FACTA National Study - SS - 10-16-09 - Part A - FINAL_mtd

Supporting Statement
FTC National Study Pursuant to Section 319 of the
Fair and Accurate Credit Transactions Act of 2003
To prepare the next in a series of reports required by the Fair and Accurate Credit
Transactions Act of 2003 (“FACT Act” or “FACTA”), the Federal Trade Commission (the
“Commission” or “FTC”) plans to conduct a national study of credit report accuracy. This study
is a follow up to the FTC’s two prior pilot studies (OMB Control No. 3084-0133) that were
undertaken to evaluate a potential design for a national study.1 The FTC is now requesting that
the OMB reinstate that clearance so that it can conduct the current study.
As noted in prior submissions regarding the pilot studies, accurate credit reporting is very
important to consumers and to the industry. For consumers, credit reports – and the associated
credit scores – may be the key to getting approval for such items as a mortgage, an automobile
loan, a major credit card, a job, the rental of an apartment, phone service, insurance coverage,
and other financial products or services. For the industry, credit reports provide lenders
information about a borrower’s credit history so they can estimate default risk and may thus
tailor interest rates and other credit terms to the risk that is presented by the borrower.
A.

JUSTIFICATION
1. Necessity for Information Collection

Among other purposes, the FACT Act amended the Fair Credit Reporting Act (“FCRA”)
to enhance the accuracy of consumer reports. See Pub. L. 108-159 (2003). Section 319 of the
FACT Act directed the Commission to study the accuracy and completeness of information in
consumers’ credit reports and to consider methods for improving the accuracy and completeness
of such information. The Act also required the FTC to issue a series of biennial reports to
Congress over a period of eleven years. The first interim report was due on December 4, 2004
(one year after enactment), and five additional reports were thereafter required at two-year

1

The OMB granted the FTC’s request for discontinuance of that prior clearance on
November 25, 2008.
October 2009

2
intervals. The FTC has submitted three reports thus far.2 The next one is due on December 4,
2010.
The proposed national study seeks to use a large representative sample of credit reports
so that we may draw inferences, up to a certain level of statistical confidence, about the accuracy
of credit reports in general. The Commission has found that, although prior approaches to the
study of credit report accuracy provide some useful information, none provides a reliable and
comprehensive assessment. This evaluation included a consideration of prior studies, published
reports of other governmental agencies, and an FTC roundtable convened specifically on credit
report accuracy. None of the existing sources of information relied on the participation of all
three of the key stakeholders in the credit reporting process: consumers, data furnishers, and the
Credit Reporting Agencies (or the “CRAs,” which include Equifax, Experian, and TransUnion
for our purposes).3 The proposed national study seeks to address certain shortcomings of prior
approaches. (See also Specifications 2 & 4 below).
2. How Data Will Be Collected and Used
The study is designed as follows:
•

Using a pre-tested solicitation procedure that invites consumers to participate in the study
(described in detail in Part B), a random sample of approximately 1,000 participants will
review their credit reports with an expert who helps identify potential errors.4

•

The consumers then dispute potential errors that the expert believes could have a material
effect on a consumer’s credit standing.

•

The study tracks the outcome of disputes filed with CRAs in accordance with the FCRA
dispute resolution process (explained further in next two paragraphs).
2

Report to Congress Under Sections 318 and 319 of the Fair and Accurate Credit
Transactions Act of 2003, Federal Trade Commission, December 2004, and Report to Congress
Under Sections 319 of the Fair and Accurate Credit Transactions Act of 2003, Federal Trade
Commission, December 2006 and 2008 (respectively). The three reports may be accessed at the
FTC’s Web site.
December 2004 Report: http://www.ftc.gov/reports/index.htm#2004.
December 2006 Report: http://www.ftc.gov/reports/FACTACT/FACT_Act_Report_2006.pdf ;
December 2008 Report: http://www.ftc.gov/opa/2008/12/factareport.shtm.
3

A data furnisher is any entity, such as banks, lenders, or other parties, who
provide consumer report information to CRAs.
4

The FTC has not yet designated a contractor for the study. Upon OMB clearance,
the FTC will seek competitive bids for the study and then select a contractor.
October 2009

3
•

The data collected from that process will be used to assess the frequency of confirmed
material errors via the dispute process and to categorize such errors by type and
seriousness in terms of potential consumer harm.

