In the
United States Court of Appeals For the Seventh Circuit ____________________ No. 23-2355 TAMARA S. FRAZIER, Plaintiff-Appellant, v.
EQUIFAX INFORMATION SERVICES, LLC, Defendant-Appellee.
____________________
Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:20-cv-06725 — Harry D. Leinenweber, Judge. ____________________
ARGUED JANUARY 25, 2024 — DECIDED AUGUST 7, 2024 ____________________
Before HAMILTON, BRENNAN, and KIRSCH, Circuit Judges. BRENNAN, Circuit Judge. In 2020, Tamara Frazier applied for a mortgage with Mutual Federal Bank. In considering her application, the bank reviewed a “tri-merge” report from CreditLink, which aggregated data received from Equifax, Experian, and TransUnion credit reporting agencies. After re- view, the bank denied her application. 2 No. 23-2355
Frazier brought Fair Credit Reporting Act (“FCRA”) claims against Equifax, alleging it reported inaccurate late payments in violation of 15 U.S.C. § 1681e(b) and § 1681i(a)(1)(A). Specifically, Frazier claims Equifax’s con- sumer report and file of her credit history contained inaccura- cies. The district court granted summary judgment to Equifax, ruling that the information “furnished and reported by Equifax … was all true” and, as a result, there “was no inac- curacy in Equifax’ report.” We affirm. I. In 2007, Frazier obtained a home mortgage. She made monthly payments through September 2015, but she stopped in October 2015. By January 2016, she was 90 days delinquent. To resolve the delinquency, Frazier negotiated and settled her debt through a short sale of her home, which closed on Janu- ary 14, 2016. Frazier knew the sale would “be reported to the credit bureau(s) [as] ‘settled in full for less than total payoff.’” Dovenmuehle Mortgage, Inc. (“DMI”) had acted as Fra- zier’s subservicer. A mortgage subservicer helps lenders ad- minister mortgage loans by accepting and keeping track of payments. It also furnishes payment data to credit reporting agencies including Equifax, Experian, and TransUnion. Those agencies compile and process that consumer credit infor- mation and produce a credit report for end-users, such as banks and landlords. Sometime between 2019 and 2020, Frazier realized that her closed mortgage account was reported as delinquent on her credit reports—namely, that she was at least 90 or more days late on her mortgage payments, even though her mortgage debt was extinguished through the short sale. Frazier No. 23-2355 3
disputed this information to several credit reporting agencies, including Equifax. Frazier sent Equifax several letters chal- lenging this and other entries on her credit report. When a consumer notifies a credit reporting agency that information on a credit report is incorrect, the agency sends the relevant data furnisher an Automated Consumer Dispute Verification (“ACDV”) form. The ACDV form contains the ac- count payment data the credit reporting agency possesses and the relevant data items the consumer disputes. Once no- tified of a dispute, the data furnisher has a statutory duty to investigate and correct or verify the disputed data. This is done by returning the ACDV form to the credit reporting agency with any amended or verified data inserted next to the old data. See generally 15 U.S.C. § 1681s-2(a)(2)(b). To confirm the accuracy of its records on Frazier’s mort- gage, Equifax sent each of Frazier’s dispute letters to DMI and asked DMI to confirm or update the information in Frazier’s credit file. Each time, DMI confirmed the reporting. In one instance, DMI updated its records to include dashes in the ac- count history for all months after December 2015. As we ex- plained in Frazier v. Dovenmuehle Mortg., Inc., 72 F.4th 769, 777 (7th Cir. 2023), the dashes meant no reporting for all months following the short sale. In turn, Equifax updated its information consistent with DMI’s reporting and sent Frazier a letter reflecting any changes or confirmation of the information in her credit file. Like DMI, Equifax recorded that the current balance, amount past due, and actual payment were $0; the “Date of Last Pay- ment” was September 2015; the account had been closed in January 2016 and “Paid for Less Than Full Balance”; and the account status was listed as “90-119 Days Past Due.” Equifax, 4 No. 23-2355
though, in its consumer file reported a number of dates in- stead of dashes. In 2020, Frazier applied for a mortgage with Mutual Fed- eral Bank. As part of that process, the bank procured a “tri- merge” report from CreditLink that aggregated data received from Equifax, Experian, and TransUnion. The CreditLink re- port detailed that Frazier’s loan had been “paid for less than full balance” with a “date of last activity” in October 2015. But unlike Equifax’s disclosures, the CreditLink report did not contain the “closed date” on the account or indicate that the short sale had occurred. The bank denied Frazier’s loan application because, as its loan officer later testified, Frazier’s student loan obligations made her debt-to-income ratio unacceptably high. The ad- verse action letter the bank sent to Frazier also indicated the loan was denied due to “Excessive obligations” and “Insuffi- cient income for total obligations.” Frazier filed separate lawsuits against DMI, Equifax, and CreditLink. In one, we affirmed judgment for DMI, holding that the information DMI furnished to Equifax was “not ma- terially misleading as a matter of law.” Dovenmuehle, 72 F.4th at 777. In another, the parties settled, and the suit was dis- missed with prejudice. See Frazier v. CreditLink LLC, No. 1:22- cv-05226 (N.D. Ill. filed Sept. 26, 2022) (settled on October 10, 2023 and dismissed with prejudice on December 4, 2023). In the third, this case, Frazier brings FCRA claims against Equifax for allegedly reporting inaccurate late payments in violation of 15 U.S.C. § 1681e(b) and § 1681i(a)(1)(A). The dis- trict court granted summary judgment to Equifax, ruling that the information “furnished and reported by Equifax … was No. 23-2355 5
