The Pandemic Press Scorecard

The pandemic gave reporters two beats at once: the economic rescue and the health response. The beats looked different. The test was the same.

Did the reporter keep the nouns straight: potential fraud or proved fraud, lender or processor, EUA or full approval, authorization or mandate, aerosols or droplets, content moderation or state action, a vendor estimate or a government estimate, a staff report or an adjudicated fact?

How the grades work. This scorecard grades the published record, not private motive. Each entry rests on named pieces, read against the later official record: inspector-general reports, GAO estimates, court dockets, agency statements and a congressional committee's own releases. Five things decide a grade: documented factual errors, weighted by where they sat; whether the outlet corrected them; how the piece handled the source hierarchy; whether it updated when the record changed; and whether it claimed more certainty than the evidence allowed on the day it ran. A correction made counts for the reporter. A correction owed and not made counts against. An A means no located error across years of work, plus outside recognition or a call that agencies or auditors later adopted. A D means a headline or lead figure that an official inquiry later found unsupported and that stayed uncorrected, or repeated large errors in the same direction. The research cutoff is September 30, 2026. The full method and sources follow the entries.

The ranking

Each name links to its entry below.

1. Ed Yong, The Atlantic - A

Beat: Health response · Record: No error located; 2021 Pulitzer

Yong sits first because he did the hardest kind of pandemic journalism: he explained the system while the system was still moving.

His Atlantic series won the 2021 Pulitzer Prize for Explanatory Reporting, "for a series of lucid, definitive pieces on the COVID-19 pandemic that anticipated the course of the disease, synthesized the complex challenges the country faced, illuminated the U.S. government's failures and provided clear and accessible context for the scientific and human challenges it posed."¹ Many later pandemic arguments were over single artifacts: one model, one mask rule, one vaccine document, one court order, one fraud estimate. Yong's best work showed how those artifacts fit into a larger machine: testing capacity, PPE shortages, hospital strain, federal coordination, risk communication, and public trust.

Good: he got the big nouns right early: testing, PPE, hospital capacity, transmission, vaccine timelines, and institutional failure.

Bias: toward system failure as the master frame. That was mostly right in 2020, and it left less room for later counterfactuals about the costs of school closure, mandates, and prolonged restrictions.

Wrong: nothing central in the reviewed record.

Correction record: none owed.

Back to the ranking

2. Helen Branswell, STAT - A

Beat: Infectious disease · Record: No error located; Polk Award

Branswell is the outbreak reporter's reporter. She covered infectious disease before COVID-19 was a national identity test, and it showed.

STAT announced in February 2021 that Branswell had won a George Polk Award for Public Service for pandemic coverage dating back to the earliest reports of unexplained pneumonia in Wuhan.² Her value was pattern recognition: variants, vaccines, public-health agencies, WHO, CDC, flu-season analogies, and the difference between uncertainty and ignorance.

Good: early, technically careful infectious-disease coverage with the right level of alarm.

Bias: institutional fluency can become institutional patience. Her work was built for outbreak science, less for the adversarial record fights that followed: EcoHealth, FOIA, mandates, jawboning.

Wrong: no central failure in the reviewed record.

Correction record: none owed.

Back to the ranking

3. Lauren Hepler and Stephen Council, CalMatters - A

Beat: UI, EDD, Bank of America · Record: No error located; state auditor confirmed the story

CalMatters' California unemployment reporting understood the core trade before most national coverage did: anti-fraud controls and claimant harm were the same story seen from different ends of the payment rail.

In October 2020, Council wrote the "freeze first, verify second" version of the EDD story: California and Bank of America were trying to stop fraud, and real claimants were being frozen out of money they had already been told was theirs.³ In November, Hepler and Council followed the bank side: Bank of America's exclusive prepaid-card contract, its fraud-monitoring promises, and the finding that "to this day, no one knows how much the bank has made off the deal."⁴

Good: the shape of the UI problem. Pandemic UI was both under-controlled and over-blocking.

Bias: toward the human case file. That is defensible on an unemployment beat, and they did not use it to deny fraud.

Wrong: nothing central in the reviewed record.

Correction record: none owed. Two months after the November story, the California State Auditor reported that EDD had paid about $10.4 billion on claims it had since determined may be fraudulent, and that $1 billion of it followed EDD's decision to remove a safeguard against paying claimants whose identities it had not confirmed.⁵

Back to the ranking

4. Zeynep Tufekci, The Atlantic / New York Times - A-

Beat: Masks, aerosols, institutional failure · Record: No error located; early on ventilation

Tufekci was early on two of the pandemic's biggest health-communication failures: masks and airborne transmission.

Her work pushed against early official caution that treated public masking as confusing or supply-threatening rather than as a harm-reduction tool. She pressed the aerosol and ventilation point before agencies said it plainly. UNC's account of her pandemic work says she "wrote about the role of ventilation and the behavior of COVID-19 clusters before those points were incorporated into how the CDC made recommendations," and that she "publicly pushed back on those who shamed people for visiting parks and beaches."⁶ Her Atlantic archive is a running fight with agencies that were too slow, too timid, or too complicated.⁷

On origins she came back to the record in March 2025, after the CIA's shift. She wrote that "to promote the appearance of consensus, some officials and scientists hid or understated crucial facts," and that "to this day, there is no strong scientific evidence ruling out a lab leak or proving that the virus arose from human-animal contact in that seafood market."⁸

Good: masks, aerosols, ventilation, and the idea that public-health communication should tell the truth even when the truth creates operational problems.

Bias: the same posture that caught official timidity can make agency constraints look simpler than they were.