The methodology for reviewing credit report information and establishing material errors
will be essentially the same as in the second pilot study. The contractor will help participants
obtain their credit reports from Equifax, Experian, and TransUnion, and then conduct an in-depth
review of these reports with each participant. Specific criteria for materiality will be employed.5
If the consumer alleges a material error, an FCRA dispute will be filed in order to obtain a
review of the challenged items by data furnishers and CRAs.6
When an item is disputed with a CRA, the latter sends the disputed information to the
relevant data furnisher for review. By instruction of the data furnisher, one of the following
outcomes may occur: delete the item, change or modify the item (specifying the change), or
maintain the item as originally reported. A CRA may also delete a disputed item due to
expiration of the statutory time frame.7 The study will use the FCRA dispute process and its
related outcomes to categorize outcomes either as confirmed error or unresolved. The study will
focus on confirmed material errors. Specifically, an outcome is counted as a confirmed material
error via the dispute process if the consumer alleges a material error with a CRA/data furnisher,
5

In the second pilot study, the contractor used the following criteria for materiality:
the consumer had a credit score less than 760 (a cutoff widely used to identify consumers with
lowest credit risk and for extending credit on most favorable terms) AND the consumer alleged
an error regarding any of the following matters: (i) negative items (such as late payments); (ii)
public derogatories (such as bankruptcy); (iii) accounts sent to collection; (iv) number of
inquiries for new credit; (v) outstanding balances not attributable to normal monthly reporting
variation; (vi) accounts on the report not belonging to the person who is the subject of the report;
or (vii) duplicate entries of the same information (e.g., late payments or outstanding obligations)
that were double-counted in the reported summaries of such items.
In the proposed national study, we do not plan to use any cutoff score for materiality, but
will retain the stated categories for identifying a dispute that is material to creditworthiness.
6

The consumer will submit a signed dispute letter to a CRA, while the contractor
may help the consumer in preparing the letter. The contractor will not encourage the filing of
immaterial disputes. Nonetheless, the contractor will help prepare dispute letters for all
consumers who want to dispute items on their credit reports, whether or not the item is material
to creditworthiness. (The FCRA permits a consumer to dispute any credit report information
that the person believes to be inaccurate.)
7

The FCRA limits the process to 30 days, but the time may be extended to 45 days
if a consumer submits relevant information during the 30-day period.
October 2009

4
and upon review, the CRA/data furnisher changes or deletes the item in keeping with the
consumer’s allegation. In cases where the consumer maintains one thing and the lender another,
the study renders no conclusion and leaves such instances as unresolved.8
The second pilot study indicated that it is feasible to measure the impact of confirmed
material errors by rescoring the credit report. More specifically, if a consumer alleges material
error(s), the contractor will mark the credit reports (the frozen files)9 with explanations of the
discrepancies and employ a credit score expert to rescore the report (just provisionally, using the
consumer’s information as though it were accurate). If via the FCRA dispute process, changes
are made by CRAs/data furnishers in keeping with the consumer’s allegations, the rescore of the
frozen file quantifies the impact of the confirmed error(s); i.e., the difference between the rescore
of the frozen file and the original score of the credit report will be used to measure the impact of
the error(s) on the consumer’s credit standing.
3.

Use of Information Technology

The study participant’s credit reports will be obtained via the Internet. Additional use of
electronic methods pursuant to the Government Paperwork Elimination Act, Pub. L. No. 105277, Title XVII, 112 Stat. 2681-749, would be impractical. The contractor will review reports
with participants over the telephone. The review of credit report information will involve a
number of back and forth questions and answers – a procedure facilitated by telephone contact.
4.

