Consumer Credit: Limited Information Exists on Extent of Credit Report Errors and Their Implications for Consumers

Published by the Government Accountability Office on 2003-07-31.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                             United States General Accounting Office

GAO                          Statement for the Record
                             Before the Committee on Banking,
                             Housing, and Urban Affairs, U.S. Senate

For Release on Delivery
Expected at 10:00 a.m. EDT
Thursday, July 31, 2003      CONSUMER CREDIT
                             Limited Information Exists
                             On Extent of Credit Report
                             Errors and Their
                             Implications for Consumers
                             Statement of Richard J. Hillman Director,
                             Financial Markets and Community Investment

                                                July 31, 2003

                                                CONSUMER CREDIT

                                                Limited Information Exists on Extent of
Highlights of GAO-03-1036T, a statement         Credit Report Errors and Their
for the record to Senate Committee on
Banking, Housing, and Urban Affairs             Implications for Consumers

Accurate credit reports are critical            Information on the frequency, type, and cause of credit report errors is
to the credit process—for                       limited to the point that a comprehensive assessment of overall credit report
consumers attempting to obtain                  accuracy using currently available information is not possible. Moreover,
credit and to lending institutions              available literature and the credit reporting industry strongly disagree about
making decisions about extending                the frequency of errors in consumer credit reports, and lack a common
credit. In today's sophisticated and
highly calibrated credit markets,
                                                definition for “inaccuracy.” The literature and industry do identify similar
credit report errors can have                   types of errors and similar causes of errors. Specifically, several officials and
significant monetary implications               reports cited collection agencies and governmental agencies that provide
to consumers and credit granters.               information on bankruptcies, liens, collections, and other actions noted in
In recognition of the importance of             public records as major sources of errors. Because credit report accuracy is
this issue, the Senate Committee on             essential to the business activities of consumer reporting agencies and credit
Banking, Housing, and Urban                     granters, the credit industry has developed and implemented procedures to
Affairs asked GAO to (1) provide                help ensure accuracy. However, no study has measured the extent to which
information on the frequency, type,             these procedures have improved accuracy. While the Federal Trade
and cause of credit report errors,              Commission (FTC) tracks consumer complaints on FCRA violations, these
and (2) describe the impact of the              data are not a reliable measure of credit report accuracy. Additionally, FTC
1996 amendments to the Fair
Credit Reporting Act (FCRA) on
                                                has taken eight formal enforcement actions directly or indirectly related to
credit report accuracy and                      credit report accuracy since Congress enacted the 1996 FCRA amendments.
potential implications of reporting             Neither the impact of the 1996 FCRA amendments on credit report accuracy
errors for consumers.                           nor the potential implications of errors for consumers is known.
                                                Specifically, because comprehensive or statistically valid data on credit
                                                report errors before and after the passage of the 1996 FCRA amendments
The lack of comprehensive                       have not been collected, GAO could not identify a trend associated with
information regarding the accuracy              error rates. Industry officials and studies indicated that credit report errors
of credit reports inhibits any                  could either help or hurt individual consumers depending on the nature of
meaningful discussion of what                   the error and the consumer’s personal circumstances. To adequately assess
could or should be done to improve              the impact of errors in consumer reports would require access to the
credit report accuracy. Because of              consumer’s credit score and the ability to determine how changes in the
the importance of accurate credit               score affected the decision to extend credit or the terms of the credit
reports to our national credit
                                                granted. Ultimately, a meaningful independent review in cooperation with
system, it would be useful to
perform an independent                          the credit industry would be necessary to assess the frequency of errors and
assessment of the accuracy of                   the implications of errors for individual consumers.
credit reports. Another option for
improving the accuracy of credit                Common Causes of Errors in the Consumer Credit Reporting Process
reports would be to create more
opportunities for consumers to
review credit reports. Such added
reviews would likely help further
ensure the overall accuracy of
consumer credit reports.


To view the full product, including the scope
and methodology, click on the link above.
For more information, contact Rick Hillman at
(202) 512-8678 or Harry Medina at (415) 904-
Mr. Chairman and Members of the Committee:

I appreciate the opportunity to provide this committee with information
on the accuracy of consumer credit reports. Accurate credit reports are
critical for all consumers attempting to obtain credit and for lending
institutions in making appropriate and timely decisions about extending
credit. Information from credit reports is used to compile credit scores,
which in turn are used as the basis for deciding whether to extend credit,
and for setting rates and terms for mortgages and other consumer loans.
Thus, inaccurate credit report data could have significant monetary
implications for individual consumers and credit granters in today’s
sophisticated and highly calibrated credit markets.

To help promote the accuracy, fairness, and privacy of personal
information assembled by consumer reporting agencies (CRAs), Congress
enacted the Fair Credit Reporting Act (FCRA) in 1970.1 Under FCRA, CRAs
must “follow reasonable procedures to assure maximum possible
accuracy” in credit reports. In 1996, amendments to FCRA expanded the
responsibilities of data furnishers, prohibiting them from knowingly
providing inaccurate consumer information to a CRA in certain
circumstances.2 Additionally, FCRA gave the Federal Trade Commission
(FTC or Commission) responsibility for enforcing compliance with the
act’s provisions—to the extent that this authority did not overlap the
authority of other financial regulators for specific institutions.

In a series of hearings this committee has recently held on FCRA issues,
questions concerning the accuracy of credit reports have surfaced. In
recognition of the importance of this issue, you asked us to provide the
committee with information on (1) the frequency, nature, and cause of
consumer credit report errors and (2) the impact of the 1996 FCRA
amendments on credit report accuracy and the potential implications of
credit reporting errors on consumers.