Wrong: no major factual failure in the reviewed record.

Correction record: none located.

Back to the ranking

5. Derek Willis and Lydia DePillis, ProPublica - A-

Profiles: Derek Willis, Lydia DePillis · Beat: PPP, Kabbage · Record: No error located; DOJ settlement followed

ProPublica's Kabbage fake-farms investigation is the model for pandemic-relief reporting that aged well.⁹

The story had loans, names, places, business registries, and absurd facts: Kabbage had sent hundreds of PPP loans to supposed farms in places where the farms did not appear to exist.

That reporting had consequences. The House Select Subcommittee opened its inquiry into Kabbage, BlueVine and their partner banks ten days later, "following reporting by ProPublica and other news outlets."¹⁰ In May 2024, DOJ announced that Kabbage's successor would pay up to $120 million to resolve False Claims Act allegations, and admitted specific loan-calculation errors: double-counting state and local taxes in gross wages, failing to exclude compensation above $100,000 per employee.¹¹

Good: specific loans and specific lender conduct. The reporting did not need to turn every flag into a conviction.

Bias: later coverage used Kabbage as a bridge to "fintechs" generally. That was not their doing, but their work became the first link in a chain that later covered weaker claims with stronger ones.

Wrong: the core Kabbage story held up.

Correction record: none owed. The gap that followed belongs to the press system: once "online lenders" became "fintech fraud," the banks of record did not receive equivalent national scrutiny.

Back to the ranking

6. Carmen Reinicke, CNBC - B+

Beat: PPP access, borrower pipeline · Record: No error located in one piece

Reinicke's May 2021 PPP-exhaustion piece caught a story the later fraud cycle mostly erased: millions of borrowers were stranded in the pipe when the money ran out.¹²

The piece walked through the pipelines at Customers Bank, Fountainhead, Cross River, Kabbage and Biz2Credit, and at Womply, "a fintech that matches borrowers with lenders," which had 2.5 million applications in its system: "of those, 1.6 million are in the hands of lenders that can't send them to the Small Business Administration."¹²

Good: the operational facts, and the gap between "application started" and "loan funded," a distinction that later mattered to both the FTC's theory and the academic fraud-flag debate.

Bias: toward borrower frustration and program failure, which was the piece she was writing.

Wrong: nothing in the reviewed piece. One piece was reviewed, which caps the grade.

Correction record: none owed. CNBC as an outlet also did strong later work on ERC promoters, especially Innovation Refunds, after the IRS moratorium exposed the mill economy.¹³

Back to the ranking

7. Stacy Cowley and Ella Koeze, New York Times - B

Profile: Stacy Cowley · Beat: PPP banks, lenders, processors and fees; fraud screening; House oversight · Record: No error located in three pieces; Dec. 2022 story did not disclose a quoted lender's fee arbitration; no update

On April 22, 2020, in the program's first round, she and Emily Flitter reported that JPMorgan Chase, Citibank and U.S. Bank had put their wealthiest clients' applications ahead of everyone else's, according to half a dozen bank employees and financial industry executives. One person familiar with JPMorgan's operation called the service "concierge treatment." At JPMorgan, they wrote, "nearly all of the 8,500 commercial and private banking clients who applied for a loan got one," while "only 18,000 of more than 300,000 small-business banking customers who applied through Chase's retail bank, where they normally did business, got loans, according to the bank."¹⁴

The three pieces graded here cover the 2021 round: the processors' fees, a fraud screen's flags and the House report.

In June 2021 she and Koeze reported on Blueacorn and Womply, two companies that between them "processed a third of all P.P.P. loans made this year." Their description was exact: "Blueacorn and Womply aren't banks, so they couldn't actually lend any money. Rather, they acted as middlemen, charging into a gap between what big banks wouldn't do and what small banks couldn't do."¹⁵ The story followed borrowers through the pipe, among them an Uber driver whose loan Fountainhead funded in seventeen days and another whose approved loan Benworth had not paid out. Its fee section rested on lenders, most of them unnamed: seven lenders' terms from "five people with knowledge of the financial arrangements," and one lender's complaint, "At those rates, I'm in the hole and losing money on many of these loans." It told readers that "Womply is locked in fee disputes with some of its partners."¹⁵

Her August 2021 story on the Texas fraud-indicator study was the careful version: "could be fraudulent," false positives acknowledged.¹⁶ Its four objecting lenders, Capital Plus, Prestamos, Cross River and Harvest, were defending their own loan books. One of them, Capital Plus, was sued over fees by Womply three weeks later, on Sept. 9, 2021; the case settled and was dismissed with prejudice on Jan. 5, 2023.¹⁷

Her Dec. 1, 2022 story covered the House Select Subcommittee's staff report on fintechs in PPP.¹⁸ Its named critic of Womply was Fountainhead's chief executive, Chris Hurn. The story quoted him telling House investigators that Womply's fraud systems seemed "put together with duct tape and gum," and reported that he "had to get a restraining order to prevent Womply from destroying the loan records, according to the report." The court record supports the order: on Oct. 7, 2021, a Florida state judge temporarily restrained Womply "from altering or destroying any of the Loan Files" pending a hearing, while making "no determination as to the merits."¹⁹ The hearing never took place: Fountainhead dismissed its suit with prejudice on Nov. 18, 2022, thirteen days before the story ran.²⁰ The story recalled the Times' 2021 coverage of "the disputes that developed between Womply and some of its lending partners" and said Womply did not respond to requests for comment. It did not tell readers that Fountainhead had been a respondent in Womply's fee arbitration since June 15, 2021, or that its suit had ended.²¹²² The staff report does not mention Womply's arbitrations with Fountainhead or Benworth.²³ In May 2023 the arbitrator in Womply's case against Benworth refused to admit the report, writing that it "relies heavily upon statements by Benworth and Fountainhead, both of which were in active arbitration with Womply."²⁴