Efforts to Identify Duplication/Availability of Similar Information

FTC staff have reviewed all prior studies of credit report accuracy and also conducted a
roundtable on June 30, 2004.10 Roundtable participants included various consumer-interest
groups, industry representatives, academic researchers, and participants from other government
agencies. Notably, all of the principal parties involved in the prior studies on credit report
accuracy were at the roundtable. The FTC’s December 2004 Report (at 21-31) reviews all prior

8

See also 2008 Report to Congress (at 3-4), where this same methodology is noted.
Despite its limitations, use of the FCRA dispute process appears to be the only feasible way of
performing a national study, in view of the enormous difficulty and cost of attempting to
ascertain the ultimate accuracy regarding alleged errors.
9

The files are called frozen because no new credit information is added to the
consumer’s original credit reports obtained in the study; any rescoring thus applies only to
potential changes or actual changes that are directly related to the contractor’s review.
10

A full transcript of the proceedings, agenda, list of participants, and a further
summary of the roundtable are available at
http://www.ftc.gov/be/workshops/methodologiesaacc/index.htm.
October 2009

5
studies and a related GAO report that also assessed the state of knowledge regarding credit
report accuracy.11 In summary, the Commission concludes (at iii):
Prior studies of consumer report accuracy and completeness essentially fall into
three categories – consumer surveys, studies based on dispute data statistics, and
studies based on anonymous data provided by the CRAs about a large number of
individual consumers. The FTC’s review of these studies determined that,
although each approach provides some useful information about credit report
accuracy and completeness, none provides a comprehensive view. Indeed, none
of the existing studies relied on the participation of all three of the key stakeholders in the credit reporting process: consumers, data furnishers, and the
CRAs.
Since the time of that report, the FTC has conducted the two pilot studies referenced above, and
there have been no further studies of credit report accuracy.
5.

Efforts to Minimize Small Organization Burden

Not applicable. The study participants are individual consumers.
6.

Consequences to Federal Program and Policy Activities/Obstacles to
Reducing Burden

If this national study is not undertaken, the FTC will not have the necessary information
to carry out the congressional mandate of section 319 of the FACT Act.
7.

Circumstances Requiring Collection Inconsistent with Guidelines

The collection of information in the proposed survey is consistent with all applicable
guidelines contained in 5 C.F.R. § 1320.5(d)(2). BE staff in consultation with the contractor will
ensure a nationally representative sample of 1000 individuals as study participants broadly
representing the characteristics of U.S. adult members of households who have credit histories.
Part B of this Supporting Statement has further detail about survey methodology.
8.

Public Comments Sought/Consultation Outside the Agency

As referenced in our response to Specification 4 above, the FTC conducted a roundtable
which (among other things) presented a proposed design to study credit report accuracy using an
expert-assisted survey approach. This design was reviewed by the roundtable participants; the

11

GAO Report, No. GAO-03-1036T, Consumer Credit: Limited Information Exists
on the Extent of Credit Report Errors and their Implication for Consumers (July 31, 2003).
October 2009

6
latter included various consumer-interest groups, industry representatives, academic researchers,
and participants from other governmental agencies.
As required by 5 C.F.R. § 1320.8(d), the FTC published a notice seeking public comment
on the study and its proposed information collection request, 74 Fed. Reg. 35,191 (July 20,
2009), attached. The FTC received 3 comments, which are available at the FTC’s Web site
(http://www.ftc.gov/os/comments/facta319study/index.shtm). The commenters were the
Consumer Data Industry Association (CDIA), Mr. Chris Hoofnagle of the Berkeley Center for
Law & Technology, and Privacy Times submitted by Mr. Evan Hendricks (and signed by
additional parties). Pursuant to the OMB regulations that implement the PRA (5 C.F.R. Part
1320), the FTC is providing a second opportunity for public comment while seeking OMB
clearance for the study.
No comments addressed the cost and hour burden estimates nor challenged the need or
the importance of the study. Overall, the comments addressed the qualifications of any potential
contractor, the universe of participants to be covered by the study, and some concerns about
specific parts of the methodology of the study.
The comment from the CDIA, submitted by Mr. Stuart Pratt, is generally supportive
while expressing certain concerns. The CDIA (at 2) believes that the FTC’s use of consumer
interviews combined with the FCRA dispute process “compares favorably to the flawed
methodology employed by consumer groups in their ‘studies’ of credit report accuracy.” The
CDIA recommends the FTC highlight these differences in its communications about the study.
As discussed above, in its 2004 Report to Congress, the FTC reviewed all prior studies and
created a design for a national study to specifically address certain shortcomings of prior
approaches. In an upcoming report to Congress about the results of the national study, the FTC
will again point out the ways in which the study has addressed prior shortcomings.
The CDIA (at 1) expresses the concern that the methodology may over-sample
consumers with low credit scores; it recommends the ultimate study group have credit scores
that “are reflective of the distribution of scores in the databases of the nationwide consumer
reporting agencies.” FTC staff agrees with the stated recommendation. As discussed in the
referenced FR notice of July 20, 2009 (at 35194), the second pilot study confirmed that purely
random sampling of potential participants yields too few actual participants with low credit
scores. In the national study, invitation letters will be sent in progressive waves, and
proportionally more invitation letters will be sent to groups having lower credit scores. Based on
our knowledge from the second pilot study and also the knowledge that will be gained from the
response rates of the earlier waves of letters in the national study, FTC staff will be able to adjust
subsequent waves of letters to the potential respondents in certain score ranges so as to achieve a
total set of respondents whose credit scores are indeed in line with national norms. (See also
discussion below in Part B, Specification 2).