all true” and, as a result, there “was no inaccuracy in Equifax’ report.” Frazier appeals. II. We “review the district court’s summary-judgment order de novo and construe the record in the light most favorable to [Frazier].” Persinger v. Sw. Credit Sys., LP, 20 F.4th 1184, 1194 (7th Cir. 2021). Frazier “must do more than simply show that there is some metaphysical doubt as to the material facts.” Sarver v. Experian Info. Sols., 390 F.3d 969, 970 (7th Cir. 2004) (quotation marks omitted). “Where the record taken as a whole could not lead a rational trier of fact to find for [her], there is no ‘genuine issue for trial,’” and summary judgment is proper. Id. Congress enacted the FCRA “to ensure fair and accurate reporting, promote efficiency in the banking system, and protect consumer privacy.” Safeco Ins. Co. v. Burr, 551 U.S. 47, 52 (2007). “To safeguard these interests, the FCRA provides a private right of action for injured consumers.” Persinger, 20 F.4th at 1194. To prevail on her FCRA claims, Frazier must prove that Equifax prepared an inaccurate consumer report, see 15 U.S.C. § 1681e(b), or kept inaccurate information in a consumer’s file, see § 1681i(a)(1)(A). Equifax’s “liability under both § 1681e(b) and § 1681i(a) depends on inaccurate information—if the credit report is accurate, [Frazier] has suffered no damages.” Chaitoff v. Experian Info. Sols., Inc., 79 F.4th 800, 811 (7th Cir. 2023); see Denan v. Trans Union LLC, 959 F.3d 290, 294 (7th Cir. 2020) (“[The FCRA] requires a plaintiff to show that a consumer reporting agency prepared a report containing inaccurate information.” (quotation marks omitted)). An item on a credit report can be incomplete or inaccurate within the meaning of the FCRA because it is 6 No. 23-2355
patently incorrect, or because it is “misleading in such a way and to such an extent that it can be expected to adversely affect credit decisions.” Chaitoff, 79 F.4th at 812 (citing Dovenmuehle, 72 F.4th at 776). A. Accuracy of Equifax’s Report Frazier argues that Equifax’s consumer report of her credit history was inaccurate and that Equifax kept inaccurate infor- mation in her consumer file. The only evidence of what Equifax would have reported to prospective lenders comes from the information Equifax kept in Frazier’s file. DMI fur- nished and reported that information to Equifax, and Equifax reflected it in the consumer disclosure provided to Frazier. The parties agree that much of the information in the con- sumer disclosure is true. For example, Equifax accurately re- ported that the “Date of Last Payment” on Frazier’s mortgage with DMI was in “09/2015” and the “Date of 1st Delinquency” on the mortgage was in “10/2015.” Equifax also accurately re- ported that Frazier’s mortgage was “Closed” in “01/2016,” the “Account [Was] Paid For Less Than Full Balance,” and the “Balance Amount” owed on the reported date was “$0.” But Frazier contends other facts in the consumer disclo- sure must be interpreted to show that her mortgage loan ac- count was inaccurately reported as delinquent at the time the bank reviewed her file. Those facts include that Equifax re- ported Frazier’s: “Status” as “90-119 Days Past Due”; and “Account History with Status Codes” as “2” or “3,” meaning 90 plus days late, on eight occasions—“12/2015,” “12/2018,” “01/2019,” No. 23-2355 7
“06/2019,” “08/2019,” “10/2019,” “11/2019,” and “07/2020”. As noted, Frazier filed a separate lawsuit against DMI rais- ing the same argument she raises here. See Dovenmuehle, 72 F.4th 769. That decision governs the accuracy of Equifax’s re- porting and forecloses Frazier’s argument. In Dovenmuehle, we held that the very information furnished to Equifax in this case would not “materially mislead a reasonable observer to conclude that Frazier is currently delinquent.” 72 F.4th at 777. Reporting information must be “reviewed in context.” Id. And in “full context” any information indicating the account was “90 days delinquent” was not misleading. Id. “A debtor can- not be currently delinquent on a loan that no longer exists.” Id. The information that Frazier’s account was “90 days delin- quent” was “directly beside” other information indicating that her “account is closed” and her “loan was paid in full for less than the remaining balance.” Id. Additionally, it is “not clear” whether the status reflected as “90-119 Days Past Due” is “an incorrect indicator of current delinquency or a correct one of historical delinquency.” Id. 1 So, the status “is not pa- tently incorrect.” Id.