The companies the report grouped together met different ends. Blueacorn's co-founders were each sentenced to ten years: Stephanie Hockridge after a jury convicted her of wire-fraud conspiracy on June 20, 2025, and Nathan Reis after a guilty plea in August 2025.²⁵²⁶ Kabbage's successor settled False Claims Act allegations.¹¹ SBA suspended Womply on Dec. 8, 2022.²⁷ In April 2024 Womply settled the FTC's "PPP Fast Lane" advertising case with a $26 million judgment and no admission or denial,²⁸ and in June 2024 it won its contract claims against Benworth at a lower fee tier than it had sought.²⁹ The outcome of Womply's arbitration against Fountainhead is not in the public record.

Good: the 2021 round's fee economics and middleman model, the borrowers in the pipe, and a careful account of the fraud-indicator study. The 2021 fee story reported the fee disputes alongside its anonymous lender quotes.

Bias: toward lenders' accounts in their fee fights with a processor. In December 2022 the named voice on Womply's systems belonged to a lender then in fee arbitration with Womply, and readers were not told.

Wrong: no false sentence located in the three pieces reviewed. The restraining-order line matches the Oct. 7, 2021 order. The December 2022 story left out its source's stake in the dispute.

Correction record: none located. The June 2021 story carries an Oct. 11, 2021 update stamp; its text matches an Aug. 6, 2021 archived copy, with no correction note. A July 2025 archived copy of the December 2022 story shows its original date and no update line. No stand-alone Times follow-up incorporating the FTC case or the arbitration was located. The missing source context and the missing update place the entry in the B band. The Times should add to its December 2022 story that Fountainhead, whose chief executive it quoted on Womply's systems, was then a respondent in Womply's fee arbitration, and that Fountainhead had dismissed its Florida suit with prejudice on Nov. 18, 2022.

Back to the ranking

8. Zachary Warmbrodt, Politico - B+

Beat: PPP politics, banks, SBA oversight · Record: No error located in 30 pieces, 2020–21

Warmbrodt's PPP coverage shows how the program looked from Washington in April 2020: banks warning of chaos, bankers pleading with the administration, lenders pressing Treasury and SBA to fix glitches.³⁰ His April 1, 2020 piece on Wall Street firms lobbying for access to the small-business program treated finance as an interested actor.³¹

He reported the verification trade-off as it was made. On April 2, 2020 the banks' "paramount concern" was "the extent to which they'll be expected to vet borrowers," and they wanted assurance that they "won't be held liable if a borrower obtains a loan after providing misleading information." The next day's guidelines "offered relaxed loan-verification rules." By April 30 he wrote that the design "required lenders to sacrifice traditional due diligence to an extent, exposing the program to potential fraud."³² He covered the watchdogs as they caught up: GAO's June 2020 warning of "a significant risk that some fraudulent or inflated applications were approved," and the Inspector General's October finding that SBA had "lowered the guardrails" on EIDL, with loans approved "with little to no vetting."³³

None of the pieces reviewed names the check that was missing. The CARES Act barred SBA from requiring EIDL applicants to submit tax returns, no PPP rule required an IRS check, and SBA began pulling IRS transcripts for EIDL only on April 28, 2021.³⁴ GAO's June 2020 report said SBA "allowed lenders to rely on borrower certifications to determine borrowers' eligibility, raising the potential for fraud."³⁵ His story on that report quoted its fraud-risk sentence, not that one.

Good: the launch, the verification trade-off in real time, and the fights between SBA and its watchdogs.

Bias: protagonist selection. Verification enters as the banks' liability and delay problem, and the 2021 fraud checks as a "processing backlog" the banks warned about.

Wrong: no error located in 30 pieces from April 2020 to July 2021.

Correction record: none owed on the text reviewed.

Back to the ranking

9. Katherine Eban and Jeff Kao, Vanity Fair / ProPublica - B

Beat: Origins, WIV, EcoHealth · Record: Two assertions removed by editor's note; translation caveat added

Eban's Vanity Fair work, later with Kao at ProPublica, moved the lab-origin debate from vibes to documents: Senate staff materials, Wuhan Institute of Virology dispatches, EcoHealth records, and the gap between public certainty and private uncertainty.³⁶

Good: the reporting made a formerly taboo question documentable. Eban's June 2021 Vanity Fair investigation opened on the argument over whether to investigate at all: State Department investigators, she reported, "were repeatedly advised not to open a 'Pandora's box.'"³⁷ It showed that public confidence had outrun the record; it did not prove the origin of SARS-CoV-2. ODNI's 2023 report, with agencies split three ways, made the same point.³⁸

Bias: the piece leaned on one reader's translation of its documents.

Wrong: two assertions in the October 2022 piece did not survive. ProPublica's editor's note says the story "has been corrected to remove assertions that Gabriel Gras was the last French expert at the Wuhan Institute of Virology and that Chinese military vaccinologist Zhou Yusen was the first researcher in the world to apply for a patent for a vaccine against COVID-19." After observers challenged the translation of a Nov. 12, 2019 party dispatch, three Chinese-language experts commissioned by the outlets "said the passage was ambiguous" and the translation "was plausible but not the only way to represent it."³⁶

Correction record: corrected, with the caveats appended in the text. The central claim, that the record was less settled than public officials said, stands; two supporting details and one translation did not.