The information that we are providing is based on a review of the limited
literature on the subject, and on interviews and supporting documentation
obtained from the three major CRAs; the Consumer Data Industry
Association (CDIA), a trade association for the consumer reporting
agencies; the National Foundation for Credit Counseling (NFCC), a

Pub. L. No. 91-508, (15 U.S.C. § 1681 et. seq.).
Consumer Credit Reporting Reform Act of 1996, Pub. L. 104-208, 110 Stat. 3009-426.

Page 1                                                                    GAO-03-1036T
national nonprofit credit counseling network; the Federal Trade
Commission (FTC); the Federal Reserve; and five data furnishers.3 While
we asked the three major CRAs to provide data on the frequency, type, and
cause of errors in credit reports, they told us that they did not have data
that would specifically respond to our request. The CRAs also told us that
they compete with each other on the basis of the accuracy and
completeness of their credit reports and were reluctant to provide us with
any data they considered proprietary. However, they did agree to provide
available information on consumer disputes to CDIA, their trade
association, which provided that data to us in aggregated form.
Consequently, we were unable to independently verify the accuracy of this
data. Except for this limitation, we conducted our work in accordance
with generally accepted government auditing standards from June through
July 2003.

In summary, we found that information contained in the literature and the
available industry data on the frequency, types, and causes of credit report
errors are limited. Moreover, there is a large variance in the frequency of
errors presented by the literature and industry data. Unfortunately, we
cannot determine a definitive level of credit report accuracy because of
the data limitations inherent to both the literature and industry data.
However, the literature and industry had identified similar types of errors
in credit reports, including the inclusion of incorrect information and the
exclusion or incomplete reporting of information. Additionally, the
literature and industry consensus was that the causes of errors included
consumers, data furnishers, and CRAs. However, several industry officials
and reports identified collection agencies and organizations providing
public records data—on actions such as bankruptcies, liens, and
collections—as being major sources of errors in credit reports. In an effort
to ensure accuracy of credit report data, the credit industry has developed
and implemented procedures that standardized the manner in which
information was collected and transmitted. The FTC tracks consumer
complaints regarding possible FCRA violations and has taken eight
enforcement actions as of July 24, 2003, directly or indirectly related to
credit report data accuracy since the passage of the 1996 FCRA

 The consumer reporting agencies we contacted were Equifax, Experian, and TransUnion.
The data furnishers that we contacted were Bank of America, Citigroup, Discover, MBNA,
and JP Morgan/Chase.

Page 2                                                                  GAO-03-1036T
We cannot readily determine the impact of the 1996 FCRA amendments on
credit report accuracy or the potential implications of credit report errors
on consumers. This is attributable to the lack of trend data available on
credit report errors. Specifically, no entity collects or maintains the
necessary data for such an assessment. Similarly, we could not determine
the potential implications for consumers of credit reporting errors due to
the lack of quality information on the frequency of errors. However,
industry officials and studies suggested that errors and inaccuracies in
credit reports have the potential to both help and hurt individual
consumers. Minor inaccuracies in a consumer’s credit file may not hurt a
consumer if that individual had a very good credit history. On the other
hand, errors or inaccuracies in the credit report of a consumer with a less
than perfect credit history could result in the denial of credit or an offer of
less favorable credit terms. So, the impact of any particular error or
inaccuracy in a credit report is dependent on the specific circumstances of
the consumer.

The lack of comprehensive information regarding the accuracy of credit
reports inhibits any meaningful discussion of what could or should be
done to improve credit report accuracy. Because of the importance of
accurate credit reports to our national credit system, it would be useful to
perform an independent assessment of the current level of accuracy of
credit reports. The assessment would then form the basis for a more
complete and productive discussion of the costs and benefits of making
changes to the current system of credit reporting to improve credit report
accuracy. Another option for improving the accuracy of credit reports
would be to create more opportunities for consumers to review credit
reports. When consumers see their credit reports, they have a chance to
identify errors and ask for corrections, thus helping to ensure greater
overall accuracy of consumer credit reports.

Page 3                                                           GAO-03-1036T
                            Available studies and credit reporting industry data disagree on the extent
Information on              of errors in credit reports. The limited literature on credit report accuracy
Frequency, Type, and        indicated high rates of errors in credit report data. In contrast, the major
                            CRAs and CDIA stated that they did not track errors specifically but that
Cause of Credit             the data the credit industry maintained suggested much lower rates of
Report Errors Is            errors. Both the literature and the data provided by the credit industry had
                            serious limitations that restricted our ability to assess the overall level
Limited; Industry Data      credit reporting accuracy. Yet, all of the studies identified similar types
and Available Studies       and causes of errors. While data provided by the credit industry did not
Disagree on                 address type and cause of errors, representatives from the three major
                            CRAs and CDIA cited types and causes similar to those cited in the
Frequency of Errors         literature. The credit industry has developed and implemented procedures
                            to help ensure accuracy of credit report data, although no one has
                            assessed the efficacy of these procedures. Moreover, FTC tracks
                            consumer disputes regarding the accuracy of information in credit reports
                            and has taken eight enforcement actions directly or indirectly involving
                            credit report accuracy since 1996.

