that claims may have withdrawn or settled before the payment of all these initial

283

See section 3.4.10.

284

See AAA, Commercial Arbitration Rules R-49 (effective June 1, 2009, through October 1, 2013).

285

See CAL. CODE CIV. P. 1284.3 (cited in AAA, Supplementary Procedures C-8). The California waiver applies to all
consumer agreements subject to the California Arbitration Act and to all consumer arbitrations conducted in
California.
286

See Drahozal & Zyontz, Empirical Study, supra note 140, at 887-88 n.165; AAA, Commercial Arbitration Rules,
Administrative Fee Schedule.

114

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

fees. In addition, we have not determined how fees were finally allocated in these cases.287 As
discussed above, during the period of our review, the arbitrator had discretion to reallocate fees
in the award, and final fee allocations may also address fee advances made at the front-end of
the process. The amounts shown aggregate initial fees for each covered dispute. Thus, they may
include administrative fees, arbitrator deposits, and any additional initial fees assessed to each
party.288

FIGURE 30:

DISTRIBUTION OF INITIAL CONSUMER FEE ASSESSMENTS 2010-12

Figure 30 illustrates the initial distribution of consumer fee assessments. The figure does not
include information about 161 disputes for which we were unable to identify information about

287

We plan to cover the ultimate distribution of fees in the next phase of our work.

288

For example, an entry reflecting $1,100 in company fees may represent: administrative fees of $775 plus $125 for
the arbitrator deposit plus $200 for a live hearing; or administrative fees of $975 plus $125 for the arbitrator deposit.

115

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

the fees assessed to the consumer.289 It does include entries for 320 filings involving fee advance
requests, nine involving AAA hardship requests, and 15 involving California waiver requests.290

FIGURE 31:

DISTRIBUTION OF INITIAL COMPANY FEE ASSESSMENTS 2010-12

As noted, we have not completed our assessment of fees incurred during this period. Although
consumer obligations would increase markedly above $75,000, based on the schedule outlined,
and consistent with the data shown, at lower claim amounts the consumer’s incurred obligation
was significantly smaller, with the company bearing the predominant share of initial fee

289

In 129 of these 161 disputes, we were unable to identify specific initial fee assessments for either party. In the
remaining 32 cases, we were able to ascertain that we were missing information on initial consumer fee assessments.
We excluded data for one case in which the AAA sought payment of a $2,800 initial fee for a three-arbitrator panel
from the parties collectively, without specifically assigning this initial fee to one side or the other.
290

116

These various fee advance and waiver requests overlap in part.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

obligations.291 Figure 31 illustrates the initial distribution of fees assessed to companies. It does
not include information relating to 229 proceedings for which we were unable to locate
information about the fees assessed to the company.292 It does include information relating to
317 fee advance requests, eight AAA Hardship Requests, and 15 California Waiver requests.
Finally, we also recorded the number of instances in which the AAA provided formal notice to
the parties that it declined to administer the arbitration because of the company’s failure to pay
required fees or deposits. This formal notice also states that because the company has failed to
comply with AAA policy on consumer arbitrations, the AAA will not administer further disputes
concerning that company. The notice requests that the company remove the AAA from its
arbitration clause. In the AAA Case Data, we observed 29 instances of companies receiving this
letter. All but six concerned credit card disputes. Of the 29, all but one were consumer-filed
disputes; the final dispute was a mutual submission.

291

Filing fees are generally $400 for United States District Courts. State court filing fees are more varied. One recent
summary, “Civil Filing Fees in State Trial Courts April 2012,” is available at www.ncsc.org, the website of the National
Center for State Courts. That summary indicates that filing fees are under $300 (and often significantly so) in almost
all state trial courts. Small claims court fees are lower, typically on the order of $20 to $75, though in a few states
exceed $100 for larger claim amounts. See id.
292

This excludes the same 129 cases for which we have no initial fee information, see supra note 289, as well as 100
cases for which we could ascertain that we were missing company initial fee information. It excludes the one case
where the AAA sought collective payment of a $2,800 initial fee associated with a three-arbitrator panel. See supra
note 289.

117

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

5. Small claims court data
This section presents data from our review of small claims court filings. Our analysis draws on a
number of different state and county small claims court databases. We reviewed these for suits
filed against and by ten large credit card issuers, which collectively account for a very significant
portion of the consumer credit card market.
We reviewed small claims court data for two reasons. First, as discussed in section 3.4.2, the
clear majority of the arbitration clauses within our review, specifically recognize—and allow—
access to small claims court as an alternative to arbitration.293 In addition, small claims courts
are reported to handle a significant share of some disputes about consumer financial
products.294 Furthermore, small claims courts are generally considered faster, cheaper, and
simpler than ordinary trial courts of plenary jurisdiction, and are intended to be easier for
individuals to pursue claims without using a lawyer.295 (In fact, some small claims court

293

Industry groups also urged the Bureau to study small claims courts, especially in the context of small claims court
carve-outs to arbitration clauses. See, e.g., Chamber of Commerce RFI Comment at 12-13 (“The study should compare
the features of different types of arbitral and litigation forums, including . . . small claims courts.”).
294

See Richard M. Hynes, Broke But Not Bankrupt: Consumer Debt Collection, 60 FLA. L. REV. 1, 27-28 (2008)
(“[M]ost unsecured consumer debts fit comfortably within the jurisdiction of the limited-jurisdiction courts, and the
overwhelming majority of suits are filed in these courts when they are available.”); see id. at 28 (noting that “the
exclusion of courts of limited jurisdiction renders most prior studies inapplicable to the questions of how much
consumer debt collection litigation exists and how it has changed over time.”).
295

See Suzanne E. Elwell & Christopher D. Carlson, The Iowa Small Claims Court: An Empirical Analysis, 75 IOWA L.
REV. 433 (1990); John C. Ruhnka & Steven Weller, Small Claims Courts: A National Examination at 2-3 (Nat’l Ctr. for
State Courts 1978) (noting broad consensus as to the goals of the small claims court process, including “accessibility,
speed, low cost, simplicity, self-representation, fairness, [and] effectiveness.”); John A. Goerdt, Small Claims & Traffic
Courts: Case Management Procedures, Case Characteristics, and Outcomes in 12 Urban Jurisdictions (Nat’l Ctr. for
State Courts ed. 1992).

118

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

jurisdictions generally require parties to proceed without representation.296) We, therefore,
wanted to see what use parties made of them with respect to consumer financial disputes.
Second, some commenters argue that the presence of small claims court carve-outs in
arbitration clauses alleviates any need for any other litigation method to resolve small-dollar
disputes.297 The argument is that small claims courts are “consumer friendly”298 so that having a
small claims court carve-out helps endow an arbitration clause with “fundamental fairness.”299
Both the AAA’s Due Process Protocol and the JAMS Minimum Standards of Procedural
Fairness, which are applicable to most consumer arbitrations, require that pre-dispute
arbitration clauses permit access to small claims courts.300

296

See, e.g., California Department of Consumer Affairs, Basic Considerations and Questions: What Is Small Claims
Court?, at http://www.dca.ca.gov/publications/small_claims/basic_info.shtml (visited Nov. 20, 2013) (“In most
situations, parties to a small claims action must represent themselves. As a general rule, attorneys or non-attorney
representatives (such as debt collection agencies or insurance companies) may not represent you in small claims court.
Self-representation is usually required. There are, however, several exceptions to this general rule: If the court
determines that a party is unable to properly present his or her claim or defense for any reason, the court may allow
another individual to assist that party.”).

297

See Peter Rutledge, Whither Arbitration, 6 GEORGETOWN J.L. & PUB. POL’Y 549, 570 (2008) (asserting that the
argument that a mandatory arbitration provision combined with a class action waiver denies consumers access to
justice is exaggerated by noting that, amongst other things, most arbitration clauses provide access to small claims
court); see also AFSA RFI Comment at 8 (“[M]any arbitration agreements will allow consumers to bring claims
otherwise subject to arbitration in small claims court. Thus, for smaller claims, consumers frequently have a choice
between arbitration and small claims court. This context is important in order to understand the dynamics of the
claims that are (and are not) asserted in arbitration versus other methods of resolving disputes between consumers
and covered persons.”).
298

ABA/CBA/FSR RFI Comment at 14.

299

ABA/CBA/FSR RFI Comment at 6. See also Mark Furletti, Mandatory Arbitration Provisions in the Credit Card
Industry at 5-6 (Jan. 2003) http://www.phil.frb.org/consumer-credit-and-payments/payment-cardscenter/events/workshops/2003/MandatoryArbitrationClauses_012003.pdf. Dean Roscoe Pound of Harvard Law
School is often credited as being the first to articulate a vision of a court to “‘provide for disposing quickly,
inexpensively, and justly of the litigation of the poor, for the collection of debts in a shifting population, and for the
great volume of small controversies which a busy, crowded population, diversified in race and language necessarily
engenders.’” Commonwealth of Massachusetts, DISTRICT COURT DEPARTMENT OF THE TRIAL COURT, REPORT OF THE
SMALL CLAIMS WORKING GROUP (Aug. 1, 2007) at 3, http://www.lawlib.state.ma.us/docs/smallclaimreport.pdf (citing
Roscoe Pound, The Administration of Justice in the Modern City, 26 HARV. L. REV. 302, 315 (1912-1913)).
300

The AAA’s Protocol states that: “Consumer ADR Agreements should make it clear that all parties retain the right
to seek relief in a small claims court for disputes or claims within the scope of its jurisdiction.” Principle 5, Consumer

119

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

5.1

Previous studies

We have not identified significant numbers of recent empirical studies on the incidence and
nature of consumer financial claims in small claims courts.301 Two older studies cover small
claims courts in multiple jurisdictions, but neither report on the incidence of consumer financial
disputes.302 Some studies do suggest the companies make more use of small claims court than
consumers do, but these are not specific to consumer financial products. Goerdt’s 1992 study of
12 urban small claims courts showed that business-initiated disputes against individuals made
up 53% of all cases studied; by contrast, individual-initiated disputes against businesses made
up 13% of the sample.303 Similarly, in a 1990 study, Elwell & Carlson found that in the Iowa
small claims court system many more businesses sued individuals than individuals sued
businesses. 304 In their sample of 1,802 disputes, individuals filed against companies in 4% of the

Due Process Protocol. The JAMS’ Minimum Standards state that: “[N]o party shall be precluded from seeking
remedies in small claims court for disputes or claims within the scope of its jurisdiction.” Minimum Standard No. 1,
JAMS Minimum Standards. Explaining the basis for the carve-out, the Reporter for the AAA’s Protocol indicated that
“[t]he Advisory Committee concluded that access to small claims tribunals is an important right of Consumers which
should not be waived by a pre-dispute ADR Agreement.” AAA, Consumer Due Process Protocol at 18 (Apr. 17, 1998),
available at www.adr.org.
301