1 As the district court stated:
Of course, inaccurate information can harm the credit rat- ing of borrowers and make obtaining credit more difficult and costly. The other side of the coin, however, is that in- accurate information as to credit worthiness of the pro- posed borrower is important to the lender because inac- curate information can lead to default and non-payment. When evaluating the credit worthiness of Plaintiff, it would be important to know that she had a delinquent 8 No. 23-2355
As the district court explained, “a report of a current balance due of zero, together with a report that the account was delinquent when closed, do[es] not violate FCRA.” Frazier, 2023 WL 4134907, at *2; see Bibbs v. Trans Union, LLC, 43 F.4th 331, 343–44 (3d Cir. 2022) (credit report notation not materially misleading as to current—rather than historical— delinquency where “multiple conspicuous statements reflect[ed] that the accounts are closed and Appellants have no financial obligations to their previous creditors”); Gross v. Priv. Nat’l Mortg. Acceptance, 512 F. Supp. 3d 423, 426–27 (E.D.N.Y. 2021) (“If a creditor read the ‘Pay Status’ entry in isolation, the creditor might conclude that the account was currently [or historically] past due. But when the creditor read the rest of the entries, the creditor would surely forego that conclusion.”); Frazier v. DMI, 2022 WL 3445801 (N.D. Ill. Aug. 17, 2022). Equifax’s reporting of Frazier’s “Account History with Sta- tus Codes” as being 90 plus days late on “12/2015,” “12/2018,” “01/2019,” “06/2019,” “08/2019,” “10/2019,” “11/2019,” and “07/2020” does not change this analysis. As with the indication that Frazier’s account was “90 days delinquent,” the “Account History with Status Codes” was di- rectly below information that her “loan was paid in full for less than the remaining balance.” Dovenmuehle, 72 F.4th at 777. This included codes “2” or “3”, and the dates. Further, the “Account History with Status Codes” was “a few” rows below
payment history which caused a previous lender to lose money on a mortgage loan. Frazier v. Equifax Info. Servs. LLC, 2023 WL 4134907, at *3 (N.D. Ill. June 22, 2023). No. 23-2355 9
information indicating that her “account is closed.” 2 Id. So again, in “full context,” id., Frazier’s status as “90-119 Days Past Due” did not mean that she was currently delinquent on her debt when the bank reviewed the report. Therefore, the status is not materially misleading as a matter of law, and the district court correctly concluded that the information in the Equifax disclosure “was all true.” Our dissenting colleague contrasts Equifax’s consumer file here and the parallel document in Dovenmuehle. Equifax used dates, while DMI used dashes in the late payment section. The dissent reads Dovenmuehle as not answering the question here of what to make of the dates. We agree that the only difference between the documents in Dovenmuehle and in this case is the use of dashes versus dates. But this does not weaken our horizontal precedent in Dovenmuehle that whether information is “materially misleading” is necessarily answered by the context of the document in which it appears. Dovenmuehle applies to this case. The same legal reasoning which explained that “Status” and “Account History” are not “materially misleading”— in Dovenmuehle and in this case— applies to Equifax’s use of dates instead of dashes, neither of which are “materially misleading.” 72 F.4th at 777. B. Frazier’s Reliance on CreditLink’s Report As explained above, the only evidence of what Equifax would have reported to prospective lenders comes from the information Equifax kept in Frazier’s file. Frazier argues Equifax’s reporting was inaccurate based on a report pre- pared and sent to Mutual Federal Bank by a different com- pany, CreditLink. But Equifax cannot be held liable for a
2 Short Appendix 6. 10 No. 23-2355
report it neither prepared nor sent. And Equifax’s report did not cause the denial of Frazier’s loan application. Equifax cannot be held liable for CreditLink’s report. The report prepared and sent by CreditLink cannot establish Equifax’s liability because there is no evidence that Equifax dissemi- nated the alleged inaccuracies in that report.3 The CreditLink tri-merge report combines data collected from all three major consumer reporting agencies. Mortgage lenders commonly request these combined reports because they provide a more comprehensive analysis of a borrower’s ability to make timely monthly payments. Importantly, tri-merge reports combine only some of the reported information and do not necessarily replicate the full context of what each individual consumer reporting agency relays. Because tri-merge reports are not prepared by a con- sumer reporting agency, like Equifax, and do not always re- port all the information reported by a particular agency, courts have ruled that tri-merge reports, standing alone, can- not establish a consumer reporting agency’s liability. See, e.g., Zotta v. NationsCredit Fin. Servs. Corp., 297 F. Supp. 2d 1196, 1205–06 (E.D. Mo. 2003) (granting summary judgment to Experian given “no evidence that an Experian report—as op- posed to a tri-merge report, which Experian did not issue— was seen by a third party”). 4 We agree with this reasoning.
3 As noted supra n.1, Frazier separately sued and subsequently settled
with CreditLink over its report. See Frazier v. CreditLink LLC, No. 1:22-cv- 05226 (N.D. Ill. filed Sept. 26, 2022). 4 See, e.g., Weeks v. Equifax Info. Servs., LLC, No. 8:21-cv-2384, 2022 WL
685665, at *6 (M.D. Fla. Mar. 8, 2022) (collecting cases and dismissing No. 23-2355 11
Frazier agrees that CreditLink—not Equifax—sent the re- port that she primarily challenges. And she concedes that the CreditLink report is “completely missing” certain pertinent information that Equifax indisputably included in its own files: namely, that the short sale occurred and the account closed in January 2016. Both facts support our conclusion that there were no inaccuracies in Equifax’s report. Frazier argues “CreditLink declared that it did not alter any information re- ceived from Equifax, thus, Equifax clearly never provided the information to CreditLink.” But her understanding of Credit- Link’s declaration rests on unfounded assumptions. Yes, CreditLink declared it did not alter any information Equifax sent it, but that does not mean CreditLink included all the in- formation Equifax sent it. Contrary to the dissent’s claim, we are not speculating, but relying on record evidence, specifi- cally Equifax’s consumer file report. Given that the CreditLink report was not issued directly by Equifax and is missing certain relevant information that was reflected in Equifax’s credit files, that report cannot estab- lish that Equifax inaccurately reported Frazier’s information. The dissent asserts that, contrary to CreditLink’s declara- tion, we speculate that CreditLink deleted information from Equifax’s consumer file report. This is not the case. Rather, the CreditLink report had a field to list Frazier’s date of last activ- ity, but nowhere to note Frazier’s date closed. Thus, it is not a