Back to the ranking

10. David Zweig - B

Outlet: New York / The Free Press / book · Beat: Schools · Record: No error located; reviewed through secondary accounts

Zweig covered school closure, one of the largest health-policy decisions of the pandemic and one of the slowest press corrections.

He argued, early and persistently, that the evidence on child risk, in-school transmission and European reopening did not support prolonged American closures. By 2025, reviews of his book An Abundance of Caution in New York magazine and The 74 treated the closures as a mistake and the question as how it happened.³⁹⁴⁰

Good: he forced a cost ledger into a conversation that often treated "abundance of caution" as self-justifying.

Bias: his frame can make the decision look cleaner in retrospect than it felt in real time. Early 2020 uncertainty, teacher risk, and household vulnerability were not imaginary.

Wrong: nothing located. This entry rests on the 2025 reviews rather than on a re-reading of his 2020 and 2021 pieces, which keeps it out of the B+ band.

Correction record: none located.

Back to the ranking

11. Ken Schwencke, ProPublica - B-

Beat: PPP report week · Record: No error located; no later update

Schwencke's December 2022 report-week piece did some things better than the national pack. It identified the House inquiry as one that "began after reporting by ProPublica," and it compressed the strongest Blueacorn facts: "lax anti-fraud standards, executives who cashed in for themselves and contempt for small loan applications that would generate minimal fees. 'Delete them,' one executive wrote."⁴¹ The Blueacorn convictions that followed in 2025 supported that part of the report.²⁵²⁶

The staff report was not an adjudication. It bundled Blueacorn, Kabbage, Womply and BlueVine into a single "fintech" story that the later legal record split apart: a criminal-founder case, a False Claims Act settlement, and an advertising settlement plus an arbitration award that went against the lender.

Good: the Blueacorn material.

Bias: toward the report as the organizing record. Understandable on release day, less so afterward.

Wrong: less a false sentence than a missing second act.

Correction record: no ProPublica correction located, and none owed on the text. The update owed is the later record: which allegations became convictions, which became settlements, and which were never tried.

Back to the ranking

12. Ben Wieder, Miami Herald / McClatchy - C+

Beat: PPP lenders, bank fees, borrower eligibility, Kabbage, Benworth · Record: One uncorrected name error (syndicated copies); one correction appended; 2024 Rubio story's vetting line did not reflect a public award; no update

Wieder's PPP stories reviewed here run from September 2020 to September 2025. They follow the program through its lenders: the banks' fees, red flags among an online lender's borrowers, loans to "a felon, an alleged drug dealer and a comic," and one Miami lender's profits and politics. No fabricated claim was located.

His September 2020 Kabbage reporting with Meghan Bobrowsky came eight months before ProPublica's fake-farms story and was among the pieces the subcommittee cited when it opened its inquiry.⁴²¹⁰ The broad Kabbage line was later vindicated by DOJ's settlement.¹¹ His December 2020 analysis with Kevin G. Hall and Nicholas Nehamas put the fee story on the banks first: more than $18 billion in PPP fees to the nation's banks.⁴³

He also got the verification gap into print early. A December 2020 analysis by Wieder and three colleagues for the Herald and the Anti-Corruption Data Collective quoted Jim Richards, a former risk officer at Wells Fargo and Bank of America: "Banks were told they didn't have to look." Richards blamed the government for not using information it already held, "such as tax filings from prospective borrowers."⁴⁴ GAO had said it in June 2020: SBA "allowed lenders to rely on borrower certifications to determine borrowers' eligibility, raising the potential for fraud."³⁵ In the stories located, the point stayed a source's quote and did not return for EIDL or the ERC.

From 2023 to 2025 he gave Benworth's own conduct sustained attention. He covered the Federal Reserve Bank of San Francisco's suit, which said more than $49 million in dividends went to Bernie Navarro,⁴⁵ and Navarro's 2025 settlement with the Fed.⁴⁶ In 2024 he reported Senator Marco Rubio's 2021 letter to SBA on the lenders' fee fight with Womply, written the year Rubio borrowed $850,000 from Benworth.⁴⁷ A May 2023 story reported that William Manger, who as SBA's chief of staff oversaw the program's implementation, later did work for Womply; Womply's lawyer told the Herald it was as a paid "testifying expert" in the Benworth arbitration.⁴⁸ Each story named the fee dispute, and the 2024 and 2025 stories reported the $118 million award.⁴⁸⁴⁶

None reported what the arbitrator decided on the merits. The award denied every Benworth counterclaim for money, including a demand for more than $420 million back, and found that "Benworth was always responsible for conducting a good faith review of a borrower's application."²⁹ A year earlier the arbitrator had refused to admit the House staff report as evidence, calling its content "rank hearsay": it "relies heavily upon statements by Benworth and Fountainhead, both of which were in active arbitration with Womply."⁴⁹ Both rulings were on a public federal docket from July 1, 2024. Five weeks later the Herald described Benworth's loans as "processed and ostensibly vetted by Womply," and gave Womply's reply one sentence.⁴⁷

His July 24, 2024 story on the Federal Reserve suit said the congressional investigation "was triggered in part by the Miami Herald's reporting."⁴⁵ That is accurate for the inquiry's opening: the Select Subcommittee's May 2021 letters on Kabbage and BlueVine cited reporting by Bloomberg, the Miami Herald, the Project on Government Oversight and ProPublica.¹⁰ Its second phase, on Womply and Blueacorn, came six months later, and the subcommittee tied it to a University of Texas at Austin study, not to press reporting.⁵⁰

His Sept. 12, 2025 story on Bernie Navarro's settlement with the Federal Reserve Bank of San Francisco, as syndicated by AOL and InsuranceNewsNet, identified Blueacorn's co-founders as "Nathan Reis and Jennifer Lockridge."⁴⁶ The defendant is Stephanie Hockridge, also known as Stephanie Reis, convicted by a jury on June 20, 2025 and sentenced on Nov. 21, 2025.²⁵

Good: Kabbage, the bank-fee analysis, the 2020 verification quote, and sustained reporting on Benworth itself.