Using data from 2007, Drahozal & Zyontz randomly sampled 500 cases filed in Oklahoma small claims court,
finding 336 disputes related to consumer debt cases, of which “330 were brought by creditors seeking to recover
unpaid debts [and] only six were brought by consumers against businesses.” Drahozal & Zyontz, Creditor Claims,
supra note 140, at 88 (analyzing sample of cases closed between March 31, 2007, and January 1, 2008). Of the six
consumer-initiated claims in Drahozal & Zyontz’s sample, just one dispute involved a credit card. This same work
cites to a number of older studies of small claims court proceedings that focus primarily on debt collection issues,
including default incidence. See id. at 82 n.18.
302

See, e.g., Ruhnka & Weller, supra note 295, at 49-50 (allocating the subject matter of disputes in small claims
courts in 15 cities into just five categories: “Consumer Plaintiff”, “Seller Plaintiff”, “Property Damage”, “LandlordTenant,” and “Other”); Goerdt, supra note 295, at 7 (analyzing small claims court data from the cities of Cambridge,
Denver, Des Moines, Fairfax, Hartford, Minneapolis, Portland, Sacramento, San Diego, Seattle, Washington, and
Wichita); id. at 47-48 (noting incidence of general categories of claims including, “consumer complaints” “debt
collection, “torts,” “housing,” and “other”).
303

Goerdt, supra note 295, at 7 (analyzing small claims court data from Cambridge, Denver, Des Moines, Fairfax,
Hartford, Minneapolis, Portland, Sacramento, San Diego, Seattle, Washington, and Wichita); id. at 43-44 (showing
filing rates by individuals and businesses).

304

120

Elwell & Carlson, Iowa Small Claims Court, supra note 295, at 487.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

cases, and businesses filed against individuals in 47% of the cases.305 The significant use of small
claims court for debt collection is well documented.306

5.2

Data sources

No centralized, comprehensive and searchable source of small claims court cases or dockets
exists.307 Documents related to small claims cases are generally only available on databases
maintained by the administrative staffs of the small claims courts themselves (or by physically
requesting the materials from the clerk’s office). These databases appear to have been designed
with individual litigants in mind. Working with these databases for the purpose of collecting
large volumes of data is painstaking. This may be why there is such a dearth of empirical data
and academic literature on the topic—or why such studies tend to look at fragmentary data.
Our main sources of data were online small claims court databases for states that offered free or
reasonable cost access to case information. In addition to these cost criteria, we used three
searchability criteria for the state court databases. To be included, such databases must: (1)
purport to provide statewide data; (2) permit searches by party name (or ready identification by
party if keyword searches are not possible); and (3) allow for by-year date sorting. Databases for
13 state jurisdictions and the District of Columbia met these criteria. We refer to this as the
“state-level sample.” We identified 12 states via this method, plus the District of Columbia. We
also added New York because with the exception of New York City—for which we were able to
obtain data by other means, as described further below—it otherwise met these same criteria.
The 14 jurisdictions in the state-level sample, excluding New York City, cover approximately 52
million people of all ages.

305

See id.

306

In 2007, a working group of Massachusetts trial courts judges and administrators “recognized that a significant
portion of small claims cases involve the collection of commercial debts from defendants who are not represented by
counsel.” Commonwealth of Massachusetts, District Court Department of the Trial Court, Report of the Small Claims
Working Group (Aug. 1, 2007) at 3, http://www.lawlib.state.ma.us/docs/smallclaimreport.pdf.

307

There is no analog to the PACER system for federal courts. Westlaw and LEXIS provide limited small claims court
data, mainly dockets, for certain jurisdictions. We used these to check our primary research. As of our search on
October 16, 2013, Westlaw and LEXIS provided only a limited number of small claims court dockets, from one entire
state (Wisconsin) and 9 counties out of the top 30 across two different states, California and Florida.

121

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

To supplement the state-level sample, we also used county-level data. We selected the 30 most
populous counties in the United States, which largely included areas not in the state-level
sample. We then reviewed all of the databases in that sample of 30 that met the same cost and
equivalent searchability criteria as for the state-level sample: free or reasonably priced;
searchable county-wide; searchable by party name; and restrictable by date. We refer to this as
the “county-level sample.” Seventeen of the 30 counties met these criteria.308 These 17 counties
include jurisdictions covering approximately 35 million people of any age. We are not aware of
another study with coverage as broad as our combined county-level and state-level samples.309
In all these jurisdictions, we looked for potential credit card cases involving a set of 10 large
credit card issuers. Given the relative concentration of the consumer credit card market, those
credit card issuers cover a predominant share of that market.310 In addition to covering very
large players, our sample covers companies with small claims court carve-outs for both
consumer and company claims (Citibank, Wells Fargo, USAA, and US Bank); non-mutual small
claims court carve-outs, meaning that only consumers may require that a dispute be brought in
small claims court (American Express and Discover); no small claims court carve-out at all
(Fifth Third); and with no arbitration clause in 2012 (Bank of America, JPMorgan Chase, and
Capital One).311

308

These added partial geographic coverage in six states not covered under our state-level review—California, Florida,
Nevada, Pennsylvania, Texas, Washington—and coverage of a number of major cities or urban areas, including
Houston, Philadelphia, and Silicon Valley. Three New York counties met these criteria as well; we included those
results under the county search, not the New York state-level search.

309

Under the auspices of the National Coalition on State Courts, Ruhnka & Weller conducted a study in 1978 of 15
separate jurisdictions, mainly urban small courts. See generally Ruhnka & Weller, supra note 295. In 1992, also
working with the National Center for State Courts, Goerdt conducted an analysis of 12 small claims and traffic courts
in urban jurisdictions. See generally Goerdt, supra note 295.

310

As of year-end 2012, our sample covers $564.75 billion in credit card outstandings, which represents a
predominant share of the consumer credit card market. Our sample for the small claims court analysis covers 84% of
the outstandings covered by the credit card contract sample used for section 3. Concentration at the top-end of the
credit card market has remained roughly the same over the last decade. Compare The Nilson Report, Issue #1,012 at 1,
8 (Feb. 2013) (as of 2012, top 10 accounted for 85% of market) with Hynes, supra note 294, at 51 (in 2004, top-10
accounted for almost 90% of market).
311

A small claims court carve-out is not necessary for either party to bring a claim in small claims court. Very few predispute clauses require arbitration of all disputes. Instead, the clauses enable either party to invoke arbitration

122

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

As the issuers in our sample cover such a large share of credit card loans outstanding, we should
uncover most consumer-issuer disputes involving credit card accounts in the small claims courts
we selected. We will only uncover suits that can be filed in small claims court. The small claims
courts covered by our review use a range of jurisdictional limits on the amount that can be
claimed in small claims court. These limits—which are generally between $2,500 and $15,000—
are laid out in detail in Appendix E.
To identify consumer credit card suits, we used small claims court dockets to identify cases filed
by individuals312 against specific issuers and against colloquial names for these entities (such as
“Bank of America” or “Chase”).313 By attempting to capture all credit card suits filed by
consumers against these issuers, our search will tend to be over-inclusive. It will include, for
example, some consumer cases that are not about credit cards.314 It should not, however, under
count consumer credit card cases against these issuers.315 Our consumer-filed case numbers,
therefore, can be seen as an outer limit on the number of consumer credit card cases against
these issuers.
To identify suits that credit card issuers filed against consumers, we focused on suits in which
the docket listed: (1) the correct legal name of one of the specific credit-card issuing

unilaterally. A small claims court carve-out, therefore, simply immunizes a small claims court filing from this
potential effect of the arbitration clause. Entities with a clause but no carve-out, therefore, may still be sued or sue in
small claims court. It is simply that the other party could then invoke arbitration, although that does not mean that
the other party will do so.
312

We excluded cases filed by non-natural persons such as corporations or partnerships.

313

We included potentially common misspellings, and alternative arrangements of character strings. We excluded
suits by individuals against entities within the corporate family that included in their name terms such as “mortgage,”
“home loan,” “auto,” and “insurance.” We would not expect such suits to be credit card cases.
314

For example, we may capture consumer cases about checking accounts (at least if there are any in small claims
court). As discussed in Appendix E, our searches using company names that a consumer might associate with our ten
issuers will also uncover cases against eight of the largest U.S. retail banks.
315

Our data do not capture credit card-related claims by these issuers’ cardholders against other parties besides the
issuer. So we are not capturing claims against private label partners. It is possible that consumers intending to sue
their “credit card company” sue such partners instead. For Philadelphia County, where detailed data are available, we
did look for suits against the major retail partners of an issuer in our sample that has a large private label business.
We did not find any such suits. In addition, our data also will not capture claims against (or by) debt collection
companies that act for issuers or purchase debt from them.

123

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

subsidiaries316 in our review; and (2) an individual as the defendant. This search will tend to
over count such suits to the extent that any of the credit card issuers: (1) also provide consumers
with products other than credit cards; and (2) that non-credit card activity causes companies to
sue consumers in small claims court. However, our review of public data suggests that this effect
should not be significant as to most of the issuers.317
Further details about our sources and methodology are included in Appendix E.

5.3

Incidence

We start, first, with data from Philadelphia County and Alameda County. These were the two
jurisdictions in our defined samples for which we were able to obtain underlying case
documents on a systematic basis. These allowed us to establish, definitively, the nature of the
claims at issue.
Using the broad case identification method outlined above, we identified for Philadelphia some
2,245 cases filed by issuers in our sample. (We identified no such cases in Alameda. In that
county and in the rest of California, companies face severe limits on bringing collections claims

316

Thus, if there was ambiguity in the name of the company claiming (e.g., the docket identified the plaintiff only as
“Chase” and not “Chase Bank USA, National Association”), we excluded the dispute from the company-filed results.
Although we consider it reasonable to assume that outside or in-house counsel will ensure that companies sue using
their correct legal name, to the extent that is not the case for credit card company suits, we will undercount such suits.

317

According to call reports, at year end 2012, three issuers in our sample—American Express, Discover, and USAA—
had no consumer loans other than credit card loans. Based on call reports, the other issuers we studied generally had
as their primary business large consumer loans, such as mortgages, or smaller-dollar loans in the form of credit card
borrowing. Our results from Philadelphia (see section 5.3) suggest that companies are unlikely to file mortgagerelated claims in small claims court. Leaving aside mortgage, therefore, call reports indicate that most of the
remaining issuers had a low volume of non-mortgage consumer loans compared to their credit card holdings. This is
true for Citibank, National Association ($106.7 billion in credit card loans and $894 million in other consumer loans),
Chase Bank USA, National Association ($93.3 billion in credit card loans and $616.3 million in other consumer loans),
Capital One ($77.8 billion in credit card loans and $277.5 million in other consumer loans), and JPMorgan Chase,
National Association ($21.5 billion in credit card loans and $3 billion in other consumer loans). Three of our ten
issuers, however, had a significant volume of consumer loans other than credit cards: Bank of America, National
Association and FIA Card Services, National Association ($94.8 billion in credit card loans, $26 billion in auto loans,
and $22.8 billion in other consumer loans), Wells Fargo Bank, National Association ($16.3 billion in credit card loans,
$12.5 billion in auto loans, and $8.9 billion in other loans) and Fifth Third Bank ($2.1 billion in credit card loans,
$11.3 billion in auto loans, and $445 million in other loans).