FCRA claims based on tri-merge report because it “was not issued directly by Trans Union … and is missing certain pertinent information”); Troy v. Equifax Info. Servs., LLC, No. CV-20-01447, 2021 WL 5998518, at *7 (D. Ariz. Dec. 20, 2021) (“[T]his Court cannot find—based on a [tri-merge] report alone—that a Trans Union credit report containing the alleged inaccuracy was actually transferred to a third-party creditor”). 12 No. 23-2355
question of deleted information, but a recognition that there was no field for CreditLink to include the information. The evidence from Equifax included the date closed; the evidence from CreditLink did not. Equifax’s report did not cause the bank to deny Frazier’s loan application. Frazier’s alleged financial and emotional injuries stem from Mutual Federal Bank’s denial of her 2020 mortgage application. But there is no evidence that the bank reviewed an Equifax report. Instead, the bank reviewed only Credit- Link’s report. And as we have explained, Equifax cannot be held liable for any possible inaccuracies in that report. But even if Equifax could be held liable, the CreditLink re- port did not mislead Mutual Federal Bank, and the bank did not deny Frazier’s loan application due to a late-payment his- tory in that report. Frazier’s loan officer at the bank testified that he understood the CreditLink report meant that “Frazier settled her [DMI] account for less than the full balance in or around October of 2015”—i.e., that it did not inaccurately re- port late payments. More importantly, the bank’s adverse ac- tion letter stated it denied Frazier’s loan application due to her high debt-to-income ratio—i.e., due to “Excessive obliga- tions” and “Insufficient income for total obligations,” not due to “Unacceptable payment record on previous mortgage.” Frazier tries to rebut this evidence by pointing to a Sep- tember 10, 2020 email from her loan officer which states, “[T]he wrong information … showing late payments in 2018 … is hurting the approval process.” But the loan officer testi- fied that, when he sent the September 10 email, he had “not review[ed]” the CreditLink report itself and had “not calcu- lated yet” the student loans that ultimately doomed her ap- plication. In addition, both that same loan officer and the No. 23-2355 13
underwriter who denied the loan two weeks later confirmed that the ultimate reason Frazier was denied a loan was that her student loan obligations made her debt-to-income ratio unacceptably high. So, the bank denied Frazier’s loan application for reasons unrelated to any late-payment history reported by Equifax. Equifax had no connection to Frazier’s high debt-to-income ratio. Frazier’s claimed injury—the loan application denial— therefore was not causally connected to Equifax’s report. Frazier argues that if not for Equifax’s report, her debt-to- income ratio would have satisfied the bank’s loan require- ments. She claims the late payments triggered a manual un- derwriting of her application, in which a lower debt ratio was required. But the record shows that Frazier’s debt ratio was 57.986 percent, above the 56.999 percent necessary even with- out the downgrade to manual underwriting. Frazier’s claim that she could have lowered the ratio even more and met the threshold is merely hypothetical. The only record evidence about Frazier’s ability to satisfy a lowered debt ratio comes from unsupported expert testi- mony. Frazier’s expert witness offered at deposition that Fra- zier “indicated to me that [reaching the 56.999 percent neces- sary] would have been no problem for her at the time … .” Equifax had filed a motion to exclude the expert’s testimony and report, alleging it violated Daubert v. Merrell Dow Pharma- ceuticals, Inc., 509 U.S. 579 (1993). The district court did not rule on that motion because it found there was no FCRA vio- lation. We are not judging the persuasiveness of the expert’s con- clusions. But that does not mean the expert’s opinion must be 14 No. 23-2355
accepted blindly. Gopalratnam v. Hewlett-Packard Co., 877 F.3d 771, 782 (7th Cir. 2017) (“[I]f the district court failed to conduct a Daubert analysis, then we review de novo whether the expert’s testimony was admissible under Federal Rule of Ev- idence 702.”). In the end, no admissible evidence was pre- sented to support those conclusions. At summary judgment, more is needed. Flowers v. Kia Motors Finance, 105 F.4th 939, 946 (7th Cir. 2024). III. Equifax’s consumer report and file of Frazier’s credit his- tory did not contain inaccuracies. Equifax cannot be held lia- ble for CreditLink’s report, and Equifax’s report did not cause the denial of Frazier’s loan application. For these reasons, we AFFIRM the district court. No. 23-2355 15
HAMILTON, Circuit Judge, dissenting. We should reverse summary judgment for defendant Equifax. Substantial evi- dence shows that when Equifax was told of false information in plaintiff Frazier’s credit report, Equifax failed to correct it. Its failure undermined plaintiff Frazier’s application for a mortgage in 2020. Based on the evidence that Equifax was re- sponsible for the bad information and factual disputes about causation, this should be a straightforward reversal of sum- mary judgment. The majority opinion instead affirms. It does so only by overlooking and misunderstanding critical evi- dence and by speculating in favor of the defendant. I respect- fully dissent. Consumer credit reporting in the United States is a high- volume system, processing well over one billion consumer credit accounts each month. 1 Mistakes happen, of course, and credit reporting agencies are not strictly liable for them. In- stead, Congress protected consumer-borrowers under the Fair Credit Reporting Act (FCRA or the Act), 15 U.S.C. § 1681 et seq. The Act grants consumers a right to obtain corrections and imposes on credit reporting agencies duties to correct mistakes promptly. Under the Act, Equifax had a duty here to “follow reasonable procedures to assure maximum possible ac- curacy of the information concerning the individual about whom the report relates.” 15 U.S.C. § 1681e(b) (emphasis added).