Bias: toward the lenders' side of their fee fight with a processor. The vetting failure sat with the processor in the paper's own voice; the arbitration's finding put the review duty on the lender.

Wrong: the August 2024 Rubio story's vetting line. It put the lenders' account in the paper's voice five weeks after a public award placed the review duty on Benworth, and no update followed.⁴⁷²⁹ The "Jennifer Lockridge" error in the syndicated copies of the September 2025 story.⁴⁶ The "triggered in part by the Miami Herald's reporting" line is accurate about the inquiry's opening and silent about how its second phase began.⁴⁵⁵⁰

Correction record: one correction appended, to the Aug. 6, 2024 Rubio story, which had misstated when the senator repaid a Benworth loan.⁴⁷ The name error stood in both syndicated copies reviewed, and no correction of it was located; the Herald's own page was not reviewed. McClatchy should correct the Hockridge name error in the syndicated copies, and in the source version if it appears there. The update owed on Benworth is the award's merits: the counterclaims denied, the review duty on the lender, and the staff report excluded as evidence.

Back to the ranking

13. Matt Taibbi, Michael Shellenberger, Bari Weiss, and the Twitter Files team - C+

Outlet: Substack / The Free Press · Beat: Platform moderation · Record: Not re-reviewed for this revision

The Twitter Files put internal platform communications into public circulation. For pandemic health policy, that meant visible records about moderation decisions, public-health claims, outside pressure, internal doubt, and the channel between government officials and private platforms.

The Twitter Files are a source set. They can show what Twitter employees saw, how they discussed requests, and how internal rules worked. They cannot, alone, prove what the government knew, what other platforms did, or whether a constitutional violation occurred.

Good: they made platform moderation legible. Without the documents, the public-health speech fight would have had more suspicion and fewer exhibits.

Bias: curated access. The reporters worked from a set selected and released under conditions shaped by Twitter's new owner and by their own editorial choices. The documents are real; the frame is incomplete.

Wrong: the broadest readings treated internal platform documents as if they settled the state-action question. The Supreme Court disposed of Murthy v. Missouri on standing in June 2024 and did not reach the First Amendment merits.⁵¹

Correction record: the correction owed is methodological: every Twitter Files claim checked against discovery, depositions, court orders, and platform policy.

Back to the ranking

14. Richard Lardner, Jennifer McDermott, and Aaron Kessler, AP - C

Beat: Fraud totals · Record: Frame outran caveats; no note

The AP's June 2023 "Great Grift" package is better inside than outside.

Outside, the package said that "fraudsters potentially stole more than $280 billion in COVID-19 relief funding; another $123 billion was wasted or misspent," and called the combined figure "10 percent of the $4.2 trillion the U.S. government has disbursed in pandemic aid." The lede called it "the greatest grift in U.S. history."⁵²

Inside, the reporters quoted Michael Horowitz, chair of the Pandemic Response Accountability Committee, telling Congress in February 2023 that fraud was "clearly in the tens of billions of dollars" and might eventually exceed $100 billion.⁵² Those two numbers sat in the same package. AP used the qualifier "potentially," but the display copy compressed proved fraud, potential fraud, improper payment, waste and estimate into one figure. SBA's Inspector General, in the same month, put potential fraud across EIDL and PPP at more than $200 billion, and called it that: potentially fraudulent, not adjudicated theft.⁵³

Good: fraud was vast, controls were weak, and the article contained countervailing voices.

Bias: toward the cinematic superlative.

Wrong: the $280 billion frame treated potential-fraud screens as close enough to theft totals to carry the story.

Correction record: none located, and no retraction was owed. The editor's note owed is narrow: explain that $280 billion aggregates potential-fraud estimates and should not be read as adjudicated theft. The gap between the package's number and the oversight chair's number inside it was the place to start.

Back to the ranking

15. Austin Fast and Sacha Pfeiffer, NPR - C-

Beat: PPP fintech report week · Record: One correction (quote); lender/LSP sentence unchanged

NPR's December 2022 fintech-fraud story carried the company responses, the underserved-borrower point, and the fact that fintechs "reached more independent contractors, as well as businesses run by women and people of color, than long-established banks did."⁵⁴

But one sentence should not still be standing without clarification. NPR wrote that fintechs "told Congress they could issue PPP loans to struggling small businesses faster than traditional banks — and they did."⁵⁴ Read as a statement about the whole category, the sentence is defensible. PayPal, Square and Intuit were approved as direct PPP lenders in April 2020, and the House report says fintechs acted as both lenders and lender service providers.⁵⁵ But the story is built around the House report's allegations against Womply and Blueacorn, and neither was a lender of record; the lenders they worked for issued the loans. AP's same-week account drew the line: "The financial technology firms reviewed PPP applications for lenders, which would ultimately distribute PPP money to businesses."⁵⁶ SBA's own statement a week later called them "non-lenders Blueacorn and Womply, companies that worked with PPP lenders."²⁷

Good: borrower-access context and company responses.

Bias: toward the House report as confirmation rather than allegation.

Wrong: the issue-loans sentence is overbroad. If a processor reads like the issuer of the loan, the bank's legal role gets smaller in the reader's mind.

