FTC Credit Report Error Study: What the Numbers Actually Mean

A primary-source reading of the FTC’s landmark credit-report accuracy studies—what “one in five” measured, what the five-percent finding means, and how to use the research responsibly.

The most repeated statistic about credit-report errors is that one in five consumers has an error. That sentence comes from a major Federal Trade Commission study, but it is often stripped of the methodology and the outcome measures that make it meaningful. An error identified by a participant, a report modified after a dispute, a score change, and a change large enough to affect a credit-risk tier are not the same event.

The FTC’s work is valuable because it followed consumers through the actual dispute process and compared reports and scores before and after modifications. It does not prove that every fifth person has a damaging error, that every disputed item was legally inaccurate, or that filing broad disputes will raise a score. Reading the study carefully produces a more useful conclusion: check all three reports, investigate specific fields, document genuine inaccuracies, and measure the result rather than assuming what a headline means.

Why the FTC conducted the study

Congress directed the FTC under Section 319 of the Fair and Accurate Credit Transactions Act of 2003 to study the accuracy and completeness of consumer credit reports. The agency issued interim reports and then a national study using consumers, the three nationwide credit reporting agencies, furnishers, and FICO scoring information. The fifth interim report was dated December 2012 and released publicly in February 2013.

Credit-report accuracy matters because the data can influence credit decisions and pricing, insurance decisions where permitted, housing applications, and some employment decisions. A report is not a score, but scoring systems use report data. An error can therefore be important even when it does not immediately change a score, while another correction may alter a score without changing whether a particular lender approves an application.

How the 2012 study was designed

The study included 1,001 participants who reviewed 2,968 credit reports with trained study associates. Participants were selected to match demographic and credit-score characteristics of consumers with credit histories. They obtained reports from the three nationwide credit reporting agencies and examined information that might be inaccurate.

When participants identified a potential material error, they were encouraged to use the Fair Credit Reporting Act dispute process. The study then compared the original report with the report produced after the dispute process. When a credit reporting agency modified information, the study used rescoring to estimate how the change affected the participant’s credit score and risk tier.

That design is stronger than simply asking people whether they believe a report contains a mistake. It connects four stages: suspected error, dispute, report modification, and score consequence. It still cannot answer every question. A modification is not always a formal finding that the original data violated the law, and a score effect depends on the scoring model and the rest of the person’s file.

The main findings, separated correctly

• About one in five participants disputed information and had at least one report modified.
• About one in four identified potential errors that might affect a score.
• Thirteen percent experienced some score change connected to a dispute-related modification.
• Approximately five percent had a maximum score change of more than 25 points.
• Roughly one in 250 had a maximum score change of more than 100 points.
• The FTC concluded that five percent had errors that may affect the likelihood of receiving credit or the terms offered.

These percentages use different denominators and thresholds. “One in five” refers to consumers who disputed information and received a modification on at least one report. It does not mean that 20 percent of every account, field, or report was wrong. It also does not mean that all modifications produced a meaningful score increase.

The five-percent result is narrower and often more relevant to lending consequences. It concerns errors associated with a score change that could move a consumer into a different credit-risk classification. Even then, the study describes possible effects on credit likelihood or terms. It does not promise that a particular lender would approve the application, lower the rate, or use the same score or tier system.

What “error” meant in the study process

Participants and study associates reviewed reports for potentially material inaccuracies. The formal dispute process then gave the credit reporting agencies and data furnishers an opportunity to investigate. If the report changed afterward, researchers measured the change. This approach recognized that consumer identification is the beginning of the process, not the final legal determination.

Some disagreements are straightforward: an account belongs to someone else, a payment is assigned to the wrong month, a balance is duplicated, or a closed account is shown as open. Others require records and context, such as the correct date of first delinquency, responsibility for a joint account, payment application during a servicing transfer, or the status of debt after a settlement or bankruptcy.

The study should not be used to justify disputing every negative item. Accurate late payments, balances, collections, charge-offs, and other lawful negative information do not become inaccurate because errors exist elsewhere in the reporting system. The FTC’s current consumer guidance distinguishes information that is wrong or incomplete from information that is accurate but unfavorable.