1 See Consumer Financial Protection Bureau, Key Dimensions and Pro-
cesses in the U.S. Credit Reporting System (December 2012). https://www.consumerfinance.gov/data-research/research-reports/key- dimensions-and-processes-in-the-u-s-credit-reporting-system/ [https://perma.cc/R3WD-RT3Z]. 16 No. 23-2355
As the majority points out, the process for challenges and corrections is highly automated. Ante at 3. This case shows the challenges consumers can face in navigating the credit re- porting dispute system. Plaintiff learned that her credit report included patently false information about her prior mortgage with Dovenmuehle Mortgage, Inc. (DMI). She used her rights under the Act to seek corrections from Equifax. The key mis- takes were not corrected. Instead, the false information was forwarded to Mutual Federal Bank, which denied Frazier’s application for a new mortgage in 2020. The principal ques- tion here is where in the chain of credit information one or more parties breached a legal duty under the Act to correct the errors. 2 I. Frazier’s Simple Case Against Equifax Frazier’s case uses the process of elimination. She was turned down for a mortgage in 2020 based on a CreditLink “tri-merge” credit report sent to Mutual Federal. The report showed incorrectly that Frazier had made payments on the DMI mortgage 90 days late in seven different months in the four years after she had settled that debt through the short sale in January 2016. Frazier disputed the report, and we know that in response, DMI sent accurate information to Equifax remov- ing those phantom seven delinquent months. Frazier v. Dovenmuehle Mortgage, Inc., 72 F.4th 769, 777–78 (7th Cir. 2023) (affirming summary judgment for DMI because undisputed facts showed it sent correct information to Equifax). We do
2 For reasons not clear to me, Frazier chose to bring three separate
lawsuits against DMI, Equifax, and CreditLink. That strategy has allowed the lone defendant in each case to try to blame the others for Frazier’s problems. Regardless of the curious strategy, the record requires reversal of summary judgment for Equifax. No. 23-2355 17
not have a copy of what Equifax sent to CreditLink, but Fra- zier has offered testimony from CreditLink both that the mis- leading information in its report came from Equifax and that CreditLink did not alter it. There is also evidence that Equifax misunderstood the cor- rected information it received from DMI and therefore failed to correct the false information in its file on Frazier. (A witness for Equifax testified as much.) It is also reasonable to infer that Equifax then sent the false information on to CreditLink. The false information then went to Mutual Federal, causing the denial of Frazier’s mortgage application. On this evidence, a reasonable jury could find against Equifax under 15 U.S.C. § 1681e(b) and § 1681i(a)(1)(A). There is no doubt that something went wrong. Some- where in the three-link chain of supposedly corrected credit information that went from DMI to Equifax, then to Credit- Link, and finally to Mutual Federal, false information was not corrected. Start at the beginning of the chain, with DMI. In late 2015, Frazier had fallen behind on an earlier mortgage loan serviced by DMI. By January 2016, she was 90 days behind on payments. She and DMI agreed to a “short sale” of her home that closed in January 2016. That short sale was reported cor- rectly to the credit reporting agencies as a debt that had been settled in full for less than the total payoff. But in 2019 or early 2020, Frazier learned that credit agen- cies were listing her in seven recent months as still having been 90 days delinquent on her mortgage payments to DMI even though the debt had been fully settled years earlier. She invoked her right to ask for a correction. That should have been fairly easy, but the system did not work for her. She be- gan by telling Equifax, in writing and accurately, that it was 18 No. 23-2355
reporting incorrect information for her, and she requested corrections. According to routine practices, Equifax forwarded those requests to DMI, which responded promptly to Equifax with corrected information. Most important here, DMI updated Frazier’s account information to show dashes (“-”) for all months after January 2016. As far as DMI was concerned, the dashes meant correctly that there was “no reporting” of data for any time after January 2016. Dovenmuehle Mortgage, 72 F.4th at 777. That meant Frazier was of course not delinquent for any monthly payments after January 2016. We held that DMI’s corrected information was not misleading in Frazier’s case against DMI. Id. at 777–78 (affirming summary judgment for DMI). When Equifax received the information from DMI, it should have been able to correct its false reports that Frazier had continued to be delinquent on the DMI mortgage for years after the debt had been resolved. Next, let’s skip ahead for a moment and work backward from the third and last link in the information chain, when Frazier applied for a mortgage with Mutual Federal in 2020. Mutual Federal received from CreditLink a so-called “tri- merge” report on Frazier. It combined information from the three major credit reporting agencies (Equifax, TransUnion, and Experian). That report to Mutual Federal showed that Frazier was still delinquent in seven monthly payments on the DMI mortgage in the years after the debt had been settled. Most important, it showed three delinquent monthly pay- ments in the twelve months before Frazier applied for a new mortgage in 2020. The CreditLink report to Mutual Federal did not list the date when the DMI account was settled; it No. 23-2355 19
