What a mixed credit file is

A mixed credit file is a consumer credit file that contains accounts, addresses, judgments or other records belonging to someone else. The Consumer Financial Protection Bureau uses the term in exactly that sense, describing it in its December 2012 study of the credit reporting industry as the “inclusion of accounts or records in a credit file that do not belong to the consumer, commonly called a mixed file.” It is not identity theft, because no one is pretending to be you and no fraud has been committed. It is not a furnisher reporting a real account of yours incorrectly. It is a filing error: the system that decides which incoming record belongs to which existing file made the wrong call, and a stranger’s payment history is now part of the document lenders, landlords and employers read when they decide about you.

That distinction matters more than it sounds, because it determines what the evidence looks like and who has to fix it. When a debt on your report belongs to a different person, the furnisher who reported that debt did nothing wrong; it reported an account that genuinely exists, owed by a genuine borrower. The error happened downstream, inside the reporting agency, when the record was filed. Disputing with the original creditor in that situation produces a verification, because from the creditor’s side the account is accurate. The problem is the assignment, and the assignment is the agency’s.

The matching logic that produces mixed files

Nationwide consumer reporting agencies do not receive files. They receive records — hundreds of millions of monthly tradeline updates, public record entries and inquiry logs — each carrying some set of identifying fields, and each has to be routed to a file. No universal consumer identifier exists in the United States for this purpose, so the routing is done by comparing the identifiers on the incoming record against the identifiers already on file, and scoring the similarity.

The critical design decision is how close a match has to be. A system that demanded exact agreement on every field would refuse most legitimate records, because real data is messy: people move, marry, abbreviate, transpose digits, and creditors truncate and misspell. A system that requires exact agreement produces fragmented files, and fragmented files are also inaccurate — a consumer whose good payment history is scattered across three partial files looks thinner and riskier than she is. So the matching is deliberately tolerant. The CFPB’s 2012 description of how these algorithms operate is that they assign a tradeline to the best available match — even where the identifiers do not all line up. That is the whole mechanism in a sentence. Tolerance for imperfect data is what makes the system work at scale, and it is also precisely what causes a record to land in the wrong file when two people’s identifiers are close enough to fool the score.

Two consequences follow, and both show up repeatedly in litigation. First, the error is not random. It concentrates in exactly the population where identifiers are legitimately similar, which is why the same fact patterns recur for decades across different agencies. Second, once a record is filed to the wrong consumer, the ordinary correction machinery does not necessarily unwind it, because the same matching rules that put it there will put it there again the next time the furnisher sends its monthly update.

Who mixed files happen to most

The circumstances are predictable enough to describe as categories.

Fathers and sons who share a name. This is the single most common pattern and the oldest one documented in the case law. A Junior and a Senior share a first name, middle name and surname; they frequently share an address for part of their lives; and the suffix that distinguishes them is the field most often dropped, truncated or ignored. A generational suffix is exactly one character of distinguishing information in a field many systems treat as optional. Where the two also share a birth month, or where one of them has thin file history, the score can clear the threshold on name and address alone.

Family members at a shared address. Address is heavily weighted in matching because it is genuinely informative — but everyone in a household shares it. Spouses, adult children living at home, and siblings all present the same address, and where surnames match too, only the first name and date of birth are doing any work. Marriage compounds this, because a name change creates a second legitimate name for one person and a period during which both names are in circulation.

Twins. Twins present the worst possible combination of inputs: identical surname, identical date of birth, identical address history for the first two decades of life, often similar or rhyming first names, and Social Security numbers issued at the same time and place, which historically meant numerically adjacent. Almost every high-weight field agrees. The distinguishing information is the first name and a small number of digits.

Common surnames. The scale problem here is easy to underestimate. When the CFPB analyzed name-only matching in 2021 it pointed to the 2010 census figure of roughly 2.4 million people in the United States sharing the surname Smith. Matching on name alone in a population that size is not matching at all. The same holds for the most common Hispanic and East Asian surnames, where a small number of surnames cover a very large share of the population, and where Western name-order conventions and paternal-plus-maternal surname structures create additional ways for a name to be recorded two different ways for the same person.