Literature Raised Serious   We identified three studies completed after the 1996 FCRA amendments
Questions Regarding Level   that directly addressed credit report accuracy, and one that indirectly
of Credit Report Accuracy   addressed the topic. One of these reports, published in December 2002 by
                            Consumer Federation of America, presents the frequency and types of
                            errors drawn from files requested by mortgage lenders on behalf of
                            consumers actively seeking mortgages.4 The Consumer Federation of
                            America initially reviewed 1,704 credit files representing consumers from
                            22 states and subsequently re-examined a sample of 51 three-agency
                            merged files. In this sample of merged files, the study found wide variation
                            in the information maintained by the CRAs, and that errors of omission
                            were common in credit reports. For example, the report stated that about:

                            •   78 percent of credit files omitted a revolving account in good standing;

                            •   33 percent of credit files were missing a mortgage account that had
                                never been late;

                            •   67 percent of credit files omitted other types of installment accounts
                                that had never been late;

                            Consumer Federation of America and National Credit Reporting Association, Credit Score
                            Accuracy and Implications for Consumers, December 2002.

                            Page 4                                                                  GAO-03-1036T
•   82 percent of the credit files had inconsistencies regarding the balance
    on revolving accounts or collections; and

•   96 percent of the credit files had inconsistencies regarding an account’s
    credit limit.

A March 1998 U.S. Public Interest Research Group (U.S. PIRG) study
found similar frequencies of errors in 133 credit files representing 88
individual consumers.5 U.S. PIRG reported that 70 percent of the files
reviewed contained some form of error. The errors ranged in severity from
those unlikely to have negative repercussions to those likely to cause a
denial of credit. For example, the report found:

•   41 percent of the credit files contained personal identifying information
    that was long-outdated, belonged to someone else, was misspelled, or
    was otherwise incorrect;

•   29 percent of the credit files contained an error—accounts incorrectly
    marked as delinquent, credit accounts that belonged to someone else,
    or public records or judgments that belonged to someone else—that
    U.S. PIRG stated could possibly result in a denial of credit; and

•   20 percent of the credit files were missing a major credit card account,
    loan, mortgage, or other account that demonstrated the
    creditworthiness of the consumer.

Similar to the U.S. PIRG study, a 2000 survey conducted by Consumers
Union and published by Consumer Reports asked 25 Consumers Union
staffers and their family members to apply for their credit reports and then
review them.6 In all, Consumers Union staff and family members received
and evaluated 63 credit reports and in more than half of the reports, they
found inaccuracies that they reported as having the potential to derail a
loan or deflect an offer for the lowest-interest credit card. The
inaccuracies identified were similar to those reported by the Consumer
Federation of America and U.S. PIRG—inclusion of information belonging
to other consumers, inappropriately attributed debts, inaccurate
demographic information, and inconsistencies between the credit reports
provided by the three major CRAs regarding the same consumer.

 U.S. PIRG, Mistakes Do Happen: Credit Report Errors Mean Consumers Lose, March
“Credit Reports: How Do Potential Lenders See You?” ConsumerReports.org, July 2000.

Page 5                                                                 GAO-03-1036T
                             While not specifically assessing the accuracy of credit reports, a Federal
                             Reserve Bulletin article found that credit reports contained
                             inconsistencies and cited certain types of data furnishers, including
                             collection agencies and public entities, as a primary source for some of the
                             inconsistencies found.7 Among the study’s findings:

                             •   Approximately 70 percent of the consumers in the study’s sample had a
                                 missing credit limit on one or more of their revolving accounts,

                             •   Approximately 8 percent of all accounts showed positive balances but
                                 were not up to date,

                             •   Between 1 and 2 percent of the files were supplied by creditors that
                                 reported negative information only, and

                             •   Public records inconsistently reported actions such as bankruptcies
                                 and collections.

                             An important aspect of the Federal Reserve study was that it used a
                             statistically valid and representative sample of credit reports, and received
                             access to this sample with the cooperation of one of the three major CRAs.
                             However, because the sample came from one CRA only, the findings of the
                             study may not be representative of other CRAs.

CRA and CDIA Data            Representatives of the three major CRAs and CDIA told us that they do not
Indicate Consumer            maintain data on the frequency of errors in credit reports. However, the
Disputes Rarely Identified   industry does maintain data that suggest errors are infrequent in cases of
                             an adverse action.8 CDIA stated that the three major CRAs provided or
Errors                       disclosed approximately 16 million credit reports, out of approximately 2
                             billion reports sold annually in the marketplace. According to CDIA data,
                             84 percent of the disclosures followed an adverse action and only 5
                             percent of disclosures went to people who requested their reports out of
                             curiosity. Out of these disclosures, CRA officials stated that an extremely
                             small percentage of people identified an error.

                             Robert Avery, Paul Calem, Glenn Canner, and Raphael Bostic, “An Overview of Consumer
                             Data and Credit Reporting,” Federal Reserve Bulletin, February 2003.
                              When a creditor or lender decides not to extend credit to an individual or not to extend
                             credit on the terms the individual requires and the individual does not accept a
                             counteroffer, this is considered an adverse action. After an adverse action, consumers have
                             the right to a free copy of their credit report.

                             Page 6                                                                     GAO-03-1036T
An Arthur Andersen study, conducted in 1992, found a similar infrequent
rate of errors arising from adverse actions. Under commission by the
Associated Credit Bureaus (now CDIA), the study reportedly found that
only 36 consumers—out of a sample of 15,703 people denied credit—
disputed erroneous information that resulted in a reversal of the original
negative credit decision.9 Similarly, in an attempt to respond to our data
request, CDIA produced data gathered by a reseller over a two-week
period that indicated that out of 189 mortgage consumers, only 2
consumers (1 percent) had a report that contained an inaccuracy. In our
conversation with data furnishers, we discovered that two conduct
internal audits on the accuracy of the information they provide to the
CRAs. These data furnishers indicated that the information they provide
and the CRAs maintain is accurate 99.8 percent of the time.