124

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

in small claims court.318) By using the detailed review of actual pleadings available in
Philadelphia, we were then able to establish that all but one of the company-filed cases were, in
fact, credit card debt collection cases. This suggests that our broad methodology provides a
reasonable, even close, approximation of issuer use of small claims court for credit card disputes
against consumers.
Our broad case identification method also identified four Philadelphia cases and 39 Alameda
cases as consumer-filed credit card cases against the issuers in our sample. When we reviewed
the actual pleadings in these cases, however, none of the four Philadelphia cases involved an
individual filing credit card claims against one of the ten issuers.319 Reviewing the pleadings in
the 39 Alameda cases, we identified only four that were clearly individuals filing credit card
claims against one of the ten issuers.320 This suggests that our broad methodology may well
overstate the actual number of small claims court cases filed by credit card consumers against
our sample of issuers.
Even using the broad methodology, however, we see relatively low outer limit estimates for
consumer-filed credit card cases. For our 31 jurisdictions combined, we were able to estimate an
outer limit of 870 such cases for all of 2012.321 We provide the detailed results of these searches
in Table 10 in Appendix E. In only three jurisdictions was our outer limit estimate higher than
70 cases against all issuers combined, and the outer limit for any one issuer was 245 across all
jurisdictions combined. The outer limit of 36 cases filed against one issuer in Orange County,
California was the maximum for any one issuer and jurisdiction.

318

See infra note 322.

319

The four consumer-initiated cases involved (1) a stolen debit card; (2) certain mortgage payments not recognized
as timely by the bank holding the mortgage; (3) an insurance claim; and (4) a tort that occurred in a bank (a bicycle
hit a customer).
320

Thirty-two disputes manifestly did not involve credit cards. These were 17 disputes involving checking accounts;
ten alleging interference with real estate listings; four about home loans; and one about insurance. Another three
disputes might conceivably have involved credit cards, but the record was insufficient to show that they actually did so.
One of these three cases did not clearly identify any product, and merely alleged in a single sentence FCRA violations
by the issuer. The other two concerned unspecified “lines of credit.”
321

This adjusts for the actual count in Philadelphia and Alameda counties.

125

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Claims filed by credit card issuers against individuals show a different pattern. We report this
data in Table 9 below. Claim numbers are either substantially lower or higher than the numbers
of consumer-filed claims. The low numbers in Table 9 predominantly correspond to
jurisdictions that impose substantial limitations on the use of small claims courts by businesses.
These include California (here represented by Orange, Riverside, San Bernardino, Santa Clara,
and Alameda Counties)322 and New York.323 Other jurisdictions have certain procedures that,
while not targeted explicitly at businesses, may dampen the number of company-initiated
claims. For instance, in Utah, the clerk of court or a judge is empowered to remove multiple
disputes filed by a same plaintiff from small claims court to the district court.324 In King County

322

In California, corporations and partnerships may bring no claim larger than $5,000 in small claims court, and no
more than two such claims of more than $2,500 per calendar year. See California Department of Consumer Affairs,
Basic Considerations and Questions: What Is Small Claims Court? (last visited Oct. 25, 2013),
http://www.dca.ca.gov/publications/small_claims/basic_info.shtml (“Corporations, partnerships, unincorporated
associations, governmental bodies, and other legal entities cannot claim more than $5,000. Also, no claimant (natural
person or legal entity) may file more than two small claims court actions for more than $2,500 anywhere in the state
during any calendar year.”).
323

New York imposes various restrictions on legal entities filing in small claims court. A corporation or partnership
may only bring a claim in small claims court if it has its principal office in New York state. Further, a corporation or
partnership must submit a demand letter before filing a claim, and can file no more than 5 claims per calendar month.
See Access to Justice NY Courts, Your Guide to Small Claims & Commercial Small Claims,
http://www.nycourts.gov/COURTS/nyc/civil/pdfs/smallclaims.pdf. But the claims that corporations or partnerships
cannot bring in New York small claims court do not disappear. Rather, it appears that they bring their would-be small
claims disputes instead in Civil Court, the next court of limited jurisdiction. Of the 523,186 cases filed in Civil Court in
the Bronx, Kings, Queens and Richmond Counties in 2006, 53% of these cases were categorized by one study as
“consumer credit litigation.” See Urban Justice, Debt Weight: The Consumer Credit Crisis in New York City and its
Impact on the Working Poor 8 (Oct. 2007). Similarly, over half of the 320,000 cases filed in New York County Civil
Court involved consumer credit. See id. (“More and more, New York City Civil Court is becoming a ‘credit card court,’
with over 50% of cases filed in that court arising out of ‘consumer credit transactions.’”) Based on a limited sample,
the study estimated that 10.7% of “consumer credit litigation” cases were, in turn, initiated by the original creditors,
including credit card issuers. Further, many if not most of these claims would have been under the small claims
jurisdictional limit of $5,000. See id. 14-16.
324

Utah Code Title 78A Chapter 8 section 102 (“If a person or corporation other than a municipality or a political
subdivision of the state files multiple small claims in any one court, the clerk or judge of the court may remove all but
the initial claim from the court's calendar in order to dispose of all other small claims matters. Claims so removed
shall be rescheduled as permitted by the court's calendar.”).

126

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

(WA), attorneys and paralegals are not allowed to appear for either party without the judge’s
permission.325
TABLE 9:

CREDIT CARD ISSUER SMALL CLAIMS COURT SUITS AGAINST INDIVIDUALS BY JURISDUCTION
2012

Citi

Wells

USB

USAA

AmX

Dis

5/3

BoA

Cap1

JPM

Total

0

10

0

0

0

0

0

0

0

0

10

262

8

0

0

91

718

0

301

4,274

0

5,654

0

0

1

0

22

444

0

0

0

0

467

76

1

0

0

22

115

0

21

865

0

1,100

Iowa

246

33

0

0

4

501

1

94

260

0

1,139

Minnesota

147

0

0

0

1

114

0

0

2,029

0

2,291

New Mexico

0

2

0

0

29

386

0

2

2

0

421

New Jersey

0

0

0

0

0

0

3

0

0

0

3

New York, ex-NYC

0

0

0

0

0

0

0

0

0

3

3

North Dakota

0

0

0

0

0

0

0

0

0

0

0

Oklahoma

0

4

0

0

0

0

*

0

0

0

4

Oregon

0

0

0

0

0

0

0

0

0

0

0

Utah

0

0

0

0

0

0

0

0

0

0

0

704

96

48

0

76

2,162

27

1

9,361

14

12,489

0

0

0

0

0

0

0

0

0

0

0

265

74

0

0

254

227

2

103

2,647

0

3,572

Clark (NV)

0

0

0

0

0

0

0

0

0

0

0

Harris (TX)

614

356

47

0

319

380

0

1,051

5,046

92

7,905

70

0

1

0

67

125

1

0

1,325

57

1,646

0

0

0

0

0

0

0

0

0

0

0

Alaska
Connecticut*
Delaware
District of Columbia

Wisconsin
Alameda (CA)
Broward (FL)

Hillsborough (FL)
King (WA)

325

See King County District Court Services, Information about Small Claims,
http://kingcounty.gov/courts/DistrictCourt/Smallclaims.aspx.

127

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Kings (NY)

0

0

0

0

0

0

0

0

0

0

0

New York (NY)

0

0

0

0

0

0

0

0

0

0

0

Orange (CA)

0

0

0

0

0

0

0

0

0

0

0

Palm Beach (FL)

174

15

0

0

177

306

0

86

1,594

2

2,354

Philadelphia (PA)

79

0

0

0

46

743

0

165

1,212

0

2,245

Queens (NY)

0

0

0

0

0

0

0

0

0

0

0

Riverside (CA)

0

0

0

0

0

0

0

0

0

0

0

Sacramento (CA)

0

0

0

0

0

0

0

0

0

0

0

San Bernardino (CA)

0

0

0

0

0

0

0

0

0

0

0

San Diego (CA)

0

0

0

0

0

0

0

0

0

0

0

Santa Clara (CA)

0

0

0

0

0

0

0

0

0

0

0

2,637

599

97

0

1,108

6,221

34

1,824

28,615

168

41,303

Totals

*Note: Search results for Capital One in Connecticut exceeded 4,274, but further information beyond the 4,274 could
not be retrieved from that database.

The higher numbers in our sample are present in jurisdictions that do not have similar limits on
company filings. In addition, two issuers accounted for the vast majority of company-filed
cases.326 This may be attributable to the fact that these two issuers are in the subprime credit
card business, where smaller credit lines are the norm. It may also reflect other issuers making
greater use of debt buyers or collection agencies. Other studies have also found relatively
concentrated use by particular issuers of small claims court.327 Appendix E depicts these
numbers relative to our outer-limit estimates for consumer-filed cases in different jurisdictions.

326

One of these issuers uses a non-mutual small claims court carve-out, which applies only to guarantee the
consumer’s right to remain in small claims court. The lack of mutuality in a clause, therefore, may not have any
impact on company use of small claims court.

327

Beth Healy, Debtor’s Hell, Part 2: A System Compromised, The Boston Globe (July 31, 2006),
http://www.boston.com/news/special/spotlight_debt/part2/page4.html.

128

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

6. Future work
In addition to the work laid out above, the Bureau has a number of phases of work that are
underway or that are under consideration for inclusion in the statutory report to Congress. We
discuss each of these below.


We will finalize our analysis of arbitration clause incidence and features. This may
expand to cover some other consumer financial product markets.



Subject to OMB approval, we plan to conduct a survey of consumers that addresses
consumer awareness of arbitration clauses and consumer perceptions of and
expectations about formal dispute resolution. The survey may also explore the role of
formal dispute resolution terms in consumer product choice. (We have already published
one detailed proposal in this respect, and have retained survey contractors. We have
received a number of comments and plan to carry out focus group work. The proposal
will be revised in light of the comments received, feedback from the survey contractors,
and any focus group input. We will submit a revised proposal for additional public
comment before we finalize the proposed survey.)



We may review arbitration case record data for filings concerning disputes about other
consumer financial products or services, such as private student loans.