Thin or shared identifiers. Young adults, recent immigrants, people who have used an Individual Taxpayer Identification Number, and anyone with little credit history has fewer records anchoring the file, so a single misrouted tradeline represents a much larger share of the file and is harder to contradict. Where a Social Security number has been transposed by a data entry error at a single creditor, that transposed number can begin to appear as an alternate identifier associated with the file, and it can pull in records belonging to whoever actually holds it.

Name changes, adoption and estrangement. A person who changed names, or who was adopted, may be matched to records belonging to a biological relative they have no relationship with and no knowledge of. The plaintiff in one of the leading employment screening cases was born under one surname and later carried another; the screening report matched him to a registry record belonging to his estranged biological father, on name alone.

Name-only matching, and why regulators singled it out

The most explicit regulatory statement on mixed files concerns the narrowest and most aggressive version of the matching problem. In an advisory opinion published at 86 Fed. Reg. 62468 on November 10, 2021, titled Fair Credit Reporting; Name-Only Matching Procedures, the CFPB stated its position without hedging: name-only matching is not a procedure that assures maximum possible accuracy, and consumer reporting agencies that use name-only matching violate section 607(b) of the Fair Credit Reporting Act. That is the statutory provision requiring agencies to follow reasonable procedures to assure maximum possible accuracy when they prepare a consumer report.

The opinion did not arrive in a vacuum. It followed two decades of litigation over watchlist screening, where agencies compared a consumer’s name against government lists of sanctioned persons and returned a hit based on name similarity alone, without comparing dates of birth or any other identifier that was sitting right there in the file.

In Cortez v. Trans Union, 617 F.3d 688 (3d Cir. 2010), Sandra Cortez went to buy a car and was held at the dealership for roughly six and a half hours while the transaction stalled. Her file carried an alert associating her with a name on the Treasury Department’s Specially Designated Nationals list — “Cortes Quintero, Sandra” — a person born in June 1971. Cortez was born in May 1944, and the middle name did not match either. The similarity that generated the alert was between the names alone. A jury awarded $50,000 in actual damages and $750,000 in punitive damages; the district court remitted the punitive award to $100,000, and the Third Circuit affirmed both the liability finding and the remitted award.

The same design produced the largest FCRA class judgment on record and then the decision that unwound most of it. In TransUnion LLC v. Ramirez, 594 U.S. 413 (2021), the agency’s OFAC Name Screen Alert compared a consumer’s first and last name against the list — and nothing else. The certified class ran to 8,185 members, of whom 1,853 had a report containing the alert actually disseminated to a third party. The jury awarded $984.22 in statutory damages and $6,353.08 in punitive damages per class member, roughly $60 million in total. The Ninth Circuit, at 951 F.3d 1008, reduced the punitive component to $3,936.88 per member, bringing the judgment to roughly $40 million. The Supreme Court then held, 6–3, that the class members whose reports were never disseminated had not suffered concrete harm and therefore lacked standing to sue in federal court. Justice Kavanaugh’s formulation — no concrete harm, no standing — has governed FCRA class litigation ever since.

Ramirez is worth understanding precisely, because it is routinely described as though it decided that inaccurate files are not actionable. It did not. It decided who may bring the claim in federal court. A consumer whose mixed file was actually sent to a lender, a landlord or an employer is on the disseminated side of that line, and the 1,853 class members on that side had their judgment left intact.

The verdicts and settlements that shaped this area

Mixed file law is unusual in that its foundational case is more than forty years old and its facts would be entirely recognizable today.

Thompson v. San Antonio Retail Merchants Association, 682 F.2d 509 (5th Cir. 1982). William Daniel Thompson Jr. owed $77.25 to Gordon’s Jewelers. That delinquent account was captured into the file of William Douglas Thompson III — a different man, with a different middle name, a different generational suffix and a different Social Security number. The mechanism the court examined was the agency’s “automatic capture” procedure, which pulled records into files without requiring the identifiers to agree. The Fifth Circuit affirmed an award of $10,000 in actual damages plus $4,485 in attorney’s fees. The dollar figures are small by modern standards; the holding is not, because it established that an agency’s matching procedure is itself the thing being judged under section 607(b), and that a procedure which merges records over a suffix mismatch and a Social Security number mismatch is not reasonable.