While consumer disputes do not provide a reliable measure of credit
report accuracy, CRA representatives told us that disputes provide an
indicator of what people perceive as errors when reviewing their credit
files. A CDIA official stated that five types of disputes comprise about 90
percent of all consumer disputes received by the three major CRAs. These
five dispute types are described as:

•   Claims account has been closed;

•   Dispute present or previous account status, payment history, or
    payment rating;

•   Dispute current balance;

•   Dispute related to disposition of account included in or excluded from
    a bankruptcy; and

•   Not my account.

Although CDIA could not provide a definitive ranking for all five types of
disputes, it did state that “not my account” was the most frequently
received dispute. After receiving a consumer’s dispute, FCRA requires a
CRA to conduct a reinvestigation. The purpose of reinvestigation is to

 This study found that 1,223 of their sample of 15,703 consumers who were denied credit
had requested their credit reports. Of those that had requested their credit reports, 304
consumers found and disputed errors. At the time of the study, 36 of those disputes had
resulted in a reversal of the original negative credit decision.

Page 7                                                                      GAO-03-1036T
either verify the accuracy of the disputed information, or to confirm and
remove an error.

CDIA provided data on the disposition of dispute reinvestigations by
categories of those received by the three major CRAs in 2002. CRA
officials explained that the data represents the first 3 quarters of 2002, and
that each CRA reported data on a different quarter. CDIA declined to
provide the total number of consumer disputes. Table 1 shows the
frequency of these four disposition categories. Specifically, the table
indicates that over half of all disputes required the CRA to modify a credit
report in some way, though not necessarily to remove an error.10

Table 1: Disposition of Consumer Disputes

 Result of Dispute                                                  Percent of Disputes
 Information verified as reported                                                      46
 Data modified/updated per furnisher’s instructions                                    27
 Data deleted per furnisher’s instructions                                           10.5
 Data deleted due to statutory time limit                                              16
Source: CDIA.

Notes: As provided by CDIA, percentages do not total to 100.

It is important to emphasize that not every dispute leads to identifying an
error. Indeed, many disputes, as the table indicates, resulted in a
verification of accuracy or an update of existing information. Additionally,
CRA and CDIA representatives stated that many disputes resulted in the
CRA clarifying or explaining why a piece of information was included in
the credit report. For example, if recently married consumers obtained a
copy of their files, they might not see their married names on file. In such
cases, the files still accurately reflected the most current information
provided to the CRA, but the consumer may have perceived the less-than-
current information as an error while the CRA would not. The CRA
representative cited another example of a consumer seeing an account

  “Information verified as reported” encompasses disputed information found to be
accurate after reinvestigation. “Data modified/updated per furnisher’s instructions”
encompasses disputed information that a CRA modified or updated after reinvestigation.
According to CDIA, the information in this category was not necessarily inaccurate. “Data
deleted per furnisher’s instructions” encompasses information identified as inaccurate
through reinvestigation. “Data deleted due to statutory time limit” encompasses
information that a CRA had to delete because the reinvestigation process exceeded the
time limits set by FCRA.

Page 8                                                                     GAO-03-1036T
                          listed with a creditor he or she did not recognize. However, the account in
                          question was with a retailer that subsequently outsourced its lending to
                          another company. In this case, the information was correct but the
                          consumer was not aware of the outsourcing. One CRA representative
                          indicated that over 50 percent of the calls they received resulted in what
                          they consider “consumer education.”

Literature and Industry   We cannot determine the frequency of errors in credit reports based on
Data Have Serious         the Consumer Federation of America, U.S. PIRG, and Consumers Union
Limitations               studies.11 Two of the studies did not use a statistically representative
                          methodology because they examined only the credit files of their
                          employees who verified the accuracy of the information, and it was not
                          clear if the sampling methodology in the third study was statistically
                          projectable. Moreover, all three studies counted any inaccuracy as an
                          error regardless of the potential impact. Similarly, the studies used varying
                          definitions in identifying errors, and provided sometimes obscure
                          explanations of how they carried out their work. Because of this, the
                          findings may not represent the total population of credit reports
                          maintained by the CRAs. Moreover, none of these groups developed their
                          findings in consultation with members of the credit reporting industry,
                          who, according to a CDIA representative, could have verified or refuted
                          some of the claimed errors.

                          Beyond these limitations, a CDIA official stated that these studies
                          misrepresented the frequency of errors because they assessed missing
                          information as an error. According to CRA officials errors of omission may
                          be mitigated in certain instances because certain lenders tend to use
                          merged credit report files in making lending decisions, such as mortgage
                          lenders and increasingly credit card lenders. CRA officials explained that
                          while complete and current data are necessary for a wholly accurate credit
                          file, both are not always available to them. For instance, credit-reporting
                          cycles, which dictate when CRAs receive data updates from data
                          furnishers, may affect the timeliness of data. CRAs rely on these updates,
                          which may come daily, weekly, or monthly depending on the data
                          furnisher’s reporting cycle. If a data furnisher provided information on a
                          monthly basis there would be a lag between a consumer’s payment, for
                          example, and the change in credit file information. Likewise, if a data

                           The Federal Reserve Bulletin article did not address the frequency of errors, although it
                          did discuss findings of inconsistencies.