October 2009

7
It is possible, although not very likely, that the methodology could render a set of
respondents having too many people with low scores. However, since the national distribution
of credit scores is known (with great refinement), there are recognized statistical procedures to
ultimately correct any over-sampling of low scores (should it occur) and to ensure the statistical
reliability of the results, including the reliability of the results for the population as a whole.12
The CDIA (at 1) also expresses a concern for reaching out to consumers who do not have
Internet access. As explained in the FR notice of July 20, 2009 (at 35195), participants will use
the Internet to register for the study. However, for those who may not have Internet access, the
contractor will also have a procedure to mail the appropriate disclosures and the study steps to
the respondent and then receive enrollment information and the consumer’s signed consent in
paper form. The in-depth review of the credit reports with the participants will occur over the
telephone and does not require Internet access.
The CDIA (at 2) recommends that upon assessing errors by a change in credit score that
is attributable to certain errors, the FTC also include measures on how a change in score would
impact a consumer’s interest rate or other credit decisions; e.g., some changes in credit score
would keep a consumer in the same “band of risk” determined by the lender, while other changes
could place the consumer in a more favorable band of risk and thus allow the lender to proffer a
noticeably better interest rate. FTC staff agrees with this recommendation. In reporting the
results of the study to Congress, staff fully intends to include the type of discussion and
associated measures here indicated.
The comment from the Berkeley Center for Law & Technology, submitted by Mr. Chris
Hoofnagle, strongly supports the FTC’s announced goal of acquiring “1000 participants who as a
group display a diversity on credit scores and on major demographic characteristics in line with
national norms.” The commenter further recommends, regarding the qualifications of any
potential contractor, that the entity be highly qualified to perform consumer surveys and that it
be a neutral entity (i.e., have no stake in the outcome of the study). FTC staff readily concurs
with the expressed concern. The FTC will publically solicit competitive bids for performing the
study in keeping with a detailed scope of work (to be announced). Staff will carefully review the
credentials associated with each bid and proposal and will seek a contractor who is highly
qualified to perform the required work and who has no stake in the outcome of the study.
The comment letter from Mr. Evan Hendricks of Privacy Times (signed by additional
parties) covers several concerns discussed above (e.g., qualifications of the study contractor and
the need for a diverse set of credit scores reflective of national norms). The commenter further
recommends that the study pay special attention to the matter of data matching procedures,
covering such maters as the use of Social Security Numbers and partial matches on consumer

12

See, for example, Harnett, Donald L., Statistical Methods (3rd ed.), AddisonWesley Publishing Co., 1984 (pages 253-254).
October 2009

8
identifiers. The matter of data matching procedures has been reviewed in the 2004 Report to
Congress, and staff does not anticipate that this study will specifically address the internal data
matching procedures used by credit bureaus. However, the contractor will keep a detailed
narrative regarding each participant, including specific errors alleged and their subsequent
disposition. In tabulating the types of confirmed errors via the dispute process, the study will
acquire a great deal of information on the main sources of error in credit reports.13 Further, in
regard to an expressed concern from Mr. Hendricks about recognizing ID theft as an important
source of error, the category of alleged error called “not mine” will be separated into the
subcategories of “mixed file” and “ID theft.”
9.