Philbin v. Trans Union Corp., 101 F.3d 957 (3d Cir. 1996). A father-and-son mixed file, between James R. Philbin Sr. and James R. Philbin Jr. The case is cited constantly, but not for a damages award — there was none. The Third Circuit reversed a grant of summary judgment against the consumer and remanded, holding among other things that a consumer can establish causation for a credit denial without producing testimony from the lender that the mixed report was the reason. That evidentiary point is why the case still matters: it removed a proof requirement that would have made most mixed file claims unwinnable at the summary judgment stage.

Miller v. Equifax Information Services, No. 3:11-cv-01231-BR (D. Or.). This is the case most often cited for a headline number, and the headline number is not what the plaintiff kept. On July 28, 2013 a federal jury in Oregon returned $180,000 in compensatory damages and $18.4 million in punitive damages — about $18.6 million — after evidence that the agency had failed over roughly two years to correct accounts belonging to someone else. In late January 2014 the court remitted the punitive award to $1,620,000, a nine-to-one ratio to the compensatory award, leaving a judgment of approximately $1.8 million, and on May 23, 2014 awarded $302,002.50 in attorney’s fees. Anyone evaluating what a mixed file case is worth should work from the post-remittitur figure. Constitutional limits on punitive damages ratios apply to FCRA verdicts like any other, and a nine-to-one ratio is at the outer edge of what courts sustain.

Erickson v. First Advantage Background Services Corp., 981 F.3d 1246 (11th Cir. 2020). This one is a defense win, and it is more instructive than most plaintiff victories. Keith Erickson, born Keith Dodgson, was matched in an employment screening report to a Pennsylvania sex offender registry record belonging to his estranged biological father. The registry the screening agency drew from did not carry full dates of birth, so the match was made on name. The report disclosed on its face that it was a name-only match. The Eleventh Circuit affirmed judgment as a matter of law for the screening agency, analyzing the claim under section 1681e(b) — the accuracy standard — and concluding on that record that the procedures were not unreasonable given the disclosure. No damages were awarded. The practical lesson is that what an agency disclosed about the limits of its own match, and what the underlying source database actually contained, can decide the case; and that a consumer contesting a name-only match should preserve exactly what the report said on its face.

Sloane v. Equifax Information Services, 510 F.3d 495 (4th Cir. 2007). Frequently listed as a mixed file case, and it is not one — it arose from identity theft by a similarly named hospital employee, Shovana Sloan, rather than from an agency merging two files. It belongs on this list anyway, because of what the Fourth Circuit did with the damages. The jury awarded $351,000, comprising $106,000 in economic damages and $245,000 for emotional distress. The court held the emotional distress award excessive and remitted it to $150,000, identifying $256,000 as the maximum the record could sustain, and separately vacated a $181,083 fee award and remanded it. The case is the leading appellate authority on how much a court will let a jury award for the non-economic consequences of a corrupted credit file, which is the largest component of damages in most mixed file cases.

What government enforcement has actually said

The regulatory record on mixed files is thinner than the litigation record, and it is often described inaccurately. What follows is what the agencies actually did.

The CFPB’s December 2012 white paper, Key Dimensions and Processes in the U.S. Credit Reporting System, is the foundational government description of the problem. It defines the term, and it explains the matching architecture — including that tradelines are assigned to the best available match even when identifiers do not fully agree. For a consumer trying to explain to a court or a bureau why the wrong record is in the file, the regulator’s own account of the mechanism is more useful than any secondary source.

Supervisory Highlights, Issue 14 (March 2017) was a consumer reporting special edition, and it reported supervisory findings on mixed file testing and on the adoption of stricter identity-matching criteria at examined entities. Supervisory findings are not public enforcement actions and do not name the companies, but they establish that the regulator examined this specific failure mode and required changes.