                          Page 9                                                                      GAO-03-1036T
                              furnisher reported to one CRA but not to another, the two reports would
                              differ in content and could produce different credit scores. It is important
                              to note that reporting information to the CRAs is voluntary on the part of
                              data furnishers. While the Federal Reserve Bulletin article noted
                              inconsistencies as an area of concern, it recognized that all credit reports
                              would not contain identical information.

                              Along with misrepresenting error frequency by counting omitted
                              information, industry officials believed that the literature misrepresented
                              the frequency of errors because the literature defined errors differently
                              than the credit industry. The CRAs and CDIA stated that they consider
                              only those errors that could have a meaningful impact on a person’s credit
                              worthiness as real errors. This distinction is critical to assessing accuracy,
                              as, according to the CDIA, a mistake in a consumer’s name might literally
                              be an inaccuracy, but may ultimately have no impact on the consumer.

                              The data provided by CDIA and the CRAs have serious limitations as well.
                              For example, neither CDIA nor CRA officials provided an explanation of
                              the methodology for the collection of data provided by CDIA and for the
                              assessments cited by the CRAs. Moreover, because these data related
                              primarily to those errors that consumers disputed after an adverse action,
                              they excluded a potentially large population of errors. Specifically, these
                              data excluded errors that would cause a credit grantor to offer less
                              favorable terms on a loan rather than deny the loan application. The data
                              also excluded errors in cases where consumers were not necessarily
                              seeking a loan and therefore did not have a need to review their credit
                              reports. Additionally, as stated earlier, only a small percentage of
                              consumers requested credit reports simply out of curiosity. While the
                              CDIA representatives felt that these data were useful for assessing a level
                              of accuracy, they agreed that by focusing on these data only, the industry
                              did not consider a potentially large set of errors.

Both Literature and           While both the literature and credit industry representatives cited similar
Industry Identified Similar   types and causes of errors, neither the literature nor the credit industry
Types and Causes of           data identified one particular type or cause of error as the most common.
                              All respondents stated that error type could range from wrong names and
Errors                        incorrect addresses to inaccurate account balances and erroneous
                              information from public records.

                              Based on the literature we reviewed and on our discussions with CRA and
                              data furnisher officials, we could not identify any one cause or source
                              most responsible for errors. However, the Consumer Federation 2002

                              Page 10                                                          GAO-03-1036T
                                                                   study, the Federal Reserve Bulletin article, and a representative from the
                                                                   National Foundation for Credit Counseling stated they felt data furnishers
                                                                   often caused more errors than did CRAs or consumers. According to
                                                                   several respondents, this was particularly true for data furnishers, such as
                                                                   collection agencies and public entities that did not rely on accurate credit
                                                                   reports for lending decisions. For example, while a bank needs accurate
                                                                   information in assessing lending risk, and thus attempts to report accurate
                                                                   information, a collection agency does not rely on credit reports for
                                                                   business decisions, and therefore has less of an incentive to report fully
                                                                   accurate information. Data furnishers told us that they did not consider
                                                                   CRAs as a significant cause of errors, but stated that difficulty in matching
                                                                   consumer identification information might cause some errors. Data
                                                                   furnishers also stated that the quality control efforts among data
                                                                   furnishers might vary due to the extent of data integrity procedures in
                                                                   place. They explained that some smaller data furnishers might not have
                                                                   sophisticated quality control procedures because implementing such a
                                                                   system was expensive.

                                                                   On the other hand, errors might occur at any step in the credit reporting
                                                                   process. Consumers could provide inaccurate names or addresses to a
                                                                   data furnisher. A data furnisher might introduce inaccuracies while
                                                                   processing information, performing data entry, or passing information on
                                                                   to the CRAs. And, CRAs might process data erroneously. Figure 1 shows
                                                                   some common causes for errors that might occur during the credit
                                                                   reporting process.

Figure 1: Common Causes of Errors in the Consumer Credit Report Process

          Consumers                                                           Furnishers                                               reporting                   m
                                                                                                                                        agency                   A
                                m                                                             rm Form Form
                            For                                                            Fo
                              A                                                                A B C                                                           Folder C

    Consumers can provide inaccurate data to                            Furnishers can input accurate information incorrectly,    Bureaus can input inaccurate information into the
    furnishers by mistake or purposefully provide false                 pass on incomplete or inaccurate data to consumer         correct file, or input accurate information into the
    information to establish new credit data.                           reporting agencies, pass on accurate information in       incorrect file.
                                                                        incorrect format, or fail to voluntarily report data on

Source: GAO analysis of credit industry and Federal Reserve interview data.

                                                                   Page 11                                                                                             GAO-03-1036T
                             CRAs and data furnishers also cited other causes of errors. For example,
                             collection agencies and public records on bankruptcies, tax liens, and
                             judgments were cited as major sources of errors. CRA officials and data
                             furnishers said the growing number of fraudulent credit “repair” clinics
                             that coach consumers to make frivolous reinvestigation requests in an
                             effort to get accurate, though negative, information off the credit report
                             also might cause errors, as disputed information a CRA cannot verify
                             within 30 days is deleted from the consumer’s credit report. File
                             segregation, a tactic in which a consumer with a negative credit history
                             tries to create a new credit file by applying for credit using consistent but
                             inaccurate information, was another reported cause for inaccurate credit