Payments or Gifts to Respondents

In an effort to maximize the response rate (see also part B, Specification 3),
compensation will be offered to participants as follows. The solicitation letters sent to
individuals with FICO14 credit scores equal to or above 700 will offer any participant $25 upon
completion of their work. Individuals with FICO scores below 700 will be offered $25 upon
agreeing to participate and a further $50 upon completion of their work. The maximum payment
to any individual is thus $75. In addition, participants will receive their credit reports and scores
at no charge.
10. & 11.

Assurances of Confidentiality/Matters of a Sensitive Nature

The information in this sample for which PRA collection clearance is being sought
includes names, addresses, telephone numbers, best times to be contacted further about the study
and credit scores for approximately 1,000 consumers that would register at a designated Web site
(or via U.S. mail) by entering an electronic signature certifying the consumer’s consent to
participate in the study (or via hard copy signature if by U.S. mail). This information is to be
obtained under applicable law and protected from disclosure by, e.g., Exemption 6 of the
Freedom of Information Act, 5 U.S.C. 552. Consumers who agree to participate will sign a
prepared consent form giving the contractor permission to review the consumer’s credit reports.
The information submitted by registrants, as well as any credit reports that individual
participants give the contractor permission to analyze for the study, will be maintained and used
subject to appropriate information security procedures and safeguards (e.g., maintaining
credit-related data separately from personal identifying information, requiring the FTC's
contractors to execute confidentiality agreements, and limiting access to those FTC and

13

See referenced Federal Register Notice at 35193 (note 9) and at 35194 (note 18)
for the types of errors to be tabulated.
14

in 2009.
October 2009

FICO was formerly known as Fair Isaac Corporation until a name change earlier

9
contractor staff who have a need to work with the data). The credit reports that the contractor
reviews with the participant will mask the participant’s account numbers and show just the last 4
digits of any social security number. Also, as explained in Part B, the FTC staff will not have
access to any personally identifying information for any participants or nonrespondents at any
time during the study. Part B also explains how the study methodology is designed to prevent
disclosure of any individual's participation in the study to any credit reporting agency.
If the FTC decides to make the results of the study publicly available (see Specification
16 below), such results would not include any personal information that identifies individual
respondents (i.e., the results would be anonymized).
To the extent, if any, that the Privacy Act, 5 U.S.C. 552(a) may apply, the credit reports
obtained in the study would be part of the agency’s existing system for legal, investigational and
other records (see http://www.ftc.gov/foia/sysnot/i-1.pdf) and, whether or not the Act applies,
the FTC will include a notice consistent with the Act on any information collection forms (e.g.,
the letter sent by the FTC to potential study participants). The letter to potential participants will
also inform them that the contractor is permitted to collect the data only for the purpose of the
FTC’s study and that other uses by the contractor have been prohibited.
12.

Estimated Hours Burden

Consumer participation in the proposed national study will involve an initial preparation
for the in-depth interview and time spent by participants to understand, review, and if deemed
necessary, dispute information in their credit reports. Invitation letters will be sent in
progressive waves in order to obtain approximately 1,000 participants. The recipients of the
letters will be asked to go to a designated Web site for enrollment if they wish to participate.
Registration is expected to take at most 15 minutes per participant.15 The registration process
thus comes to approximately 250 hours (reckoned at 1/4 hour for each of 1,000 consumers).
Regarding the review of the credit reports, FTC staff submits the following estimates
which are based on what we learned during the second pilot study. Some participants prepare
thoroughly in advance of the in-depth interview of their credit reports. In such cases, even