We will continue our review of litigation filings concerning consumer financial products.
This may involve additional research into claims filed in small claims court. We will also
review available federal court filings for select products, and, to the extent reasonably
available from electronic sources, state court filings as well. We may use some sampling
to control for case numbers. As with the arbitration filings work, this is intended to bring
to light the types and frequency of claims that consumers bring to this method of formal
dispute resolution. Viewed in conjunction with the same analysis for arbitration filings,
this work may also help us to discern the impact of arbitration clauses on the incidence

129

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

and nature of consumer claims, although we cannot, by definition, directly observe
claims that consumers may not have pursued at all because of the absence of a particular
mechanism of formal dispute resolution.


Once we have completed our front-end identification of consumer financial cases, we will
consider how—if at all—we might meaningfully compare the disposition of cases across
arbitration and litigation (including class litigation), both in terms of substantive
outcome and in terms of procedural variables like speed to resolution. We recognize that
it may be hard to identify comparable cases across these different alternatives, not least
because cases are not randomly assigned to the different streams. In addition, numerous
factors will affect variables like speed to resolution and substantive outcome that are
independent of the formal dispute resolution mechanism used.



As part of our outcome analysis, we are reviewing court records to assess the
circumstances under which companies invoke arbitration clauses in response to
consumer claims, and how that invocation may impact the outcome of such claims. One
way we are exploring this issue is to try to identify consumer challenges that have been
raised with respect to particular industry practices (or alleged practices) and then assess
how dispute outcomes are impacted, if at all, by the presence of arbitration clauses and
the availability of arbitration.



We will use public court records to try to assess consumer benefits and transaction costs
in consumer class actions involving consumer financial services. We will also attempt to
evaluate whether class actions exert improper pressure on defendants to settle meritless
claims. We intend to look at the outcomes of filed class actions, including settlements,
and on dispositive and certification motion practices.



We intend to assess the possible impact of arbitration clauses on the price of consumer
financial products.



We are also examining the interrelationship between public enforcement and private
aggregate enforcement. Although some empirical research compares public enforcement
and private class actions in securities and antitrust matters, we have not located
empirical studies that compare these in the context of consumer law. Accordingly, we are
conducting an empirical analysis in this area. It will consider the types of cases brought

130

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

by public and private actors, and the relationship between any actions against the same
defendants or challenging similar conduct.

131

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

APPENDIX A

Clause incidence methodology
This appendix sets out the methodology used in collecting the data described in section 3.
CREDIT CARD AGREEMENTS
The sample of credit card agreements studied consists of credit card issuers with agreements on
file with the CFPB as of December 31, 2010, 2011, and 2012. The Credit CARD Act of 2009
requires credit card issuers above a de minimus size to post their credit card agreements on the
issuer’s web page and to file the agreement with the appropriate regulator.328 Initially, the Credit
CARD Act required the agreements to be filed with the Federal Reserve.329 The Dodd-Frank Act
changed the requirement so that credit card agreements now must be filed with the CFPB.330
Issuers are required to update their filing only when it changes.331 Some issuers have never filed
new agreements, so the agreement they have on file has not changed since 2009. Other issuers
file new agreements every quarter; for those issuers, the agreement studied is the agreement on
file as of December 31, 2012. Many issuers have multiple agreements on file with the CFPB. In
almost every case, the dispute resolution clause in each agreement filed by the issuer is identical.
In the rare case in which an issuer has multiple agreements, the most frequently used type of
agreement is included in the sample. In a handful of cases, the issuer filed some document other
than the cardholder agreement with the Bureau. In those cases, the cardholder agreement was
obtained from the issuer’s website when possible.

328

Credit Card Accountability Responsibility and Disclosure Act of 2009, Pub. L. No. 111-24, 123 Stat. 1746, § 204(a)
(May 22, 2009); see 15 C.F.R. § 226.58(c)(5) (de minimus exception).

329

Credit CARD Act, § 204(a).

330

See 15 U.S.C. § 1632(d).

331

15 C.F.R. § 226.58(c)(3).

132

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Each credit card agreement in the sample was reviewed to determine if it included an arbitration
clause.332 If it did, the provisions of the arbitration agreement were coded for the information
described in section 3. One issuer’s agreement, while indicating that it included an arbitration
clause, incorporated by reference the details of the arbitration clause included in another
document, which was unavailable. That issuer’s agreement was coded as providing for
arbitration, but the details of the arbitration provision were coded as missing.
Data on credit card loans outstanding from the issuer’s December 31, 201o, December 31, 2011,
and December 31, 2012 call reports were used to measure the relative market share of each
issuer. Only domestic credit card loans outstanding were used, plus securitized credit card loans,
if any. Data for related issuers (i.e., those that are subsidiaries of a common parent) were
consolidated as long as each issuer used the same dispute resolution clause in its credit card
agreement. If the issuers had different dispute resolution clauses, their data were not
consolidated and each issuer was included separately in the sample.
CHECKING ACCOUNT AGREEMENTS
We examined three samples of checking account agreements: the 100 largest banks based on
consolidated deposits less than $250,000 (i.e., the deposit insurance threshold); a random
sample of 150 banks not among the 100 largest (referred to as small and mid-sized banks); and
the 50 largest credit unions based on the amount of insured deposits. We used insured deposits
as a proxy for consumer accounts.
For each sample, we first attempted to collect the institution’s checking account agreement
(typically its general deposit account agreement) from its web page. If the agreement was not
available, or if the available agreement was undated or dated prior to 2011, the Bureau directed
the institution to provide the agreement pursuant to section 1022(c)(4) of the Dodd-Frank
Act.333 For all of the largest banks, 141 of the 150 small and mid-sized banks, and 49 of the 50
largest credit unions, we were able to obtain some version of its checking account agreement.
From that agreement, we determined whether the institution used an arbitration clause, and, if
so, coded it as described above.

332

Agreements that provided for use of California’s judicial reference procedure were not treated as including an
arbitration clause.
333

133

12 U.S.C. § 5512(c)(4).

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

As noted above, we used consolidated insured deposits, obtained from December 31, 2011 and
December 31, 2012 call reports, as a measure of market share. As a general matter, insured
deposits were consolidated across all affiliated institutions. If, however, affiliated institutions
used different forms of dispute resolution or different arbitration clauses, we included each
institution separately in the sample.334
GPR PREPAID CARD AGREEMENTS
As noted above, market data are much less complete for GPR prepaid cards than for either credit
cards or checking accounts. We included in the sample all GPR prepaid cards (1) listed on the
Visa, MasterCard, or NerdWallet web pages advertising such cards;335 or (2) examined in
various recent studies of the terms of prepaid cards.336 We also included agreements from the
two leading credit union GPR card programs, by PSCU and CUNA.337 Most of the cards in the
sample (44 of 63, or 69.8%) were listed in multiple sources. We collected the cardholder
agreements for the cards from the Internet, and excluded eight cards for which cardholder
agreements were unavailable and three cards that had been discontinued.
Market share data came from the Aité Group’s November 2012 report on the prepaid card
market.338 Because of limited market share data, the results for GPR prepaid cards based on the
dollar amounts loaded on the cards are limited to the firms for which market share data are

334

As a result, for example, our sample of the 100 largest banks includes 103 observations.

335

MasterCard, Get a Prepaid Card That’s Right for You, http://www.mastercard.us/get-a-prepaid-card.html (visited
Aug. 2, 2013); Visa, Get a Visa Prepaid Card, http://usa.visa.com/personal/cards/prepaid/prepaid-card-online.html
(visited Aug. 2, 2013); NerdWallet, Prepaid Debit Cards, http://www.nerdwallet.com/prepaid.
336

Aité Group, supra note 62, at 18-19; Bankrate.com, 2013 Prepaid Debit Cards Survey,
http://www.bankrate.com/finance/banking/best-prepaid-debit-cards.aspx; Bretton Woods, Inc., Analysis of Branded
General Purpose Reloadable Prepaid Cards: A Comparative Cost Analysis of Prepaid Cards, Basic Checking Accounts
and Check Cashing 9 (Feb. 2012); CardHub, Prepaid Cards Report ― 2013, http://www.cardhub.com/edu/prepaidcards-report-2013/; Consumer Reports, Prepaid Cards: How They Rate on Value, Convenience, Safety and Fee
Accessibility and Clarity 9 (July 2013); Pew Charitable Trusts, supra note 40, at 26.

337

CUNA, Learn More About Our Prepaid Debit Cards,
http://www.cunastrategicservices.com/ICUL_Service_Corporation_CUMONEY_Debit_Cards_134.html (last visited
Oct. 17, 2013); PSCU, Prepaid – Reloadable Cards, http://www.pscu.com/solutions/prepaid-reloadable-cards.html
(last visited Oct. 17, 2013).
338

134

Aité Group, supra note 62, at 19.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

available, which make up 82.9% of the entire market.339 As a result, most of the prepaid cards in
the sample are not included in the market share figures. Moreover, because two of the firms that
are included use two different form cardholder agreements, the market share data sometimes
are presented as ranges rather than as a single figure.

339

135

Id. at 18-19.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

APPENDIX B

Arbitration data procedures
This appendix describes our methodology for reviewing the data described in section 4.
OVERVIEW
The AAA provided us with electronic documents from their case management records regarding
all non-class consumer arbitrations from January 1, 2010 to February 2013. The AAA grouped
these .pdf files together in folders by arbitration proceeding. Each folder could contain as few as
one document to over 400 documents.340 The folders generally all contained some variation of
the AAA claim form. Otherwise, there were few consistent markers across the documents.
Accordingly, unless there was a cross-reference from another document, we would likely be
unaware if documents were missing from the files. Sometimes folders simply stop, seemingly
mid-way through a proceeding. As we will discuss in future work, it is difficult to tell from the
case record whether such files reflect claimant withdrawals, settlements, or some other
outcome.
We manually reviewed the documents to identify cases relating to our product markets of
interest: credit cards, checking accounts, and payday loans. This process resulted in 1,241
folders (each representing a single consumer arbitration) relating to the credit card, checking, or
payday loan markets from the years 2010, 2011, and 2012. We also identified four prepaid cases,
which, to date, we have not considered further. We have begun to identify other product
markets, such as student loan cases, that we may consider in future work.

340

For about two-dozen cases, we were not able to access one or more individual .pdf files provided by the AAA
because of technical issues. We are working with the AAA to resolve this issue. It is unlikely that these documents
contain information relating to the data we collected. Once we have resolved the issue, we will update our calculations
as necessary, and reflect any changes in our statutorily mandated report.

136

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Through multiple rounds of manual review, we then coded data points from each of the files
within our product markets. For some data, the collection process was only an exercise in simple
data entry: copying information from a particular standard document. In other cases, the
required information was similarly objective, but required gathering information from multiple
documents. And finally, one specific type of coding was necessarily more subjective. We discuss
each of these three “tiers” of data collection below.