The state attorney general settlements of 2015 are the most consequential regulatory intervention, and they are two separate agreements that are constantly conflated. On March 9, 2015 the New York Attorney General reached a settlement with the three nationwide agencies that produced the National Consumer Assistance Plan, the industry program that changed how medical debt, public records and dispute handling were treated. That agreement was New York’s alone. A separate multistate settlement followed on May 20, 2015, involving 31 states, a $6 million payment, and led by then-Ohio Attorney General Mike DeWine. The mixed file provisions are in the multistate agreement, not the New York one. That agreement requires an escalated process for handling complicated disputes, including those involving identity theft, fraud, or mixed files — described in the settlement as cases in which one consumer’s information is mixed with another’s — and requires each credit reporting agency to notify the other agencies if it finds a mixed file. That cross-notification obligation is the one most consumers have never heard of and the one most worth invoking, because a file that is mixed at one agency is very often mixed at the other two, for the same reason.

The CFPB’s annual report to Congress under section 611(e), published January 5, 2022, quantified the complaint volume. From January 2020 to September 2021 the Bureau received more than 800,000 credit or consumer reporting complaints, more than 700,000 of which concerned Equifax, Experian or TransUnion. The largest single sub-issue by volume was “information belongs to someone else” — the complaint category that captures mixed files. The Bureau does not publish a percentage for that sub-issue, so no one honestly can state one, but it is the top category out of a set that includes every other kind of credit reporting problem. The same report found the three agencies provided relief in under two percent of covered complaints in 2021, down from roughly a quarter in 2019. The Bureau’s most recent edition, published in December 2025, reported approximately 4.8 million of 5.6 million total complaints between January 2024 and June 2025 concerned consumer reporting, with roughly 3.9 million about the three nationwide agencies, characterized as an almost 3,000 percent increase since January 2020.

The FTC’s Section 319 accuracy study is the most-cited and most-misquoted data set in this field. Congress directed the FTC to study credit report accuracy, and the study surveyed 1,001 participants across 2,968 reports. Four distinct figures come out of it and they are not interchangeable. Twenty-six percent of participants identified at least one potentially material error. Twenty-one percent had a report modified after dispute — this is the “one in five” figure the press coverage used. Thirteen percent saw a credit score change. And 5.2 percent had an error corrected that moved them into a better credit risk tier, meaning it could have affected the terms of credit they were offered. The fifth interim report was issued in December 2012 and announced on February 11, 2013; the sixth interim and final reports followed in January 2015, announced January 21, 2015.

Two older FTC actions are commonly miscited in this context and should not be relied on as accuracy penalties. On January 13, 2000 the FTC obtained $2.5 million in total penalties — $1 million each from Experian and TransUnion, $500,000 from Equifax — for failing to staff their toll-free consumer telephone lines adequately, a violation of FCRA section 609(c)(1)(B). On July 30, 2003 Equifax paid $250,000 for violating that 2000 consent decree. Both concern telephone access, not file accuracy. Citing them as accuracy enforcement misstates the record, and a consumer who does so in a dispute letter loses credibility.

Why the ordinary dispute process struggles with mixed files

Most credit report disputes describe a fact about an account: the balance is wrong, the account was paid, the late payment did not happen. The reinvestigation system is built for that. It converts the dispute into a standardized code, sends it to the furnisher, and asks the furnisher to confirm or correct what it reported.

A mixed file dispute is a different kind of assertion. It says the account is not mine at all — it belongs to a different person who really does owe it. Routed through the same machinery, that dispute arrives at a furnisher who looks at its own records, confirms that the account exists and is delinquent, and verifies it. Nothing in that exchange tests the only question that matters, which is whether the agency was right to file the record to this consumer. The verification is accurate and the outcome is wrong.