Correction record: one correction. NPR appended it on Dec. 12, 2022: "Due to incorrect information provided by the House Select Subcommittee on the Coronavirus Crisis, a quote by the CEO of Prestamos was incorrect in an earlier version of this story." The lender/LSP sentence was unchanged in our copy of the page, made June 25, 2026.⁵⁴ A fair clarification would not say fintechs never issued PPP loans. It would say some fintech lenders did, while Womply and Blueacorn worked for the lenders that issued the loans at issue.

Back to the ranking

16. Tony Romm, Washington Post - C-

Profile: Tony Romm · Beat: Covid relief oversight · Record: Headline carried the staff report's conclusion; no update

Romm's pandemic-relief work had virtues: he followed oversight bodies, DOJ, congressional investigators, and program administrators over a long period, and treated pandemic fraud as a government-performance story.

His Dec. 1, 2022 story ran under the headline "'The more you submit, the more we get paid': How fintech fueled covid aid fraud," with a subhead that firms "allegedly collected taxpayer-funded fees as they overlooked signs of grift."⁵⁷ The subhead hedged; the headline did not. The story carried KServicing's statement that it "vigorously disputed" the fraud allegations and Chairman Clyburn's statement of the report's conclusions.⁵⁷

Good: the scale and persistence of pandemic-relief oversight failures.

Bias: toward official oversight sources as the story's spine.

Wrong: the headline stated a staff report's conclusion as the finding. It was a plausible allegation and a good same-day story.

Correction record: none located on the headline, and no stand-alone update incorporating the SBA suspension, the FTC case, the arbitration award, or the Blueacorn convictions was located.

Back to the ranking

17. Ken Dilanian and Laura Strickler, NBC News - C-

Beat: PPP, UI, fraud enforcement · Record: $400 billion ceiling repeated; no update

NBC's March 2022 PPP fraud piece had a good instinct and a misordered denominator.⁵⁸

The instinct: fraud was big, early, and under-controlled, and DOJ was overwhelmed.

The denominator: the headline called PPP the "biggest fraud in a generation." The opening put PPP theft at "as much as $80 billion — or about 10 percent — of the $800 billion," then "the $90 billion to $400 billion believed to have been stolen from the $900 billion Covid unemployment relief program," then "another $80 billion potentially pilfered from a separate Covid disaster relief program."⁵⁸ The piece did name that separate program lower down: "the SBA's inspector general has identified $78.1 billion in potentially fraudulent Economic Injury Disaster Loans, another Covid relief program for businesses."⁵⁸ So the denominator was in the story; the headline and lede put PPP first when the agency's own figure already put EIDL level with it. Fifteen months later the Inspector General's 2023 estimate made the order plain: more than $136 billion in potentially fraudulent COVID-EIDL funds (33 percent of disbursements) against $64 billion in PPP (8 percent).⁵³ EIDL was a government-run program with no private lenders or processors to blame. GAO's later UI estimate was $100 billion to $135 billion, not $400 billion.⁵⁹

Good: speed and weak controls created real fraud.

Bias: prosecutor-and-watchdog sourcing, with private actors more visible than government-administered failure.

Wrong: the piece repeated a UI range that ran to $400 billion, the vendor-shaped ceiling NBC had carried as a "credible estimate" the previous August, and that GAO and the House committees later undercut.⁶⁰⁵⁹⁶¹

Correction record: no correction or update located. An update was owed on two points: the $400 billion ceiling, which NBC had also carried in August 2021 as the top of a range of "credible estimates," and the EIDL-PPP order once the Inspector General published the split.

Back to the ranking

18. Matt Egan, CNN Business - C-

Beat: PPP fintech fraud study · Record: Headline outran a proxy study; no update

CNN's August 2021 piece took the Griffin, Kruger and Mahajan study and told readers, in its headline, that fintechs "have a fraud problem with their Covid relief loans." The story said fintechs were "almost five times more likely than traditional banks to have made 'highly suspicious' loans" and that researchers "found that 1.8 million PPP loans with a total value of $76.3 billion had suspicious characteristics."⁶²

The study was serious and later appeared in the Journal of Finance. Its indicators were proxies for fraud, not proof, and the researchers said so. The Times' account of the same study said the $76 billion "contains some false positives, the researchers acknowledged," gave the example (multiple loans to businesses at one residential address), and reported that all four named lenders "strongly objected to the study's methodology, data and conclusions." The residential-address indicator can catch fraud and can also count sole proprietors and home-based businesses, the borrowers fintech channels served.¹⁶ CNN's body copy hedged ("may also have opened the door"), quoted the study's "highly suspicious" in quotation marks, and carried Cross River's statement that it had "deployed fraud detection standards that far exceeded" SBA's requirements.⁶² The headline did not hedge.

Good: the study was newsworthy and the responses were in the story.

Bias: toward the clean study headline.

Wrong: "fraud problem" in the headline arrived before the evidentiary standard did. One piece was reviewed.

Correction record: none located. The update owed is the distinction between flagged loans, funded fraud, and the enforcement record that followed.

Back to the ranking

19. Apoorva Mandavilli, New York Times - D+

Beat: Vaccines, children, origins · Record: Two large corrected errors, same direction; deleted tweet

Mandavilli covered the science beat through the pandemic; the grade rests on the size and direction of the errors that were corrected, not on one tweet.