Industry Has Implemented     The credit industry has been working on systems to help ensure accuracy
Procedures to Ensure Data    since the “reasonable procedures” standard took effect under FCRA in
Consistency and Accuracy,    1970. Within the last decade, CDIA has led efforts to implement industry
                             systems and processes to increase the accuracy of credit reports. In
but Efficacy of Procedures   commenting upon accuracy, representatives from CDIA, the CRAs, the
Not Known                    Federal Reserve, and the data furnishers stated that credit score models
                             were highly calibrated and accurate and, on the aggregate level, credit
                             reports and scores were highly predictive of credit risk.12

                             During the 1970s, the Associated Credit Bureaus (now CDIA) attempted to
                             increase report accuracy by introducing Metro 1, a method of
                             standardizing report formats. The goals of Metro 1 were to create
                             consistency in reporting rules and impose a data template on the industry.
                             In conjunction with the industry, in 1996 CDIA created Metro 2, an
                             enhancement of the Metro 1 format that enables a finer distinction for
                             reporting information. For example, Metro 2 allowed CDIA to implement
                             an “Active Military Code” to protect the credit reports of troops serving
                             overseas. Since active military personnel are legally entitled to longer
                             periods to make credit payments without penalty, this new code ensured
                             that data furnishers did not incorrectly report accounts as delinquent.

                             While use of the Metro format is voluntary, CRAs currently receive over 99
                             percent of the volume of credit data—30,000 furnishers providing a total of

                               Because many credit grantors are also data furnishers, it is generally in their best interest
                             to report accurate information to the CRAs, as they rely on credit reports received from the
                             CRAs in assessing risk. Likewise, the CRAs depend on ensuring accuracy in their credit
                             reports in order to provide a quality product to their customers, the credit grantors.

                             Page 12                                                                        GAO-03-1036T
                             2 billion records per month—in either Metro 1 or Metro 2 format, with
                             over 50 percent sent in Metro 2. One data furnisher who recently switched
                             from Metro 1 to Metro 2 found that data accuracy improved overall as
                             evidenced by the reduction in the number of data rejections by the CRAs
                             and dispute data. Those data furnishers that do not use the Metro formats
                             provide data on compact disc, diskette, tape, or other type of electronic
                             media. While use of standardized reporting formats ensures more
                             consistent reporting of information, because the industry has never
                             conducted a study to set a baseline level of error frequency in credit
                             reports, and does not currently collect such data, no one knows the extent
                             to which these systems have improved accuracy in credit reports.

FTC Has Taken                FTC has taken eight formal enforcement actions since the passage of the
Enforcement Actions          1996 FCRA amendments against CRAs, data furnishers, and resellers that
Related to the Accuracy of   directly or indirectly relate to credit report accuracy.13 FTC receives and
                             tracks FCRA complaint data against CRAs by violation type and uses this
Credit Reports Since 1996    data to identify areas that may warrant an enforcement action. While these
FCRA Amendments              data cannot provide the number of violations or frequency of errors in
                             credit reports, since each complaint does not necessarily correspond to a
                             violation, they can give a sense of the relative frequency of complaints
                             surrounding CRAs. We discuss complaint data in more detail in the next

                             According to FTC staff, accuracy in the context of FCRA means more than
                             the requirement that CRAs establish “reasonable procedures to assure
                             maximum possible accuracy of their reports.” They explained that the
                             statute also seeks to improve accuracy of credit reports by a “self-help”
                             process in which the different participants comply with duties imposed by
                             FCRA. First, creditors and others that furnish information are responsible

                               Prior to 1996, FTC carried out actions involving procedures to ensure the accuracy of
                             credit reports against TransUnion in 1983, TRW (which would later become Experian) in
                             1991, and Equifax in 1995. According to FTC, these “omnibus” actions differed in detail but
                             generally covered a variety of FCRA issues including accuracy, disclosure, permissible
                             purposes, and prescreening. While we limited this review to FTC’s accuracy-related efforts,
                             we are currently conducting additional work as part of another ongoing engagement
                             looking at FTC and the banking regulator’s enforcement of FCRA. A number of other
                             federal agencies have responsibilities under FCRA including the Office of the Comptroller
                             of the Currency, Federal Reserve Board, Office of Thrift Supervision, National Credit Union
                             Administration, Federal Deposit Insurance Corporation, Department of Transportation, and
                             Department of Agriculture. Each entity can pursue FCRA enforcement actions against their
                             respective regulated institutions as identified in FCRA.

                             Page 13                                                                    GAO-03-1036T
for accuracy. Second, credit bureaus must take reasonable steps to ensure
accuracy. Finally, users of credit reports must notify consumers (provide
adverse action notices) about denials of a loan, insurance, job, or other
services because of something in their credit report. FTC staff stated that
it is crucial that consumers receive adverse action notices so that they can
obtain their credit reports and dispute any inaccurate information. For
that reason, the Commission has made enforcement in this area a priority.

FTC staff stated that their primary enforcement mechanism is to pursue
action against a CRA or data furnisher that showed a pattern of repeated
violations of the law identified through consumer complaints. According
to FTC staff, the Commission has taken eight enforcement actions against
CRAs, furnishers, or lenders, since 1996 that directly or indirectly
addressed credit report accuracy.14 One case pertained to a furnisher
providing inaccurate information to a CRA, two cases pertained to a
furnisher or CRA failing to investigate a consumer dispute, and two
actions were taken against lenders that did not provide adverse action
notices as required by statute. The remaining three cases were against the
major CRAs for blocking consumer calls and having excessive hold times
for consumers calling to dispute information on their credit reports.15 In
addition to enforcing FCRA, FTC also provides consumer educational
materials and advises consumers on their rights (such as the right to sue a
CRA or data furnisher for damages and recoup legal expenses).