showed that the last activity occurred in October 2015. See Bodnar Decl. ¶ 4, Dkt. No. 113–1. 3 So: we have evidence that in the first information link, DMI-to-Equifax, DMI provided accurate information to Equifax. And we have evidence that at the third information link, CreditLink-to-Mutual Federal, CreditLink forwarded misleading information to Mutual Federal. By process of elimination, we need to focus on the middle link, Equifax-to- CreditLink. Equifax received accurate information directly from DMI, and it forwarded information to CreditLink. The question is whether Equifax sent bad information to Credit- Link or whether CreditLink made a mistake with accurate in- formation. Equifax cannot tell us. It did not keep a record of the infor- mation it sent to CreditLink. Oral Arg. 14:30–15:00. And given this gap in the evidence, it is difficult to understand how the district court granted summary judgment on the theory that the information “furnished and reported by Equifax … was all true” and that there was “no inaccuracy in Equifax’[s] re- port.” 2023 WL 4134907, at *3 (N.D. Ill. Jun. 22, 2023). The dis- trict court simply did not have such information, let alone ev- idence making this point undisputed. The majority tries to fill in the gap with evidence about what it says Equifax “would have reported.” Ante at 9. That evidence is also a step away from what Equifax actually reported to CreditLink. Moreover, that file shows the false information about the seven
3 Frazier has offered evidence that the incorrect information about de-
linquent payments hurt her ability to obtain the mortgage she sought in 2020. Causation is discussed below in Part III. 20 No. 23-2355
delinquent monthly payments. If that’s what Equifax sent, Equifax should not be off the hook. Frazier, on the other hand, has offered testimony from the president of CreditLink that it did not alter any of the infor- mation provided to it by Equifax. The same witness testified that the critical and false “Late Dates” on the report it pro- duced specifically came from Equifax, not from the other two credit reporting agencies. Bodnar Decl. ¶¶ 4–5, Dkt. No. 113– 1. By process of elimination, DMI and Mutual Federal are off the hook. The bad information must be attributable to Equifax or CreditLink, or perhaps both. Plaintiff has offered evidence that Equifax is to blame. Equifax cannot even dispute those facts because it did not keep a copy of the report it sent to CreditLink. And the evidence that Equifax offers of what it “would have” sent CreditLink still has the false reports of seven delinquent monthly payments. That’s why this should be an easy reversal of summary judgment. II. The Majority Opinion’s Errors To avoid reversal, Equifax and the majority opinion offer several mistaken theories: (a) we decided in Dovenmuehle Mortgage that Equifax was not to blame; or (b) maybe Credit- Link deleted some key information; or (c) the false infor- mation was so obviously wrong that it could not have misled anyone. None of these theories offers a sound basis for affirm- ing on summary judgment. A. We Decided This in Dovenmuehle? No, we did not. In Dovenmuehle Mortgage, we affirmed summary judgment for DMI. Frazier complained about the false information in her credit report. Equifax forwarded her No. 23-2355 21
complaint to DMI, which responded with corrected infor- mation showing no late payments after January 2016. That un- disputed fact led us to affirm summary judgment for DMI. 72 F.4th at 777–78. The majority writes here: “That decision gov- erns the accuracy of Equifax’s reporting and forecloses Fra- zier’s argument. In Dovenmuehle, we held that the very infor- mation furnished to Equifax in this case would not ‘materially mislead a reasonable observer to conclude that Frazier is cur- rently delinquent.’” Ante at 7, quoting 72 F.4th at 777. That second quoted sentence correctly states our holding on the information DMI sent to Equifax. The problem is that evidence shows that the “very information” DMI sent on did mislead Equifax itself. Equifax misunderstood the corrected in- formation from DMI. An Equifax employee testified that the Equifax computer system misunderstood the “dashes” from DMI as “no change,” rather than what DMI meant: “no infor- mation,” meaning no delinquent payments. See Willis Dep., Dkt. No. 86–5 at 27, 35. The majority writes that “the only dif- ference between the documents in Dovenmuehle and in this case is the use of dashes versus dates.” Ante at 9. Yes, exactly. That’s the critical difference, between zero delinquent pay- ments (the dashes in DMI’s corrected report) and seven delin- quent monthly payments (the dates in the Equifax file) in the years after the account was closed. So, no, we did not decide in the Dovenmuehle Mortgage de- cision that Equifax is off the hook. Quite the contrary. We held that DMI forwarded correct information to Equifax and that DMI was not responsible for Equifax’s further report of “the amended data to indicate she was currently delinquent on the mortgage with missed payments in months following the set- tlement in January 2016.” 72 F.4th at 774. 22 No. 23-2355
B. Maybe CreditLink Deleted Some Key Information? The majority’s next theory for affirming summary judg- ment is founded on speculation. Perhaps CreditLink deleted some key information that it received from Equifax and thus did not forward that information to Mutual Federal? See ante at 9–11. The supposedly key information is the date of Fra- zier’s settlement of her DMI account in January 2016. We know that DMI sent that information to Equifax. We also know it was not included in the CreditLink report to Mutual Federal. This theory for affirmance fails to apply the summary judgment standard. It speculates in favor of the party who moved for summary judgment rather than give the non- moving party the benefit of her evidence and reasonable inferences from it. What does the record show about what Equifax sent to CreditLink? We know that Equifax has no direct evidence to answer that question. Equifax did not keep a record of its re- port to CreditLink. The majority writes: “The only evidence of what Equifax would have reported to prospective lenders comes from the information Equifax kept in Frazier’s file. DMI furnished and reported that information to Equifax, and Equifax reflected it in the consumer disclosure provided to Frazier.” Ante at 6. Two key assertions in these sentences are factually wrong. First, as just discussed, the corrected information DMI re- ported to Equifax was not what Equifax included in the con- sumer disclosure provided to Frazier in our record. Second, we have contrary evidence of what Equifax reported to pro- spective lenders from CreditLink. The president of No. 23-2355 23