Why the score outcomes varied

A correction can have no score effect, a small effect, or a larger effect because scoring models evaluate the entire file. Correcting an address spelling is important for identity and matching but may not affect a score. Correcting a serious delinquency, ownership error, balance, limit, or duplicated account may matter more, depending on the rest of the data and the model.

The FTC reported maximum score changes across a participant’s reports, not a universal expected gain from a dispute. A consumer with several established accounts may respond differently from someone with a thin file. The same correction may also produce different results across bureaus when the underlying reports contain different accounts or balances.

Score change is only one outcome. A corrected account owner, date, status, or balance can matter during manual underwriting or identity-theft recovery even if a score barely moves. Conversely, a numerical increase does not guarantee better terms because lenders apply product rules, income and debt analysis, collateral requirements, fraud checks, and their own risk standards.

What the 2015 follow-up added

The FTC’s final follow-up focused on 121 consumers from the earlier study who had at least one unresolved dispute and participated in a follow-up survey. This was not a new nationally representative sample of all consumers. It was a specific subgroup selected because an earlier disagreement remained unresolved.

Thirty-seven of those consumers, or 31 percent, said they had come to accept the originally disputed information as correct. Eighty-four, nearly 70 percent, still believed at least some disputed information was inaccurate. Among those 84, 38 planned to continue disputing, 42 planned to abandon the dispute, and four were undecided.

Those results reveal a problem that a score-change statistic cannot capture: unresolved belief and process fatigue. Half of the consumers who still believed information was inaccurate planned to stop. That does not prove the data was wrong, but it suggests that dispute outcomes and explanations can leave consumers uncertain about what was investigated, what evidence controlled, and why the result occurred.

The follow-up also examined whether negative information removed after disputes later reappeared. The FTC found two instances, about one percent of the consumers in the relevant group. The study recommended that credit reporting agencies improve how they notify consumers about investigation results and continue consumer education about report review and dispute rights.

Important limits of the headline statistics

• The study population included consumers with credit histories; it was not a count of every U.S. adult.
• The three reports from one participant are related observations, not three unrelated consumers.
• A participant’s suspected error was not automatically proven inaccurate.
• A report modification did not always change a score.
• A score change did not automatically change an approval, price, or real-world lender decision.
• The study used scoring and risk-tier methods relevant to its design and time; lenders may use other models and policies.
• The 2015 follow-up examined a small, selected subgroup with unresolved disputes, not the full original sample.
• The research does not measure the results of indiscriminate, repeated, or unsupported disputes.

A sound article should therefore avoid claims such as “20 percent of credit reports are wrong,” “disputing guarantees a score increase,” or “the FTC proved bureaus cannot verify accounts.” Those statements collapse distinct measures and overstate the evidence. The defensible statement is that a meaningful share of study participants found potential inaccuracies, obtained modifications through the dispute process, and in a smaller share of cases saw score changes large enough to matter to credit-risk classification.

How consumers can apply the research

First, review all three reports because creditors do not always furnish identical information to every bureau, and matching or timing differences can appear. The FTC says AnnualCreditReport.com is the authorized source for the free reports provided by law, and the nationwide bureaus currently allow free weekly access through that site.

Second, classify the issue before disputing it. Record the company, account identifier, bureau, field, displayed value, correct value, and evidence. A dispute about account ownership needs different proof from a dispute about a payment date, balance, account status, or obsolete information.

Third, ask for a correction that matches the error. If one payment month is wrong, requesting deletion of a positive account may cause unnecessary harm. If the account does not belong to you, ownership and identity-theft procedures may be more appropriate than arguing about the balance.

Fourth, include copies of documents and preserve a complete record. FTC guidance recommends identifying each mistake, explaining why it is wrong, marking the relevant report entry, and providing supporting documents. Consumers may dispute with the bureau displaying the issue and with the business that furnished the information.

Fifth, inspect the result instead of relying on a status label. Compare the original and updated reports field by field. Confirm the account, balance, status, dates, payment-history grid, comments, and dispute notation. Check every bureau that displayed the issue because a correction on one file does not prove the other files changed.