An October 2021 Times story said "nearly 900,000 children have been hospitalized with COVID-19 since the pandemic began."⁶³ The corrected figure was "more than 63,000 children were hospitalized with Covid-19 from August 2020 to October 2021." The same correction fixed a description of Swedish and Danish regulators' actions on the Moderna vaccine and the date of an FDA meeting.⁶⁴ The original figure was about fourteen times the corrected one, in the direction of greater child risk.

Seven months later a correction fixed a May 2022 story that said "nearly 4,000 children aged 5 to 11" had died of MIS-C. The children had been diagnosed with MIS-C, not killed by it. The CDC count cited in coverage of the correction was 8,210 cases and 68 deaths across all ages.⁶⁵

In May 2021, Mandavilli tweeted, "Someday we will stop talking about the lab leak theory and maybe even admit its racist roots. But alas, that day is not yet here." She deleted the tweets on the night of May 26 and said they were "badly phrased."⁶⁶ Three months later, every U.S. intelligence agency assessed that "two hypotheses are plausible: natural exposure to an infected animal and a laboratory-associated incident."⁶⁷ ODNI's June 2023 declassified report said "the Department of Energy and the Federal Bureau of Investigation assess that a laboratory-associated incident was the most likely cause of the first human infection with SARS-CoV-2, although for different reasons," while other agencies assessed natural exposure as most likely and the CIA and another agency "remain unable to determine" the origin.³⁸ In January 2025 the CIA said it favored a research-related origin "with low confidence."⁶⁸ The question was open when she tweeted and is open in September 2026.

Good: she covered a hard science beat and published useful reporting on masks, vaccines, and CDC guidance.

Bias: toward high-risk framing around children and toward treating lab-origin discussion as culturally contaminated.

Wrong: the pediatric hospitalization and MIS-C errors directly affected readers' sense of child risk, and both pointed the same way. The tweet fails the certainty test: it treated a hypothesis that no investigation had ruled out as a subject to be retired.

Correction record: the Times corrected both errors, and that counts in her favor against the reporter below her, who corrected nothing. The tweet was deleted, not corrected in print. The Times should add an editor's note on its two corrected pediatric-risk errors, which both overstated child risk.

Back to the ranking

20. Felix Salmon, Axios - D-

Beat: UI fraud · Record: Headline figure found unsupported by Congress; no note

In June 2021, Axios ran the headline "Half of the pandemic's unemployment money may have been stolen" and reported that fraud "could easily reach $400 billion, according to some estimates." The estimate had one named source: "Blake Hall, CEO of ID.me, a service that tries to prevent this kind of fraud, tells Axios that America has lost more than $400 billion to fraudulent claims. As much as 50% of all unemployment monies might have been stolen."⁶⁹ ID.me sold identity verification to state unemployment systems.

UI fraud was real. GAO estimated it at $100 billion to $135 billion, about 11 to 15 percent of benefits paid, for April 2020 through May 2023.⁵⁹ That is roughly one-quarter to one-third of the Axios ceiling.

The harder finding came from Congress. In November 2022 the House Oversight Committee and the Select Subcommittee on the Coronavirus Crisis said that "despite repeated requests, ID.me could not provide the Committees with any methodology" for the $400 billion or 50 percent claim, that the sources ID.me cited did not support the total, and that one of them, a Heritage Foundation estimate, had itself relied on ID.me's earlier figure.⁶¹

Good: state UI systems were exposed, and the piece said fraud-detection vendors were a growth business.

Bias: toward a vendor source whose commercial interest should have made the number the subject, not the answer.

Wrong: the $400 billion, half-the-money frame. One piece was reviewed; it is the piece that mattered.

Correction record: none located. The page's only update line adds a White House quote: "This story has been updated with the quote from the White House."⁶⁹ A defensible editor's note would say the article's central figure came from a vendor whose estimate congressional investigators later found unsupported, and that GAO's estimate came in at a quarter to a third of it.

Back to the ranking

Other stories on the record

These stories are not ranked. Each is a clear item where the press corrected a concrete error, should have corrected one, or owed readers a later editor's note because the underlying record changed.

CNBC and AP: the Secret Service $100 billion figure

Public claim: on Dec. 21, 2021, CNBC reported that "criminals have stolen close to $100 billion in pandemic relief funds, the U.S. Secret Service said," diverted from PPP, EIDL and unemployment programs.⁷⁰

What the record showed: AP's story the same day gave the figure's composition. The estimate "is based on Secret Service cases and data from the Labor Department and the Small Business Administration," Roy Dotson, the agency's pandemic fraud recovery coordinator, told AP; "most of that figure comes from unemployment fraud," where "the Labor Department reported about $87 billion in unemployment benefits could have been paid improperly."⁷¹ The Secret Service was aggregating other agencies' estimates with its own casework. CNBC's account presented the total as the agency's finding and did not explain what it was made of.⁷⁰

Correction status: stories that treated the number as a Secret Service count of stolen funds owed readers the composition AP printed.

Washington Post / Paulina Firozi: "debunked"

Public claim: on Feb. 17, 2020, the Post's headline read "Tom Cotton keeps repeating a coronavirus conspiracy theory that was already debunked," and the story said Cotton was "fanning the embers of a conspiracy theory that has been repeatedly debunked by experts." The same story quoted Cotton saying, "Now, we don't have evidence that this disease originated there, but because of China's duplicity and dishonesty from the beginning, we need to at least ask the question." Its expert quotes addressed a weapon: "The possibility this was a deliberately released bioweapon can be firmly excluded," Richard Ebright said. The story also reported that Cotton had separated a weapon from "other possibilities such as a lab accident."⁷²

What later records showed: in August 2021 every U.S. intelligence agency assessed a laboratory-associated incident as plausible, and the community judged that the virus "was not developed as a biological weapon."⁶⁷ The weapon claim the experts dismissed stayed dismissed. The accident the headline called debunked stayed open.