This is also why mixed file entries reappear after they are deleted. If the underlying matching logic has not changed, the furnisher’s next monthly update contains the same record with the same identifiers, and the same rules route it to the same wrong file. A consumer who successfully removed an item and finds it back two cycles later has not been ignored; the file is being re-mixed. Documenting that recurrence — dated report copies before and after — is often stronger evidence than the original error, because it shows the problem is systemic rather than a one-time slip.

The practical consequence is that a mixed file dispute has to be framed as an identity question, addressed to the agency, and supported with identifiers rather than account facts. Say that the account belongs to a different person. Say what distinguishes you from that person: full legal name including any suffix, date of birth, and the fact that the Social Security numbers differ — without writing out the digits. Ask the agency to describe the procedure it used to associate the record with your file, and ask it to notify the other nationwide agencies, which the multistate settlement requires when a mixed file is found.

Mixed files are not confined to the three nationwide agencies

Every company that assembles records about people and sells them for credit, employment, tenancy or account-opening decisions runs the same kind of matching, and specialty agencies frequently run it on thinner data than the nationwide bureaus have. A consumer who checks only the big three can miss the file that is actually causing the problem.

• Experian — one of the three nationwide files, subject to the multistate cross-notification requirement.
• Equifax — operates under the same accuracy and reinvestigation standards.
• TransUnion — the agency at the center of both Cortez and Ramirez.
• Innovis — a further nationwide agency most consumers never check.
• ChexSystems — deposit account histories used to approve or deny checking accounts.
• Early Warning Services — bank-shared account and fraud records.
• SageStream — an alternative credit file with its own matching and its own dispute process.
• The Work Number — employment and income records matched to individuals by employers.
• CoreLogic — rental and public record data sold to landlords.
• RealPage — eviction and screening records used by property managers.
• First Advantage — the employment screening agency in Erickson.
• Sterling — criminal and employment record searches for employers.

Request your file from each agency that plausibly holds a record about you, not only from the bureaus. The employment screening and tenant screening files are where a mixed record does the most immediate damage, because a job offer or a lease is usually decided in days and the adverse decision is made before anyone tells you a report existed.

What to do if you think your file is mixed

Pull all three nationwide reports on the same day, plus files from any specialty agency relevant to the decision that went against you. Compare them field by field rather than reading each one for anything alarming: the personal information section is where the evidence usually is. Alternate names you have never used, a middle name that is not yours, a suffix you do not carry, addresses you never lived at, and employers you never worked for are the signature of a file that has absorbed another person’s records. Those entries are more probative than the disputed account itself, because they show how the record got in.

Write to the agency rather than using the online dispute portal for this kind of claim, and keep proof of delivery. Describe the problem as an identity assignment error, list the specific items that are not yours, and enclose copies of documents that establish your own identifiers. Ask for the description of the reinvestigation the statute entitles you to request, including who the agency contacted and what it received back. Save the dispute, the response, and a dated copy of the report before and after. If the entry returns, save that too, along with the date.

If the mixed file has already caused a denial — a mortgage, a car loan, an apartment, a job — keep the adverse action notice, because it names the agency that supplied the report and starts the clock on your right to a free copy of it. That notice is often the only document connecting the decision to the file.

When a mixed file becomes a legal claim

The Fair Credit Reporting Act requires agencies to follow reasonable procedures to assure maximum possible accuracy, and to conduct a reasonable reinvestigation when a consumer disputes an item. A mixed file that survives a properly framed written dispute is evidence that at least one of those obligations was not met, and the case law from Thompson forward has consistently treated the matching procedure itself as the thing to be examined. Damages in these cases are typically dominated by the non-economic harm — the denial, the delay, the time spent, the effect of being told repeatedly that a stranger’s debt is yours — and Sloane remains the leading discussion of what courts will sustain on that element. Where the agency knew of the error and left it uncorrected, as in Miller, punitive damages are available, subject to the ratio limits that reduced the award in that case.

Our mixed credit file attorneys handle these claims nationwide, and the broader FCRA practice covers the related failures that often accompany a mixed file, including credit reporting errors generally, identity theft where a real fraud sits underneath the merge, and deceased indicators reported on living consumers, which is a mixed file problem with a particular and unusually severe consequence.