Direct coding of information from claim forms provided our most “objective” data, as it
allowed no room for interpretation. This process means that our results capture and
reflect any errors in the original data.



Objective coding across multiple documents was sometimes necessary for certain data
points. For example, the claim forms may touch on descriptions of the parties’ claims or
the amounts in dispute, but these descriptions may be more thoroughly elucidated in
party-drafted summaries of claims, exhibits to claim forms, or correspondence from the
parties or AAA. The supplemental documents also provided information not raised in the
claim forms at all, such as fee assessments. Relying on documents outside of the claim
form, however, introduces some risk of inconsistency because other than the AAA claim
forms, the case folders did not contain other specific documents on a uniform or nearuniform basis.



Finally, our attempt to code the subject matter of claims is more subjective. We used a
maximum of three such subject matters from a set list for each dispute with the intent of
capturing the essence of each dispute. The procedure was intended to cover a range of
claim specification—from a very short case description in a hand-written arbitration
filing to a 50-page complaint drafted by a team of lawyers in a different dispute. This
process required a holistic review of each proceeding. To mitigate the risk of
inconsistent coding across individual reviewers, we relied on multiple rounds of review.

DIRECT CODING FROM SINGLE DOCUMENTS
The information we directly coded from AAA claim forms includes but is not limited to:


Party and representation information:


137

We coded this information for up to three parties on each side.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



We used a binary code to indicate whether more than three parties were involved on
either “side.”



We recorded whether the parties were represented by counsel (including in-house
counsel for companies) and listed the names of any outside counsel.341



Basic case information:


We recorded whether the claimant(s) purported to represent one or more classes of
claimants.



We listed the date of the claim.


If a case file had been closed and then reopened, we recorded the earlier
date.



If the claim form lacked a clear statement about the date, we used other
information, such as a date header created by a fax-machine, to
approximate the date.



We recorded information from claim forms regarding the relief sought by the
claimants:


We recorded whether claimants sought a specific amount of money; “at
least” or “at most” certain amounts; or whether they sought a range of
relief (e.g., “$1,000 to $10,000”).



If a claim form attached a supplementary document, like a long-form
complaint, and included instructions akin to “see attached” in the claim
form itself, our claim form claim information includes data from those
incorporated documents.



If the documents did not indicate a single overall claim amount for
multiple claims, we followed the language of the claim form (and attached
documents), with the exception that if multiple claims related to the same
cause of action or statute, we would not sum separately delineated claim
amounts.

341

If any party changed counsel, we recorded the most recent counsel of record.

138

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



In many cases, the parties described a specific debt amount in dispute. We
recorded this number.



We also recorded whether the parties sought injunctive relief, declaratory relief, or
punitive damages.342



We recorded who filed the claim. The AAA claim form has three checkboxes
regarding “who filed,” indicating: consumer-filing, company-filing, or mutual-filing.
We have not attempted to verify whether the checkbox selected on the claim form is
an accurate reflection of the submission.



We used the address of the first consumer in the AAA claim forms relating to credit card
disputes to determine a relevant nine-digit zip code. We used that information to
compare the median income of the consumers’ location against the median income of
locations for the overall credit cardholding population. The home address information
was entered into the U.S. Post Office’s on-line “Look up a Zip Code” tool.343 We did not
record any address information about consumers in this process other than their ninedigit zip codes. If the consumers listed a P.O. Box as their address, we would use the
nine-digit zip code for the P.O. Box. If they listed their attorney’s address, however, we
did not include a zip code for them in our analysis.

DIRECT CODING ACROSS MULTIPLE DOCUMENTS
We looked to information outside of the AAA claim forms to record certain other information,
including:


Filing fees:


We recorded amounts for:


Initial administrative fees;

342

We ultimately determined that our data was not sufficiently specific regarding injunctive relief, declaratory relief,
or punitive damages for reporting purposes. Parties frequently indicated in their claim forms that they were seeking
such relief, but there was rarely additional information regarding these requests. Similarly, the requests for punitive
damages were often vague or unbounded.
343

139

https://tools.usps.com/go/ZipLookupAction!input.action.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS





Neutral compensation (money paid by each party to the arbitrator); and



Any other fees.

In gathering fee information, we also recorded instances relating to consumer fee
waiver requests and fee advance requests.



We recorded whether there was an indication of a prior litigation proceeding and, if so,
who invoked arbitration as to that prior litigation proceeding. (By “invoking arbitration”
as to prior litigation, we mean conduct that ends or stays the prior litigation by some
form of reliance on an arbitration clause. We do not mean filing an arbitration per se.)


In many cases documents clearly showed that there was previous litigation and, in
some cases, they spelled out who invoked arbitration as to those proceedings. For
example, the arbitration record might contain a court order granting an identified
party’s motion to compel arbitration. In other cases, the record would include
documents showing that the consumer initiated the arbitration proceeding in lieu of
filing an answer in state or federal court.



In one specific set of cases, we made an assumption about who invoked arbitration as
to prior litigation. In those cases, the arbitration record: (1) had a clear indication of
prior litigation; (2) indicated that only one side brought claims in the prior litigation;
and (3) did not directly identify who invoked arbitration as to the prior litigation.
When all three criteria were satisfied, we assumed that arbitration was invoked by
the party who did not file the claims below. Our reasoning was that a party was
unlikely to file a claim in litigation and then, in the absence of counterclaims, invoke
arbitration as to its own claim or claims.344



We have begun attempting to record information about claim amounts to the extent that
they evolve over time prior to the arbitrator’s resolution or case closure for other
reasons. This effort involves documents beyond the claim form and attachments to it.

344

In a handful of cases, we found case records that indicated that some companies filed complaints in litigation and
then actually moved to compel arbitration regarding their own claims—even when there was no consumer
counterclaim at issue in the court proceeding.

140

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



We recorded instances where the AAA stated to a party that it had not complied with the
AAA Due Process Protocol: e.g., by not paying fees. We also recorded stated terminations
by the AAA for failure to comply with the Protocol.



We recorded basic information about the types of claim brought by each party,
specifically whether they brought:


Federal statutory claims;



State statutory claims;345



Fraud claims;



Contract claims;



Tort claims;



“General, unspecified claims”: given the informal requirements for pleading in
arbitration, we used this category to record consumer claims that were more
colloquial in nature—such as “I want my money back,” “This is unfair,” or “I was
discriminated against”—that the consumer did not identify as falling under any of the
alternative options above; and



“Refutes debt”: we recorded instances in which consumers affirmatively claimed that
they did not owe an alleged debt, at least to the extent that a consumer did not clearly
state a contract claim to the same effect. If a consumer did not bring a claim to this
effect, however, but simply faced a company claim in the arbitration for debt, then we
did not use this notation. (The fields here were intended to capture claim types, not
defenses.)



We specifically noted instances when consumers brought actions under the following
statutes:


The Alternative Mortgage Transaction Parity Act of 1982, 12 USC § 3801;



The Consumer Leasing Act of 1976, 15 USC § 1667;



The Electronic Fund Transfer Act, 15 USC § 1693;



The Equal Credit Opportunity Act, 15 USC § 1691;

345

As a subset of the statutory claims, we recorded instances where a party brought claims under an unfair or
deceptive acts or practices statute or rule.

141

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



The Fair Debt Collection Practices Act, 15 USC § 1692;



The Fair Credit Billing Act, 15 USC § 1601;



The Fair and Accurate Credit Transaction Act of 2003, 15 USC § 1601;



The Fair Credit Reporting Act, 15 USC § 1681;



The Homeowner Protection Act of 1998, 12 USC § 4901;



The Real Estate Settlement Procedures Act, 12 USC §2601;



The Truth in Savings Act, 12 USC § 4301;



The Truth in Lending Act, 15 USC § 1601;



The Interstate Land Sales Full Disclosure Act, 15 USC § 1701;



Antitrust statutes;



The Credit Repair Organization Act, 15 USC § 1679;



The Telephone Consumer Protection Act, 47 USC § 227;



The Communications Act of 1934, 47 USC § 609; and



The Racketeering Influenced and Corrupt Organizations Act, 18 USC § 1961.

CLAIM SUBJECT MATTERS
Finally, we attempted to identify the claim subject matter of each dispute, meaning a holistic
statement about the essence of each arbitration filing. We worked with subject matter experts
within the Bureau, who reviewed secondary sources and manually reviewed samples of
complaints in federal and state courts, to help us identify different subject matter areas for
disputes that might appear between consumers, on the one hand, and companies in connection
with credit cards, checking accounts/debit cards, and payday loans, on the other hand.
We organized the subject matters by product in a general hierarchy to mitigate the risk of
inconsistent coding across multiple reviewers. If application of this hierarchy would cause us to
miss what we considered to be the “core” subject matters of the claims, then we deviated from it
to allow for more accurate coding. For each case, we used a maximum of three fields in order to
capture the “core” subject matters of the dispute. We did not use all three fields unless we
thought that was necessary. (We frequently used under three.)
For Filings Relating to Credit Cards

We coded documents in the following priority order using our three fields. We also noted if:
(a) the dispute related to other issues not listed among these choices; or (b) there was
insufficient information to characterize the dispute. Not all of the subject matter categories
actually appeared in our review of the AAA documents.

142

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

We could choose one of three different debt-related categories, if any applied:


Substantive debt dispute – general claims about debt amounts owed or not owed (often
brought by companies in the form of accounts stated or breach of contract claims; or
brought by consumers as requests for declaratory relief that the consumers did not owe
certain amounts alleged by companies). We also included in this category claims under
the FDCPA if the only such claim was that the company misrepresented the amount of
debt in its collection efforts. (If either of the following two debt categories applied,
however, we did not also code the case as a “substantive debt dispute.”)



Debt collection process (harassment) – allegations that a company called consumers
after hours, used rude or obscene language, or made other threats during debt collection
activity.



Debt collection (process) – other claims relating to a company’s debt collection process,
such as FDCPA claims that a company contacted consumers known to be represented by
counsel.346

We also chose any of the following, to the extent that they applied:
Payments


Payment allocation – allegations that payments were applied to the wrong balances
(such as lower-interest rate balances);



Payment period – allegations that a payment or grace period was too short; and



Payment posting – allegations that the payment posting process was flawed in some
respect, such as being too slow.

Promotions and special features


Interest rate promotions;



Balance transfer promotions;

346

Claims under the FDCPA that companies misrepresented the amount of debt owed, however, were classified as
“Substantive Debt Dispute” claims.

143

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



Convenience checks; and



Add-on products.

Interest and Fee Issues


Interest rates/charges;



Late fees – this category was used, as opposed to “Interest Rates/Charges” when the
allegations related solely to late fees, with no claims about interest rates;



Foreign currency/transaction fees – allegations about foreign currency or transaction
fees or exchange rate calculations;



Annual/monthly fees; and



Other fees.