A practical evidence hierarchy

• Strong direct records: account statements, payment ledgers, bank confirmations, court orders, identity-theft reports, settlement documents, and written creditor acknowledgments.
• Useful contextual records: emails, secure messages, call references, hardship agreements, servicing-transfer notices, and dated screenshots.
• Weak by themselves: a score drop, a monitoring alert, an unsupported template, or a statement that an item is “unverified” without identifying the inaccurate field.
• Irrelevant volume: unrelated statements or repeated legal citations that do not prove the specific fact in dispute.

Evidence quality matters because the investigation must connect a claim to a record. A short timeline with the due date, payment date, posting date, reporting month, and attached proof is often easier to investigate than a long packet of unrelated documents. Redact unrelated personal and account information while leaving enough detail to identify the issue.

When a dispute result does not resolve the issue

A verified or updated response is not necessarily the end, but sending the identical dispute repeatedly may not help. Determine whether the response addressed the precise field and evidence. Request relevant account records from the furnisher, clarify contradictions, or submit new documentation rather than merely restating the conclusion.

The FTC explains that a bureau generally has 30 days to investigate a dispute and can stop if it reasonably considers the dispute frivolous or irrelevant, while providing notice and a reason. The bureau sends relevant evidence to the furnisher, and the furnisher investigates and reports back. If the furnisher finds the reported information inaccurate, it must notify the nationwide bureaus so they can correct the files.

For persistent, well-documented inaccuracies, consumers may consider a CFPB complaint or advice from a qualified consumer-law attorney. Identity theft calls for the specialized recovery steps at IdentityTheft.gov. If the information is accurate and the real problem is unaffordable debt, a payoff plan, hardship option, or nonprofit credit counseling may address the problem more directly than a dispute.

A reader’s checklist for interpreting any credit study

• Who was studied, and how were participants selected?
• What is the denominator: consumers, reports, accounts, disputes, or score changes?
• Was an error merely alleged, modified, confirmed, or legally adjudicated?
• Was the outcome a data correction, score change, risk-tier change, approval, or price difference?
• Which scoring model and time period were used?
• Were results nationally representative or limited to a selected subgroup?
• What uncertainty, attrition, and follow-up limitations apply?
• Does the article report absolute counts as well as percentages?
• Does the conclusion stay within the outcomes the study actually measured?

This framework helps separate useful research from sales claims. A company may cite “one in five” to create urgency, then imply a guaranteed deletion or score increase that the FTC never found. The responsible response is to read the underlying measure, identify whether your own report contains a specific error, and choose an evidence-based action.

Sources and methodology

Primary sources used: FTC Fifth Interim Report page and study summary, https://www.ftc.gov/reports/section-319-fair-accurate-credit-transactions-act-2003-fifth-interim-federal-trade-commission-report ; FTC 2013 release describing the national study and sample, https://www.ftc.gov/news-events/news/press-releases/2013/02/ftc-study-five-percent-consumers-had-errors-their-credit-reports-could-result-less-favorable-terms ; FTC 2015 follow-up release, https://www.ftc.gov/news-events/news/press-releases/2015/01/ftc-issues-follow-study-credit-report-accuracy ; FTC dispute guidance, https://consumer.ftc.gov/articles/disputing-errors-your-credit-reports ; FTC free-report guidance, https://consumer.ftc.gov/articles/free-credit-reports ; 15 U.S.C. §1681i, https://www.law.cornell.edu/uscode/text/15/1681i ; and CFPB credit-report resources, https://www.consumerfinance.gov/consumer-tools/credit-reports-and-scores/ .

This article compares the FTC’s reported sample sizes, stages, percentages, and follow-up findings. It does not independently reproduce the FTC’s raw-data analysis, and it does not treat a report modification as a court finding. Figures are presented as the agencies reported them, with denominators and limitations described wherever they materially affect interpretation.

This material is general education, not legal, credit, or financial advice. Study findings cannot predict the result of an individual dispute, score change, application, or price. Verify current procedures with official sources and consult a qualified professional when the facts or consequences are significant.

The bottom line

The FTC research supports regular review and precise disputes, not panic and not mass challenges. One in five participants obtained a modification after disputing at least one report, thirteen percent experienced some score change, and a smaller five-percent group had errors associated with changes that could affect credit-risk classification. The value of those numbers is not a promise. It is a reason to inspect your own three reports, document what is actually wrong, and verify the correction.

This article is general information, not legal or financial advice. Results vary by credit profile.