15

At the registration Web site, a person may take the time to read several
disclosures, including a privacy disclosure and an outline of the various steps of the study that
every participant agrees to undertake. The consumer is then asked to enter basic contact
information (e.g., name, address, telephone number, best time to be contacted further about the
study) and to enter an electronic signature certifying the consumer’s consent to participate in the
study. For those who may not have Internet access to register, the contractor will also have a
procedure to mail the appropriate disclosures and study parameters to the respondent and then
receive the enrollment information and the consumer’s signed consent in paper form.
October 2009

10
complicated reports may generally be finished under 30 minutes. Other consumers may not find
time for significant preparation in advance of the in-depth review, and in such cases the
interview could take up to an hour. Participants in the second pilot study reported taking an
average of 69 minutes (median 53 minutes) to prepare for the interview, with 90% of them
taking between 10 and 180 minutes. The interviews themselves took an average of 19 minutes
(median 15 minutes) with 90% taking between 5 and 45 minutes. Overall, the average combined
time for preparation and the interview was about 90 minutes (1.5 hours). For a national study
involving 1,000 consumers, FTC staff thus estimates the burden hours for the review process to
be approximately 1,500 hours (1,000 consumers x 1.5 hours). Adding on the time spent for the
registration process (0.25 hours per participant), the total burden hours come to about 1,750
hours.
13.

Estimated Cost Burden

Participation by the consumer is voluntary. All participants will benefit by receiving
assistance from the contractor in reviewing their credit reports, and identifying and resolving
errors. There are no start-up or monetary costs for the consumer; specifically, participants will
not pay for their credit reports or credit scores.
14.

Estimated Cost to Federal Government

Upon OMB clearance, the FTC will seek competitive bids for the study and select a
contractor. The contract award is yet to be determined.16 In addition, the aggregate of FTC staff
time and costs already expended to design the study and, further, to identify a future contractor
plus time that will be expended to assist the contractor in completing his duties and to analyze
the data, will sum to approximately 1,800 hours of economist and research analyst time. The
associated monetary cost is approximately $117,000. The cost of FTC staff time is necessarily
an estimate because factors in this calculation may vary, including the number of staff members
involved and the amount of time required of each. These estimates include clerical and other
support services.
15.

16

Program Changes or Adjustments

Based on the contract cost for the second pilot study, staff has its own estimate of
the likely contract cost for a national study but is reluctant to publish such a figure in advance of
competitive bidding.
October 2009

11
The burden estimate for the proposed national pilot study survey represents an increase
of 1000 hours from the estimate for the second pilot study.
16.

Plans for Tabulation and Publication

The FTC has not yet determined whether or how the results would be made public in a
report or other publication. Staff believes that it is likely that anonymized results would be made
public in the same manner as in the pilot studies; e.g., a copy of the contractor’s report may be
attached as an appendix to a subsequent FTC Report to Congress under the FACT Act.
17 & 18.
Certification

Display of Expiration Date for OMB Approval/Exceptions to

The FTC will display the expiration date alongside the assigned OMB control number on
the front pages of any documentation given to the participants of the study. There are no
exceptions to certification.

October 2009

12

ATTACHMENT
Fair and Accurate Credit Transactions Act of 2003
P.L. 108-159
SEC. 319. FTC STUDY OF THE ACCURACY OF CONSUMER REPORTS.
(a) STUDY REQUIRED.--Until the final report is submitted under subsection (b)(2), the
Commission shall conduct an ongoing study of the accuracy and completeness of information
contained in consumer reports prepared or maintained by consumer reporting agencies and
methods for improving the accuracy and completeness of such information.
(b) BIENNIAL REPORTS REQUIRED.-(1) INTERIM REPORTS.--The Commission shall submit an interim report to the Congress on
the study conducted under subsection (a) at the end of the 1-year period beginning on the date of
enactment of this Act and biennially thereafter for 8 years.
(2) FINAL REPORT.--The Commission shall submit a final report to the Congress on the study
conducted under subsection (a) at the end of the 2-year period beginning on the date on which
the final interim report is submitted to the Congress under paragraph (1).
(3) CONTENTS.--Each report submitted under this subsection shall contain a detailed summary
of the findings and conclusions of the Commission with respect to the study required under
subsection (a) and such recommendations for legislative and administrative action as the
Commission may determine to be appropriate.

October 2009