Credit line issues


Credit line increases/decreases;



Account opening issues; and



Account closing issues.

Improper transactions


Unauthorized use of account – allegations that account was improperly used by known
or unknown parties;



Merchant-related billing error – allegations that a credit card company has failed to
address alleged merchant failure to refund or correct consumers’ bills; and



Set-off – allegation that party has set-off a credit card balance against some other
account.

Information sharing and credit reporting


Credit reporting;



FACTA –if a dispute involved FACTA claims, we would not use the Credit Reporting (or
the disclosure of private information) notation unless other claims required that we do
so; and



Disclosure of private information – allegations that information was improperly
disclosed, at least outside the credit reporting context.

144

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Other


Discrimination;



Telemarketing; and



Antitrust/conspiracy issues.

For filings relating to checking accounts and debit cards

We used similar categories, including options to note that (a) the disputes related to other issues
not listed among these choices or (b) there was insufficient information to characterize the
dispute.
We could choose one of three different debt-related categories, if any applied:


Substantive debt disputes;



Debt collection process (harassment); and



Debt collection process (other).

We also chose from the following:
Fee issues, generally


Foreign currency/transaction fees;



Annual/monthly fees; and



Other fees (other than overdraft fees).

Overdraft issues


Overdraft ordering/timing; and



Other overdraft issues.

Account opening and closing


Refusal to open;



Other account opening/reopening issues; and



Account closing issues.

Issues with Deposits


Deposits – process issues – allegations that deposit process was flawed in some respect
(e.g., the consumer did not receive deposited funds within the proper time frame); and

145

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Deposits – other errors – allegations of substantive error in the deposit process (e.g.,



that the consumer did not receive deposited funds in whole or in part).
Issues with Authorized Payments
Authorized payments – process issues – allegations that payments process did not work



properly (e.g., process was contrary to representations); and
Authorized payments – other errors –allegations of substantive payment error (e.g.,



wrong person paid or not paid at all).
Payment error – unauthorized payments
Unauthorized recurring payments – this also was used for allegations that a bank failed



to honor stop payments on single checks;


Merchant-related error;



Improper set-off; and



Other unauthorized use – this would cover failure to protect against fraudulent access;
failure to address fraudulent checks, ACH, debit card, or ATM transactions;
unauthorized use by known parties; or situations where consumers claimed that banks
improperly provided funds pursuant to court order.

Information Sharing and Credit Reporting


Credit reporting;



FACTA; and



Disclosure of private information.

Other

146



Add-on products;



Discrimination;



Telemarketing Issues; and



Antitrust/Conspiracy.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Finally, for filings relating to payday loans

We used similar categories, including options to note that (a) the disputes related to other issues
not listed among these choices or (b) there was insufficient information to characterize the
dispute.
Debt issues


Substantive debt disputes;



Debt collection process (harassment); and



Debt collection process (other).

Account disclosures/breach of contract


Failure to disclose or follow core terms – we defined core terms to mean the finance
charge amount and the duration of the loan; and



Failure to disclose or follow other terms – this would include disclosure issues relating to
APRs; if a case fell into both this and the prior category, we would categorize the case as
falling into this category.

Rollovers


Rollover issues.

State regulatory requirements – we used only one of these four options at most, with priority
leading from the beginning to the end of this list:


Unlicensed lending activity;



Excessive interest rates;



Improper loan duration; and



Loan amount above state cap.

Bank account debits


Improper access to DDA (ACH); and



Improper access to DDA (Check).

Refusal


147

Refusal to issue or reissue loans.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Information sharing or credit reporting


Credit reporting; and



Disclosure of private information.

Product mischaracterization


Product mischaracterization;



Credit Service Organizations – allegations that state regulations violated; and



Credit Service Organizations – allegations that federal regulations violated.

Other

148



Add-on products;



Discrimination;



Telemarketing issues;



Antitrust/conspiracy claims; and



Other unfair loan practices.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

APPENDIX C

Identifying credit card cases in
federal district courts
As a basis of comparison, we reviewed pleadings in federal district courts for the years 2010,
2011, and 2012 to ascertain the number of consumer disputes relating to the credit card product
market for those years. To gather pleadings for our review, we used the LexisNexis Courtlink
database, which LexisNexis describes as the “largest collection of dockets and documents.”347
Courtlink search strings are limited to a maximum of either 32 words or 2,000 characters,
whichever is reached first.348 Courtlink does not allow for the use of parentheses to control the
order of operations of search terms. Accordingly, we crafted a deliberately overbroad text search
in Courtlink, using a search string intended to identify documents that were credit-card related
and likely to be pleadings. We followed that with manual review.
Our search string was:
“credit card” or “credit cards” or “charge card” or “charge cards” and complaint or crossclaim or
counterclaim or crossclaims or counterclaims and not motion
Applying Courtlink’s order of operations, the search resolved in the following order:
[(“credit card” or “credit cards” or “charge card” or “charge cards”) and (complaint or crossclaim
or counterclaim or crossclaims or counterclaims)] and not motion

347

http://www.lexisnexis.com/en-us/products/courtlink-for-corporate-or-professionals.page (“Search across the full
text of more than 168 million federal and state court dockets and documents in a single click.”).
348

149

https://courtlink.lexisnexis.com/Search/Single/SingleSearchHelp.aspx.

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

The search, when applied against all of Courtlink’s federal district court data sets, returned more
than 32,000 documents. We manually reviewed these documents to isolate complaints in
federal court containing company-consumer disputes about credit cards.
We also coded other information, including:


Whether the plaintiff purported to represent a class;



The court in which the pleading was filed;



The docket number;



Whether a jury was requested;



The filing date; and



Whether the consumer or the company filed.

For manageability, we focused our review on initial complaints, not amended complaints. We
also eliminated duplicate hits. At the end of that process, we identified 3,054 cases.

150

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

APPENDIX D

Additional arbitration data
This appendix contains additional data figures that are referenced in section 4.7.
Figures 32 and 33, which relate to all credit card consumer arbitrations (debt collection, along
with non-debt collection) exclude credit card proceedings for which we were unable to associate
census tract-based median income information. In some cases, the AAA case record did not
provide sufficient location information. (For example, it might provide location information
only for the consumer’s counsel or debt settlement provider.) Thus, the analysis shown here is
based on information from 911 credit card arbitrations.
Figures 34 and 35, which relate to checking account/debit card consumer arbitrations, similarly
exclude arbitrations for which we were unable to associate census tract-based median income
information. Thus, it is based on information from 56 checking account/debit card arbitrations.
Finally, Figures 36 and 37, which relate to payday consumer arbitrations, exclude payday cases
for which we were unable to associate census tract-based median income information.
Accordingly, it is based on information from 123 payday arbitrations.

151

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 32:

MEDIAN INCOME OF CREDIT CARD HOLDER LOCATIONS AND LOCATIONS OF CONSUMERS
INVOLVED IN CREDIT CARD CONSUMER ARBITRATIONS 2010-12

FIGURE 33:

PERCENTAGE DIFFERENCES BETWEEN DISTRIBUTIONS IN FIGURE 32, CREDIT CARD
ARBITRATIONS

152

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 34:

MEDIAN INCOME OF U.S. CENSUS TRACTS AND LOCATIONS OF CONSUMERS INVOLVED IN
CHECKING ACCOUNT/DEBIT CARD CONSUMER ARBITRATIONS 2010-12

FIGURE 35:

PERCENTAGE DIFFERENCES BETWEEN DISTRIBUTIONS IN FIGURE 34, CHECKING
ACCOUNT/DEBIT CARD ARBITRATIONS

153

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

FIGURE 36:

MEDIAN INCOME OF U.S. CENSUS TRACTS AND LOCATIONS OF CONSUMERS INVOLVED IN
PAYDAY CONSUMER ARBITRATIONS 2010-12

FIGURE 37:

PERCENTAGE DIFFERENCES BETWEEN DISTRIBUTIONS IN FIGURE 36, PAYDAY
ARBITRATIONS

154

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

APPENDIX E

Small claims court data and
methodology
This appendix provides additional data for section 5. It also provides a more detailed description
of our methodology.
DATA BY JURISDICTION
Table 10 shows data by jurisdiction for our estimated outer limit on consumer claims against our
credit card issuer sample.
Our issuer sample covers on the order of 84% of the credit card market measured by
outstandings.349
We have included for each jurisdiction the estimated annual volume for credit card direct mail
for the issuer sample overall, using data from a commercial provider.350 Those volume numbers
show that our issuer sample collectively has a significant presence in each jurisdiction, at least
from a marketing perspective.351 It is true that a low number of claims for a specific issuer in a
particular jurisdiction may reflect a lack of issuance by that issuer in that area. To address this
issue, we have noted with an asterisk every instance in which a specific issuer had no direct mail
volume for 2012. Our review sample only includes one predominantly regional player. The rest
are national issuers. So this effect is primarily isolated to that one issuer.

349

See supra note 310.

350

Data were available only by state. For counties, therefore, we assume that a county’s population share in the state
is equal to its share of direct mail.
351

In addition, for every jurisdiction, the direct mail numbers demonstrate that our sampled issuers collectively have
a significant presence.

155

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

TABLE 10: INDIVIDUAL SMALL CLAIMS COURT SUITS AGAINST ISSUERS 2012

Citi

Wells

USB

USAA

AmX

Disc

5/3

BoA

Cap1

JPM

All
Issuers

2012
Mailings per
Mintel

Alaska

1

0

0

0

0

0

*

2

0

0

3

n/a

Connecticut

0

2

0

0

0

0

*

4

0

5

11

129,650,753

Delaware

1

0

0

0

0

1

1

1

1

0

5

15,665,346

District of Columbia

2

4

0

0

0

0

*

2

1

0

9

40,875,797

Iowa

1

3

4

0

0

0

*

1

5

3

17

125,424,333

Minnesota

0

17

5

2

0

1

*

1

0

0

26

260,603,424

New Mexico

0

1

0

1

0

0

*

0

0

0

2

36,658,268

New Jersey

5

19

0

3

2

0

*

17

2

9

57

271,344,574

13

9

0

0

0

0

0

18

6

21

67

70,238,046

North Dakota

0

0

0

0

0

0

*

0

0

0

0

378,259,494

Oklahoma

0

0

0

1

0

0

*

0

0

3

4

110,984,476

Oregon

0

6

2

1

0

0

0

3

1

1

14

156,191,952

Utah

1

5

1

1

0

0

*

1

0

1

10

116,951,853

Wisconsin

1

1

1

0

0

0

0

2

0

3

8

253,280,825

Alameda (CA)*

1

0

0

0

3

0

*

2

0

1

7

57,290,710

Broward (FL)

1

0

1

0

0

0

0

17

0

4

23

59,396,856

Clark (NV)

1

2

0

1

1

0

*

6

1

0

12

40,677,128

Harris (TX)