  The eight cases are First American Real Estate Solutions LLC, Docket No. C-3849 (1999);
U.S. v. Unicor Funding, Inc., Civ. No. 99-1228 (C.D. Cal. 1999); U.S. V. Equifax, No1:00 CV-
0087 (C.D. Ga. 2000); U.S. v. Experian, No. 3-00CV0056-L (N.D. Tex. 2000); U.S. v.
TransUnion, 00C 0235 (N.D. Ill. 2000); U.S. v. Performance Capital Management, Inc., No.
01-1047 (C.D. Cal. 2001); U.S. v. DC Credit Serv. Inc., Civ. No. 02-5115 (C.D. Cal. 2002);
Quicken Loans Inc. Docket No. 9304 (Apr. 8, 2003).
  FTC has also investigated landlord’s compliance with their duty to provide FCRA required
notices to consumers who suffered adverse action based on their consumer reports in
connection with apartment rental applications. The Commission did not bring any formal
actions, but published a consumer alert and a business education brochure for landlords
that resulted from this enforcement effort.

Page 14                                                                      GAO-03-1036T
                            To date, no comprehensive assessments have addressed the impact of the
Impact of 1996 FCRA         1996 FCRA credit report accuracy amendments or the potential effects
Amendments on               inaccuracies have had on consumers. In addition, because it has not
                            conducted surveys, FTC was not able to provide overall trend data on the
Credit Report               frequency of errors in credit reports. Industry officials as well as two
Accuracy and the            studies we reviewed suggest that errors and inaccuracies in credit reports
                            have the potential to both help and hurt individual consumers, while in
Potential Effects of        some instances errors or inaccuracies may have no effect on the
Errors on Consumers         consumer’s credit score. The impact of any particular error or inaccuracy
Is Not Fully Known          in a particular credit report will be dependent on the unique and specific
                            circumstances of the consumer.

Information on the Impact   Data on the impact of the 1996 FCRA amendments on credit report
of FCRA Amendments on       accuracy was not available. For instance, we could not identify impact
Credit Report Accuracy      information from the literature we reviewed and industry officials with
                            whom we spoke said they did not collect such data. Furthermore, FTC
Was Not Available           could not provide overall trend data but did provide FCRA-related
                            consumer complaint data involving CRAs.

                            FTC staff could not say what the trend in the frequency of errors in credit
                            reports has been since the 1996 amendments because that data is not
                            available. However, FTC officials provided consumer complaint data that
                            shows from 1997 through 2002, the number of FCRA complaints involving
                            CRAs received annually by FTC increased from 1,300 to almost 12,000. The
                            most common complaints cited against CRAs in 2002 pertained to the
                            violations are listed below:

                            •   Provided inaccurate information (5,956 complaints);

                            •   Failed to reinvestigate disputed information (2,300 complaints);

                            •   Provided inadequate phone help (1,291 complaints);

                            •   Disclosed incomplete/improper credit file to customer (1,033
                                complaints); and

                            •   Improperly conducted reinvestigation of disputed item (771

                            Consumer complaint data involving CRAs and FCRA provisions represent
                            3.1 percent of the total complaints FTC received directly from consumers
                            on all matters in 2002. The FTC staff explained that their knowledge was

                            Page 15                                                        GAO-03-1036T
                            limited to complaints that came into the agency and that they did not
                            conduct general examinations or evaluations that would enable them to
                            project trends. FTC staff cautioned that it would not be appropriate to
                            conclude that since the complaints against CRAs were on the rise,
                            accuracy of credit reports was deteriorating. They stated that the increase
                            in the number of complaints could be due to greater consumer awareness
                            of FTC’s role with respect to credit reporting, as well as a general trend
                            towards increased consumer awareness of credit reporting and scoring.

Errors May Result in Both   CRAs and the literature suggest that credit-reporting errors could have
Positive and Negative       both a positive and negative effect on consumers. One CRA stated that
Impacts on Consumers        errors occur randomly and may result in either an increase, decrease, or
                            no change in a credit score. Another CRA stated that information
                            erroneously omitted from a credit report such as a delinquency, judgment,
                            or bankruptcy filing would tend to raise a credit score while that same
                            information erroneously posted to the report would tend to lower the
                            score. The Consumer Federation of America study cited earlier also
                            analyzed 258 files to determine whether inconsistencies were likely to
                            raise or lower credit scores. In approximately half the files reviewed (146
                            files, or 57 percent), the study could not clearly identify whether
                            inconsistencies in credit reports were resulting in a higher or lower score.
                            The study determined that in the remaining 112 files there was an even
                            split between files that would result in a higher or lower score. The
                            Federal Reserve Bulletin article previously mentioned also concluded that
                            limitations in consumer reporting agency records have the potential to
                            both help and hurt individual consumers. The article further stated that
                            consumers who were hurt by ambiguities, duplications, and omissions in
                            their files had an incentive to correct them, but consumers who were
                            helped by such problems did not.

Impact of Errors May Be     Industry officials and the literature we reviewed suggested that the impact
Influenced by Other         of an error in a consumer’s credit report was dependent on the specific
Factors in a Credit File    circumstance of the information contained in a credit file. CRA and data
                            furnisher officials further pointed out that a variety of factors such as
                            those identified by Fair Isaac, a private software firm that produces credit
                            score models, might impact a credit score.16 According to the Fair Isaac

                              Fair Isaac Corporation produces software used by many consumer reporting agencies,
                            including the three main U.S. consumer agencies, to produce FICO scores, which according
                            to industry sources, is a commonly used credit score in the United States.