CreditLink testified in an affidavit that it prepared its report to Mutual Federal “[w]ithout altering any of the information provided by the nationwide consumer reporting agencies.” Bodnar Decl. ¶ 4, Dkt. No. 113–1. She testified further: “The ‘Late Dates’ reflected in the screenshot [the seven delinquent monthly payments] were reported to Credit Link by Equifax Information Services LLC.” Id., ¶ 5. The majority responds to this evidence: “Yes, CreditLink declared it did not alter any information Equifax sent it, but that does not mean CreditLink included all the information Equifax sent it.” Ante at 11. With respect, this is just speculation. We have testimony that CreditLink did not “alter” any information from Equifax. As a matter of the English language, the broad verb “alter” can include the more specific verb “delete.” One way to “al- ter” information is to “delete” some of it. Moreover, there is simply no evidence that a deletion occurred. Perhaps future cross-examination of Bodnar might offer support for the ma- jority’s speculation, but that has not happened. Even if one were inclined to think “deletion” was a reasonable possibility, despite the uncontradicted Bodnar testimony, the majority’s reasoning still departs from elementary rules of summary judgment practice. We draw inferences in favor of the non- moving party, not the moving party. Navratil v. City of Racine, 101 F.4th 511, 518 (7th Cir. 2024); see also, e.g., Anderson v. Lib- erty Lobby, Inc., 477 U.S. 242, 255 (1986) (“The evidence of the non-movant is to be believed, and all justifiable inferences are to be drawn in his favor.”). 24 No. 23-2355
C. So Wrong It Was Not Really Misleading? But let’s assume for the sake of argument that Equifax passed along to CreditLink the January 2016 date of Frazier’s settlement with DMI, and that CreditLink deleted that date when it forwarded its report to Mutual Federal. That still would not save the Equifax report from having been patently false and misleading. That’s because we must assume, as the majority does, that the Equifax report continued to list the seven post-settlement monthly payments as delinquent, in- cluding three within the twelve months before Frazier applied for a new mortgage. The majority seems to accept Equifax’s theory that the combination of the settlement date and reported delinquent monthly payments was so obviously wrong that it would not have misled anyone. Pause there for a moment. The defense theory is that a credit reporting agency is off the hook under the FCRA if its credit report is so wrong it’s obvious there’s a mis- take. Equifax actually made that argument during oral argu- ment. Oral Arg. 21:00–23:00. That startling theory is wrong on the law and the facts. A credit report showing delinquent monthly payments for years after the date the report shows the account was closed is false and misleading. That should be self-evident. I’m will- ing to assume that a reasonable reader who saw both the re- ported settlement date and the reported delinquent payments after that date would understand that something was wrong. But what? Maybe the settlement date was wrong. Maybe the reported delinquent payments were wrong. Maybe both were wrong. The reasonable reader knows she simply cannot rely on that report as a reliable indicator of credit risk. Maybe that No. 23-2355 25
reasonable reader asks for clarification, or maybe she just de- nies the mortgage application and moves on to the next file on her desk. Equifax’s theory that false information can be so obviously false that it cannot violate the FCRA’s duty to provide accu- rate information is downright Orwellian. It also conflicts with our more general FCRA case law. “Courts have long under- stood that, when it comes to the FCRA, ‘accurate’ means more than just ‘technically correct.’” Chaitoff v. Experian Info. Sols., Inc., 79 F.4th 800, 812 (7th Cir. 2023), citing Koropoulos v. Credit Bureau, Inc., 734 F.2d 37, 40 (D.C. Cir. 1984) (“Certainly reports containing factually correct information that nonetheless mis- lead their readers are neither maximally accurate nor fair to the consumer who is the subject of the reports.”). To support this remarkable theory, Equifax and the major- ity cite Bibbs v. Trans Union LLC, 43 F.4th 331 (3d Cir. 2022). But there is a key factual difference. In Bibbs, the credit report showed that an account was closed. For the same account, it also showed “Pay Status: >Account 120 Days Past Due Date<,” without showing clearly whether that “past due” statement was current or historic. Id. at 343. The Third Circuit resolved the ambiguity by concluding that a reasonable reader would realize that the account had been closed but had been 120 days past due when it was closed. Id. at 343–44. Even assuming that Bibbs is correct where a credit report contains such an ambiguity, its reasoning does not reach a case of flat contradiction. That’s what we have here. The Equifax re- port on Frazier showed delinquent payments to DMI in seven different months after the DMI account was reported to have been closed. The reasonable reader knows that both cannot be correct, but she has no way to know which of the 26 No. 23-2355
contradictory points, if either, is correct. Neither the majority opinion nor Equifax offers an explanation that would let the reader resolve the contradiction with confidence. For pur- poses of the FCRA, the better rule is that a credit report con- taining such a patent contradiction is patently false. It needs to be corrected, not explained away. The majority’s reliance on this theory is also wrong on the facts. At the very least, it misapplies the summary judgment standard by resolving material factual disputes against the non-moving party. After noting the contradictory infor- mation supposedly in the Equifax report, the majority tries to avoid the obvious conclusion: So again, in “full context,” [Dovenmuehle Mort- gage, 72 F.4th at 777], Frazier’s status as “90-119 Days Past Due” did not mean that she was cur- rently delinquent on her debt when [Mutual Federal] bank reviewed the report. Therefore, the status is not materially misleading as a mat- ter of law, and the district court correctly con- cluded that the information in the Equifax dis- closure “was all true.” Ante at 9. The logic of the “therefore” does not follow. This key passage fails to confront the contradiction between the re- port of delinquent monthly payments (“90-119 Days Past Due”) and the reported settlement date. This contradiction cannot be resolved by assuming that a reasonable reader would con- clude (1) that the delinquent monthly payment reports meant something else entirely and (2) that she could reliably extend credit based on that other meaning. The contradiction cer- tainly cannot be resolved on summary judgment. See Guthrie v. PHH Mortgage Corp., 79 F.4th 328, 345 (4th Cir. 2023) No. 23-2355 27