1

2

0

1

2

0

*

9

6

11

32

118,398,987

Hillsborough (FL)

0

9

0

1

0

0

0

12

1

4

27

41,767,393

King (WA)

0

8

1

1

0

0

*

4

0

13

27

71,826,030

Kings (NY)

11

2

1

0

9

1

0

13

7

32

76

84,569,238

New York (NY)

15

2

0

0

9

0

0

12

2

27

67

53,545,763

Orange (CA)

8

14

7

0

1

0

*

36

1

11

78

114,190,320

Palm Beach (FL)

0

9

1

25

5

0

0

5

2

4

51

44,856,321

Philadelphia (PA)*

0

0

0

0

0

0

0

0

0

0

0

58,925,778

Queens (NY)

8

3

0

0

3

0

0

8

10

4

36

75,318,585

Riverside (CA)

4

8

4

0

1

0

1

32

1

8

59

83,061,972

Sacramento (CA)

2

9

2

0

1

0

*

0

1

5

20

53,820,389

New York, ex-NYC

156

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

San Bernardino (CA)

0

5

4

0

0

1

*

11

0

4

25

77,203,777

San Diego (CA)

6

28

6

1

1

0

*

20

2

14

78

117,417,788

Santa Clara (CA)

0

2

0

0

1

0

*

6

2

8

19

67,584,914

All jurisdictions

83

170

40

39

39

4

2

245

52

196

870

The two figures below combine the data from Table 9 in section 5.3 with the data from Table 10
above. Figure 38 shows the data from the jurisdictions that do not limit or bar the use of the
small claims courts by companies. In these jurisdictions, company-initiated claims uniformly
outnumber our outer limit estimate for consumer-filed claims. Figure 39, by contrast, shows the
data from jurisdictions almost all of which limit or bar company use of small claims court. In
these jurisdictions, our outer limit estimate for the number of consumer-filed credit card claims
is larger than the number of company-filed claims.

FIGURE 38:

INDIVIDUAL AND ISSUER USE OF SMALL CLAIMS COURT 2012 – PART 1 (SCALED FROM
ZERO TO 14,000)

Wisconsin
Harris (TX)
Connecticut
Broward (FL)
Palm Beach (FL)
Minnesota
Philadelphia (PA)
Hillsborough (FL)
Iowa
District of Columbia

Individual-Initiated

Delaware

Company-Initiated

New Mexico
-200

157

1,800

3,800

5,800

7,800

9,800

11,800

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

13,800

FIGURE 39:

INDIVIDUAL AND ISSUER USE OF SMALL CLAIMS COURT 2012 – PART 2 (SCALED FROM
ZERO TO 100)

San Diego (CA)
Orange (CA)
Kings (NY)
New York, ex-NYC
New York (NY)
Riverside (CA)
New Jersey
Queens (NY)
King County (WA)
San Bernardino (CA)
Sacramento (CA)
Santa Clara (CA)
Oregon
Clark County (NV)
Utah
Alameda (CA)
Oklahoma
Alaska
North Dakota

Individual-Initiated
Company-Initiated

-80

-60

-40

-20

0

20

40

60

METHODOLOGY
We searched two sources of data to determine the incidence of consumer financial disputes in
small claims court: state databases and county databases.
State-level sample

We checked the online databases of the small claims courts of the 50 states and the District of
Columbia. We included a state database as a source of data if it allowed us to do all of the
following:


Perform a keyword search by party name or permit easy identification of parties;



Identify a party as a plaintiff, defendant, or third party (such as a trustee or garnishee);



Capture statewide data, at least according the database. We excluded states for which we
could tell that a number of counties were missing. The one exception was New York
which we included because we were able to capture missing counties using our countybased sample;

158



Conduct statewide keyword searches rather than county-by-county searches;



Identify suits filed in 2012; and

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS



Run searches for free or for a reasonable cost (i.e., less than $1,000 per state).

The jurisdictions meeting the above criteria included Alaska, Connecticut, Delaware, the District
of Columbia, Iowa, Minnesota, New Mexico, New Jersey, New York (excluding New York City),
North Dakota, Oklahoma, Oregon, Utah, and Wisconsin. (The databases of the remaining 36
states did not meet one or more of the criteria above.) According to the 2010 Census data, our
covered states had a combined population of 52,894,194, or about 17.1% of the total U.S.
population of 308,745,538.352 The states meeting our criteria were concentrated in the Midwest
and Northeast.
County-level sample

To supplement our state-level sources, we checked the online small claims court databases of
the 30 most populous counties.353 We included a county as a source of data if the relevant
database permitted us to do the following:


Perform a keyword search by party name or permit easy identification of parties;



Identify a party as a plaintiff, defendant, or third party (such as a trustee or garnishee);



Identify claims filed between January 1, 2012 and December 31, 2012; and



Run searches for free or for a reasonable cost (i.e., less than $1,000 per state).

Thirteen counties met our criteria above: Harris (TX), Orange County (CA), Riverside (CA), San
Diego (CA) San Bernardino (CA), Clark (NV), King (WA), Santa Clara (CA), Broward (FL),
Philadelphia (PA), Sacramento (CA), Palm Beach (FL), and Hillsborough (FL). One county
(Suffolk County, NY) met the criteria above but was excluded because it was already included in

352

All population data retrieved from http://www.census.gov. New York state results exclude the population of the
five boroughs of New York City.

353

These 30 counties in order of population were as follows: Los Angeles, CA (9,818,605); Cook County, IL
(5,194,675); Harris County, TX (4,092,459); Maricopa County, AZ (3,817,117); San Diego County, CA (3,095,313);
Orange County, CA (3,010,232); Miami-Dade County, FL (2,496,435); Kings County, NY (2,504,700); Dallas County,
TX (2,368,139); Queens County, NY (2,230,722); Riverside County, CA (2,189,641); San Bernardino County, CA
(2,035,210); Clark County, NV (1,951,269); King County, WA (1,931,249); Tarrant County, TX (1,809,034); Santa
Clara, CA (1,781,642); Wayne County, MI (1,820,584); Broward County, FL (1,748,066); Bexar County, TX
(1,714,773); New York County, NY (1,585,873); Philadelphia County, PA (1,526,006); Alameda County, CA
(1,510,271); Middlesex County, MA (1,503,085); Suffolk County, NY (1,493,350); Sacramento County, CA (1,418,788);
Bronx County, NY (1,385,108); Nassau County, NY (1,339,532); Palm Beach County, FL (1,320,134); Cuyahoga
County, OH (1,280,122); and Hillsborough County, FL (1,229,226).

159

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

a state-level source (New York state). In addition to the online small claims court databases, we
found county-level data from two other sources. First, we collected data in person from
terminals in the small claims court clerk offices of New York, Queens and Kings Counties (the
boroughs of Manhattan, Queens and Brooklyn). While not available online, these databases met
all of the other county-level criteria. Second, Westlaw and Lexis’s Courtlink provided data for
two counties not otherwise available online: Alameda and Riverside Counties (both CA).354
In all, these county-level sources added 17 jurisdictions to our dataset, or approximately
35,160,801 persons all from states (California, Florida, Nevada, Pennsylvania, Texas, and
Washington) or locations (New York City) not already included in our set of state-level sources.
In total all our sources—state and county—together cover approximately 85 million people, and
20 states (including the District of Columbia), either in whole or in part.
In these 31 states and counties, the small claims jurisdictional limit (the maximum amount of
money damages that a plaintiff may claim and still require a defendant to dispute in small
claims court) ranged from $3,000 to $15,000, as shown below.
TABLE 11: SMALL CLAIMS JURISDICTIONAL MONEY DAMAGES LIMITS
State/County

Small Claims Jurisdictional Limit

Alaska

$10,000

Connecticut

$5,000

Delaware

$15,000

District of Columbia

$5,000

Iowa

$5,000

Minnesota

$10,000

New Jersey

$3,000 ($5,000 if demand is for return of deposit)

354

Other than these two counties, Westlaw and Lexis provided no unique additional jurisdictions that were not
already present in our dataset; conversely, most of the jurisdictions in our dataset were not available on Westlaw and
Lexis. (To the extent that Westlaw and Lexis offered data available independently online, our searches of Westlaw and
Lexis either confirmed our results or showed that Westlaw and Lexis offered fewer results.)

160

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

New Mexico

$10,000

New York (excluding NYC)

$5,000 (additional jurisdictional limits for companies)

North Dakota

$10,000

Oklahoma

$7,500

Oregon

$10,000

Utah

$10,000

Wisconsin

$10,000 for replevin, $5,000 for tort, $25,000 for consumer credit transaction
(for return of personal property subject to a lease or credit from a dealer)

Alameda (CA)

$10,000 (additional jurisdictional limits for companies)

Broward (FL)

$5,000

Clark (NV)

$7,500

Harris (TX)

$10,000

Hillsborough (FL)

$5,000

King (WA)

$5,000

Kings (NY)

$5,000 (additional jurisdictional limits for companies)

New York (NY)

$5,000 (additional jurisdictional limits for companies)

Orange (CA)

$10,000 (additional jurisdictional limits for companies)

Palm Beach (FL)

$5,000

Philadelphia (PA)

$12,000

Queens (NY)

$5,000 (additional jurisdictional limits for companies)

Riverside (CA)

$10,000 (additional jurisdictional limits for companies)

Sacramento (CA)

$10,000 (additional jurisdictional limits for companies)

San Bernardino (CA)

$10,000 (additional jurisdictional limits for companies)

San Diego (CA)

$10,000 (additional jurisdictional limits for companies)

Santa Clara (CA)

$10,000 (additional jurisdictional limits for companies)

161

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Institutions

Our search for disputes in small claims court involving consumer financial products focused on
credit card issuers. The industry is concentrated, so searches for relatively few issuers cover
most of the market.355 To come up with our list, we started with the list of the largest issuers
with pre-dispute arbitration clauses in their agreements at year-end 2012.356 We ranked the top
15 by consolidated volume of credit card loans outstanding at year end 2012, based on call
report data, as reported by SNL.357 This is shown below.
TABLE 12: SUMMARY OF LARGE ISSUERS BY SMALL CLAIMS COURT ARBITRATION CARVE-OUT STATUS

Credit Card Issuer

1. Citibank

2012 Loans Outstanding
($ 000s)
107,126,000

Small Claims Court
Carve-Out?
Yes – Mutual

2. American Express

56,714,000

Yes – Consumer only

3. Discover

50,927,790

Yes – Consumer only

4. GE

26,349,553

Yes – Mutual

5.

Wells Fargo

24,651,342

Yes – Mutual

6.

U.S. Bank

17,120,000

Yes – Mutual

7.

USAA

15,879,574

Yes – Mutual

8.

Barclays

14,281,456

Yes – Mutual

9.