                            Page 16                                                                  GAO-03-1036T
               Web site, their credit score model considers five main categories of
               information along with their general level of importance to arrive at a
               score. These categories and their respective weights in determining a
               credit score include payment history (35 percent), amounts owed (30
               percent), length of credit history (15 percent), types of credit in use (10
               percent) and new credit (10 percent). As such, no one piece of information
               or factor alone determines a credit score. For one person, a given factor
               might be more important than for someone else with a different credit
               history. In addition, as the information in a credit report changes, so does
               the importance of any factor in determining a credit score. Fully
               understanding the impact of errors on consumer’s credit scores would
               require access to consumer credit reports, discussions with consumers to
               identify errors, and discussions with data furnishers to determine what
               impact, if any, correction of errors might have on decisions made based on
               the content of a credit report.

               The lack of comprehensive information regarding the accuracy of
Observations   consumer credit reports inhibits any meaningful discussion of what more
               could or should be done to improve credit report accuracy. Available
               studies suggest that accuracy could be a problem, but no study has been
               performed that is representative of the universe of credit reports.
               Furthermore, any such study would entail the cooperation of the CRAs
               data furnishers, and consumers to fully assess the impact of errors on
               credit scores and underwriting decisions. Because of the importance of
               accurate credit reports to the fairness of our national credit system, it
               would be useful to perform an independent assessment of the accuracy of
               credit reports. Such an assessment could be conducted by FTC or paid for
               by the industry. The assessment would then form the basis for a more
               complete and productive discussion of the costs and benefits of making
               changes to the current system of credit reporting to improve credit report

               Another option for improving the accuracy of credit reports would be to
               create the opportunity for more reviews of credit reports by consumers.
               One way this could be accomplished would be to expand the definition of
               what constitutes an adverse action. Currently, consumers are only entitled
               to receive a free copy of their credit reports when they receive adverse
               action notices for credit denials or if they believe that they have been the
               victim of identity theft. When consumers see their credit reports, they have
               a chance to identify errors and ask for corrections to ensure the accuracy
               of their credit reports. Expanding the criteria for adverse actions to
               include loan offers with less than the most favorable rates and terms
               would likely increase the review of credit files by consumers. Such added

               Page 17                                                        GAO-03-1036T
                  review of credit files would in all likelihood help to further ensure the
                  overall accuracy of consumer credit reports. However, the associated
                  costs to the industry would also need to be considered against the
                  anticipated benefits of increasing consumer access to credit reports.

                  For further information regarding this testimony, please contact Harry
Contacts and      Medina at (415) 904-2000. Individuals making key contributions to this
Acknowledgement   statement include Janet Fong, Jeff R. Pokras, Mitchell B. Rachlis, and
                  Peter E. Rumble.

                  Page 18                                                          GAO-03-1036T
This is a work of the U.S. government and is not subject to copyright protection in the
United States. It may be reproduced and distributed in its entirety without further
permission from GAO. However, because this work may contain copyrighted images or
other material, permission from the copyright holder may be necessary if you wish to
reproduce this material separately.
                         The General Accounting Office, the audit, evaluation and investigative arm of
GAO’s Mission            Congress, exists to support Congress in meeting its constitutional responsibilities
                         and to help improve the performance and accountability of the federal
                         government for the American people. GAO examines the use of public funds;
                         evaluates federal programs and policies; and provides analyses,
                         recommendations, and other assistance to help Congress make informed
                         oversight, policy, and funding decisions. GAO’s commitment to good government
                         is reflected in its core values of accountability, integrity, and reliability.

                         The fastest and easiest way to obtain copies of GAO documents at no cost is
Obtaining Copies of      through the Internet. GAO’s Web site (www.gao.gov) contains abstracts and full-
GAO Reports and          text files of current reports and testimony and an expanding archive of older
                         products. The Web site features a search engine to help you locate documents
Testimony                using key words and phrases. You can print these documents in their entirety,
                         including charts and other graphics.
                         Each day, GAO issues a list of newly released reports, testimony, and
                         correspondence. GAO posts this list, known as “Today’s Reports,” on its Web site
                         daily. The list contains links to the full-text document files. To have GAO e-mail
                         this list to you every afternoon, go to www.gao.gov and select “Subscribe to e-mail
                         alerts” under the “Order GAO Products” heading.

Order by Mail or Phone   The first copy of each printed report is free. Additional copies are $2 each. A
                         check or money order should be made out to the Superintendent of Documents.
                         GAO also accepts VISA and Mastercard. Orders for 100 or more copies mailed to a
                         single address are discounted 25 percent. Orders should be sent to:
                         U.S. General Accounting Office
                         441 G Street NW, Room LM
                         Washington, D.C. 20548
                         To order by Phone:     Voice:    (202) 512-6000
                                                TDD:      (202) 512-2537
                                                Fax:      (202) 512-6061

To Report Fraud,
                         Web site: www.gao.gov/fraudnet/fraudnet.htm
Waste, and Abuse in      E-mail: fraudnet@gao.gov
Federal Programs         Automated answering system: (800) 424-5454 or (202) 512-7470

                         Jeff Nelligan, Managing Director, NelliganJ@gao.gov (202) 512-4800
Public Affairs           U.S. General Accounting Office, 441 G Street NW, Room 7149
                         Washington, D.C. 20548