(reversing summary judgment for information furnisher where corrected report failed to correct false listings of ac- count as delinquent in prior months). We need not speculate how a reasonable reader would in- terpret the conflicting information here. Evidence in the rec- ord shows that the notations in the Account History section of Frazier’s credit report did mislead people in the real world in just the way Frazier argues. CreditLink interpreted those ex- act same seven months on Frazier’s credit report as represent- ing “Late Dates.” Bodnar Decl. ¶¶ 4–5, Dkt. No. 113–1. Further, the majority’s theory for resolving this contradiction—as a matter of law, no less—does not explain how Frazier could have been 90 to 119 days delinquent in the listed seven months, but not in any of the other months after the reported settlement date. Equifax had a duty under the FCRA to “follow reasonable procedures to assure maximum possible accuracy of the in- formation concerning the individual about whom the report relates.” 15 U.S.C. § 1681e(b). So even if we overlook the fac- tual disputes about what Equifax reported to CreditLink, it should not be a defense to this claimed violation that the infor- mation in Equifax’s supposedly corrected credit report was so obviously wrong that a reasonable reader could not rely on it. 4
4 Because the district court decided as a matter of law that Equifax’s
report was accurate, it did not address whether Equifax adopted “reason- able procedures to assure maximum possible accuracy” or conducted a “reasonable reinvestigation” into Frazier’s dispute. See 15 U.S.C. §§ 1681e(b) & 1681i(a)(1)(A). Based on this record, Frazier has raised a 28 No. 23-2355
III. Factual Disputes About Causation The majority opinion also offers an independent basis for affirming summary judgment: Even if Equifax submitted pa- tently false and misleading information to CreditLink, its vi- olation of its FCRA duties did not cause Frazier any harm. The theory is that Mutual Federal denied her 2020 mortgage ap- plication for unrelated reasons. Ante at 12–14. The factual question of causation is disputed. Frazier of- fered evidence that the reports of three delinquent payments in the preceding twelve months meant that a mortgage loan would violate requirements of the U.S. Department of Hous- ing and Urban Development and would, as a practical matter, make it impossible for Mutual Federal to approve her appli- cation. See King Report at 2–3, Dkt. No. 86–17. The majority relies on evidence showing that Frazier’s debt-to-income ratio was too high, but it fails to reckon with her evidence showing that she and Mutual Federal were working to try to resolve that problem. The majority notes that her debt ratio was slightly less than one percent too high. Ante at 13. If Frazier had been able to pay off credit card debt of about $1800, for example, that would have lowered her debt-to-income ratio
genuine dispute of material fact on this point. See Chaitoff, 79 F.4th at 816, 819 (explaining that reasonableness of both a credit reporting agency’s procedures and reinvestigation is a question for the jury “unless reasona- bleness is beyond dispute”); see also Collins v. Experian Info. Sols., Inc., 775 F.3d 1330, 1332–33 (11th Cir. 2015), denying reh’g sub nom. due to waiver, Collins v. Equable Ascent Fin., LLC, 781 F.3d 1270 (11th Cir. 2015) (where credit reporting agency used automated process to verify information from lender, “an issue of material fact remained as to whether [reporting agency’s] investigation was reasonable when it disregarded … infor- mation [consumer] provided and instead relied solely on [lender] to verify the debt”). No. 23-2355 29
enough that she could have been approved. King Report at 3– 4, Dkt. No. 86–17. On causation, Frazier relies on the report of an expert on consumer mortgage loans. Equifax moved to exclude that re- port as inadmissible, but the district court did not rule on that motion. The majority describes a key part of the report as “un- supported” and disregards it as inadmissible, while denying that it is judging the persuasiveness of the report. The major- ity seems to be deciding the Daubert motion in the first in- stance. That’s a further departure from sound summary judg- ment practice. As the case comes to us, plaintiff is entitled to rely on that expert report, and it is for a jury to decide whether its conclusions are persuasive or not. See generally, e.g., An- derson v. Liberty Lobby, Inc., 477 U.S. 242, 255 (1986) (“Credibil- ity determinations, the weighing of the evidence, and the drawing of legitimate inferences from the facts are jury func- tions, not those of a judge, whether he is ruling on a motion for summary judgment or for a directed verdict.”). Together with Frazier’s own testimony and the evidence from Mutual Federal, the report is sufficient to raise a genuine issue of ma- terial fact on causation. To be clear, the majority and I agree that potential lenders were entitled to know that Frazier had fallen behind on her mortgage back in 2015 and had settled that account for less than the full debt. But Frazier was entitled to have potential lenders receive accurate information and to know that she had been current in paying debts in the following years. The majority concludes: “Equifax’s consumer report and file of Frazier’s credit history did not contain inaccuracies.” Ante at 14. That assertion is simply wrong as a matter of in- disputable fact. Equifax falsely listed and reported those 30 No. 23-2355
seven delinquent monthly payments in the years after the ac- count was settled and closed. That was the difference between DMI correctly reporting “dashes” and Equifax falsely report- ing “dates.” Accordingly, we should reverse summary judg- ment and return this case to the district court for trial. I re- spectfully dissent.