Comenity

7,245,336

Yes – Consumer only

355

The credit card market is concentrated; the 10 credit card issuers we selected make up an overwhelming share of
the market, as measured by outstandings. See supra note 310.
356

See section 3, supra.

357

We adjusted two loans outstanding figures where SNL’s automated data via Excel appeared to be inconsistent with
our independent check of the regulatory filings.

162

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

10. First National Bank of
Omaha

4,229,188

Yes – Claims < $25,000

11. Fifth Third

2,085,601

No

12. BB&T

1,744,138

Yes – Consumer Only

13. Merrick

1,323,392

Yes – Consumer Only

14. Regions

899,929

Yes – Mutual

15. KeyBank

729,151

Yes – Mutual

We first selected the top four issuers from this list that had “mutual” small claims court carveouts. Excluding GE, this gave us Citibank, Wells Fargo, U.S. Bank, and USAA. We excluded GE
because its credit card business is exclusively private label, meaning that it issues no GE cards,
only cards for other businesses.358 We then added American Express and Discover, as the largest
issuers with non-mutual small claims court carve-outs (meaning that consumers may require
the resolution of a dispute in small claims court, but the issuer may not do so), and Fifth Third,
as the largest issuer with an arbitration provision with no small claims court carve-out. The
resulting issuers are shown in bold above.
As a further comparison, we selected the largest issuers with no arbitration clauses at all. The
top three issuers with no arbitration clause, by consolidated credit card loans outstanding at
year end 2012, were (in order) Chase ($117.6 billion), Bank of America ($94.8 billion), and
Capital One ($77.8 billion).359 Altogether, the 10 issuers we selected had nearly $564.8 billion in
credit card loans outstanding at year end 2012.

358

With private label credit cards, the actual credit card issuer may not be named on the card or in materials the
account holder might see, such as the credit card agreement, or the monthly credit card statements. As a result, such
an account holder might only know to file against the merchant listed on the credit card rather than the actual issuer.
Given the lengthy list of companies issuing cards through GE, a search for consumer-initiated claims against GE likely
would have been difficult, time consuming and still potentially inaccurate.
359

We adjusted one figure reported by SNL. As to Bank of America, the SNL reported number included foreign credit
card loans. We added the domestic FIA Card Services, National Association credit card loans (94,832,485,000) to the

163

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Searching for cases

To search the 31 databases meeting our criteria, we formulated search terms to find disputes in
small claims courts initiated by individuals against credit card issuers (“consumer-filed cases”)
and by issuers against individuals (“issuer-filed cases”).
To develop search terms for the consumer-filed cases, we started with the names of the parent
companies for our sample of issuers. We added all their subsidiaries included in the credit card
agreement database or reported by SNL as holding credit card loans as of year-end 2012. The
resulting list of names is in column 2 of Table 13 below. From this list, we developed search
terms. We selected the minimum number of common words or phrases that would still capture
every legal entity related to the 10 issuers we identified (e.g., “American Express” was intended
to capture both “American Express Centurion Bank” and “American Express Travel Related
Services, Inc.”; “Citi” was intended to capture “Citibank, National Association” and “Citigroup,
Inc.”).
We added other terms likely to be used by consumers as plaintiffs, based on the assumption that
consumers (or the court employees entering case information in small claims court databases)
may not always list the correct credit card-issuing legal entity in their suits. They may
erroneously list, instead, the corporate parent (“American Express Co.”), a truncated and
informal version of the name (“American Express” or “AmEx”), alternate spellings (e.g., “Capitol
One” for “Capital One”), or spacings (“J. P. Morgan” rather than the now-official “JPMorgan”).
They may also use trade names that would not be found in a search for the main entity (e.g.,
“BankAmericard”).
The full list of search terms is below in Columns 3 and 4.

domestic Bank of America, National Association credit card loans (2,822,000). After the top three issuers, the next
three issuers without arbitration clauses had significantly smaller volumes of loans outstanding: Navy Federal Credit
Union ($6.777 billion), PNC Bank, National Association ($3.821 billion), and World’s Foremost Bank ($3.852 billion).

164

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

TABLE 13: CREDIT CARD ISSUING ENTITIES AND SEARCH TERMS

Corporate
Parent

Subsidiary holding credit card loans as of
YE 2012 per SNL

Search Term
Based on Issuing
Entities + Parents

Other Terms
(Abbreviation,
Misspellings,
Trade Names)

American
Express
Co.

American Express Bank, FSB
American Express Centurion Bank
American Express Travel Related
Services Company, Inc.
American Express Credit Corporation

“American
Express”

“AmEx”
“Am Ex”

Citigroup,
Inc.*

Banamex USA
Citibank, National Association

“Banamex”
“Citibank”
“Citicorp”
“Citigroup”

“Citi”

Discover
Financial
Services

Discover Bank

“Discover”

N/A

“Fifth Third”

“Fifth 3rd"
“5th Third”
“5th 3rd”
“5/3rd"

Fifth Third
Bancorp

165

Fifth Third Bank

U.S.
Bancorp

U.S. Bank National Association
Elan Financial Services

“U.S. Bancorp”
“U.S. Bank”
“Elan”

“US Bancorp”
“US Bank”
“U S Bancorp”
“U S Bank”
“U. S. Bancorp”
“U. S. Bank”

USAA
Insurance
Group

USAA Federal Savings Bank
USAA Savings Bank

“United Services”
“USAA”

n/a

Wells
Fargo
& Co.

Wells Fargo Bank Northwest, National
Association
Wells Fargo Bank, National Association
Wells Fargo Financial National Bank

“Wells Fargo”

“WellsFargo”

JPMorgan
Chase
& Co.

JPMorgan Chase Bank, National Association
Chase Bank USA, National Association

“JPMorgan”
“Chase”

“JP Morgan”
“J P Morgan”
“J.P.Morgan”
“J.P. Morgan”
“J. P. Morgan”

Bank of
America
Corp.

Bank of America, National Association
FIA Card Services, National Association

“Bank of America”
“FIA”

“BoA”
“BofA”
“BankAmericard”

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

Capital
One
Financial
Corporation

Capital One, National Association
Capital One Bank (USA), National Association

“Capital One”

“Capital 1”
“Cap One”
“Cap 1”
“Capitol One”

* Note: Department Stores National Bank is a subsidiary of Citigroup, Inc. that reported some credit card loans on its
balance sheet. It has no arbitration clause and, as such, it was excluded from Citibank’s loan volume and search
terms.

Once we identified cases that named entities responsive to these search terms, we eliminated
duplicate entries because different search terms sometimes retrieved the same case. We also
removed cases in which the responsive entity was not the defendant (e.g., a garnishee) because
in such cases the issuer is not involved in a dispute with a consumer. We culled cases in which
the responsive entity was by its name alone identifiably not a credit card or checking-related
subsidiary. (For instance, we cut results in which “mortgage,” “home loan,” “insurance,”
“casualty,” “auto,” “investment,” or “securities” appeared in the entity name.) Finally, we also
removed cases in which the filing party was not a natural person.
For issuer-filed cases, we included only disputes in which the named plaintiff was one of the
legal entities issuing credit cards identified in Table 13. In other words, we used column 2 for
our search, not columns 3 and 4. In most jurisdictions we could not tell from available
documents if a credit card was involved in the dispute. We assumed that companies would file
credit card-related disputes only in the name of credit card-issuing legal entities and not in the
name of other subsidiaries, including the parent company. Finally, we removed cases in which
the party being sued was not a natural person.
Consumer Checking Account Volume

The search terms used for consumer-filed cases will also identify a number of cases against large
retail banks within the same corporate families. Many of the issuers covered by our sample also
offer checking account and other retail banking products. In addition, our operative search
terms for credit card issuer will identify other subsidiaries in the same corporate family that do
not issue credit cards, but do offer checking account and other retail banking products. As a
result, we should expect our outer limit estimate for credit card consumer-filed cases to include
a significant share of any checking account and other retail banking product cases filed in these
small claims courts.

166

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

APPENDIX F

Defined terms
We set out below a number of defined terms as used in this presentation:
DEFINED TERM

167

AAA

American Arbitration Association, an arbitration administrator

AAA DATA

Electronic case records provided voluntarily by the AAA to the Bureau
for all AAA-administered consumer arbitration filings since 2010

AAA CASE DATA

A subset of AAA Data pertaining only to credit card, checking
account, or payday loan disputes for 2010, 2011 and 2012

ADMINISTRATIVE FEES

Administrative fees payable to an arbitration administrator

ARBITRATION CLAUSE

A provision in a consumer contract that provides for binding arbitrated
resolution of claims that arise between the parties

ARBITRATOR FEES

Fees payable to an arbitrator to resolve claims that arise between the
parties

BUREAU

Consumer Financial Protection Bureau

CLASS ACTION

A form of litigation in which a party seeks to or does represent a
“class” of similarly-situated parties

CONSUMER DUE
PROCESS PROTOCOL

AAA-adopted principles directed to the fairness of consumer
arbitrations

COMPANY-FILED
ARBITRATION

Arbitration recorded on AAA claim form as being filed by a company
against a consumer

CONSUMER-FILED
ARBITRATION

Arbitration recorded on AAA claim form as being filed by a consumer
against a company

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS

168

DEBT COLLECTION
ARBITRATION

Company-filed arbitration seeking debt, or consumer-filed or mutually
submitted arbitration made after court collection proceedings; see
Section 4.2.2 for more details

FAA

Federal Arbitration Act, 9 U.S.C. §§ 1-16

FLESCH READABILITY
SCORE

Readability tests designed to indicate the difficulty of comprehension
of text. The tests are a widely used standard in plain language
analysis. Scores range from 0.0 to 100.0, with a higher number
indicating greater readability.

FLESCH-KINCAID
GRADE LEVEL

Measure that translates readability of a text into the grade level
required to understand the text. A lower grade level indicates greater
readability.

FORMAL DISPUTE
RESOLUTION

Litigation or arbitration mechanism that can effect binding resolution
of a dispute

GPR PREPAID CARD

General purpose reloadable prepaid card

INITIAL FEES

Administrative and arbitrator fees assessed at the beginning of a
consumer arbitration

JAMS

JAMS, Inc. (formerly known as Judicial Arbitration and Mediation
Services), an arbitration administrator

MUTUALLY
SUBMITTED
ARBITRATION

Arbitration recorded on claim form as being submitted by both parties

NAF

National Arbitration Forum, previously a significant consumer
arbitration administrator

PRIVATE LABEL
CREDIT CARD

Credit card issued and/or managed by a financial institution on behalf
of a merchant or a wholesale manufacturer for use only in that
merchant establishment.

SUPPLEMENTARY
PROCEDURES

AAA rules applicable to consumer arbitration

CONSUMER FINANCIAL PROTECTION BUREAU, ARBITRATION STUDY PRELIMINARY RESULTS