Correction status: corrected. By late May 2021 the Post had appended: "Earlier versions of this story and its headline inaccurately characterized comments by Senator Tom Cotton regarding the origins of the coronavirus. The term 'debunked' and The Post's use of 'conspiracy theory' have been removed because, then as now, there was no determination about the origins of the virus." The headline was changed to "Tom Cotton keeps repeating a coronavirus fringe theory that scientists have disputed."⁷³

New York Times / Alexandra Stevenson: "Fringe Theory"

Public claim: on Feb. 17, 2020, the Times ran "Senator Tom Cotton Repeats Fringe Theory of Coronavirus Origins," under the subheading "Scientists have dismissed suggestions that the Chinese government was behind the outbreak." The story opened on the rumor "that the outbreak now afflicting people around the world had been manufactured by the Chinese government," said "the conspiracy theory lacks evidence and has been dismissed by scientists," and added that "on Sunday, it got its biggest public boost yet" from Cotton. It then quoted Cotton: "We don't have evidence that this disease originated there."⁷⁴

What later records showed: the manufactured-weapon rumor matches what the intelligence community rejected in 2021. The lab-accident possibility Cotton raised is one that every agency called plausible.⁶⁷ The body of the story got the weapon claim right. The headline put Cotton's question under the same label.

Correction status: none located. An archived copy from March 1, 2023 carries the same headline, a "Updated Feb. 18, 2020" line and no correction. The live page could not be re-read. The Times should append a note: the headline called Cotton's question a "fringe theory," and the story folded it into the rumor that the Chinese government had manufactured the virus.

NPR / Geoff Brumfiel and Emily Kwong: "Scientists Debunk Lab Accident Theory"

Public claim: on April 22, 2020, All Things Considered aired "Scientists Debunk Lab Accident Theory Of Pandemic Emergence." The web summary said "Scientists dismiss the idea that the coronavirus pandemic was caused by the accident in a lab," and the host's introduction said "scientists worldwide agree that the virus occurred naturally."⁷⁵ The next day's web story opened: "Virus researchers say there is virtually no chance that the new coronavirus was released as result of a laboratory accident in China or anywhere else." It ran a section headed "No conspiracy theories," quoting Peter Daszak: "It's a real shame that the conspiracy theories can get to the level they've got with policymakers."⁷⁶ The story identified Daszak as president of EcoHealth Alliance. It did not say that EcoHealth had channeled National Institutes of Health money to Shi Zhengli's laboratory in Wuhan.⁷⁷

What later records showed: in August 2021 all U.S. intelligence agencies called a laboratory-associated incident plausible.⁶⁷ In June 2025 the WHO's advisory group said it was "not in a position to rule this out as a possibility."⁷⁸

Correction status: none located. Archived copies of the April 22 page (Dec. 27, 2025) and the April 23 page (Sept. 26, 2026) carry no editor's note. NPR should add notes to both pieces.

CNN / Chris Cillizza: "crushed"

Public claim: on May 5, 2020, CNN published Cillizza's analysis "Anthony Fauci just crushed Donald Trump's theory on the origins of the coronavirus." It said Fauci "was definitive about the origins of the virus" and quoted him: the evidence "is very, very strongly leaning toward this could not have been artificially or deliberately manipulated … Everything about the stepwise evolution over time strongly indicates that [this virus] evolved in nature and then jumped species." Then: "Only one of these two people is a world-renowned infectious disease expert. And it's not Donald Trump."⁷⁹

What later records showed: Fauci's own words were "very, very strongly leaning," not definitive. The accident hypothesis remained plausible to every U.S. intelligence agency in 2021, and the CIA leaned toward it at low confidence in 2025.⁶⁷⁶⁸

Correction status: none on the CNN page, which as read on Sept. 30, 2026 shows "Updated May 5, 2020." Cillizza wrote in his own newsletter on Jan. 28, 2025 that he "screwed up": "it now appears that the reality TV-star-turned-president was right. And Anthony Fauci was wrong."⁸⁰ That later sentence also runs past the record: the CIA assessment it responded to was made at low confidence. CNN should add a note to the analysis.

Vox / Eliza Barclay: "debunked"

Public claim: on March 4, 2020, Vox published "The conspiracy theories about the origins of the coronavirus, debunked," with the subheading "There's a rumor the coronavirus started in a Chinese lab. And a scientific consensus it didn't." Its photo caption said, "Conspiracy theorists have conjectured the lab is the origin of the Covid-19 coronavirus outbreak; scientific evidence shows it's not."⁸¹

What later records showed: the same assessments as in the Washington Post and NPR items above.⁶⁷

Correction status: partly updated. Vox changed some wording in April 2020 without saying so, and added an editor's note on May 24, 2021: "Since this piece was originally published in March 2020, scientific consensus has shifted. Now some experts say the 'lab leak' theory warrants an investigation, along with the natural origin theory." A second note, dated May 25, 2021, said the April 2020 changes "were not mentioned at the time." The subheading now ends "And evidence from scientists to the contrary." As read on Sept. 30, 2026, the headline still says "debunked" and the caption still says "scientific evidence shows it's not."⁸¹

PolitiFact / Daniel Funke: "Pants on Fire," withdrawn

Public claim: on Sept. 16, 2020, PolitiFact rated "Pants on Fire" Li-Meng Yan's statement on Tucker Carlson Tonight: "This virus, COVID-19 SARS-CoV-2 virus, actually is not from nature. It is a man-made virus created in the lab." The headline called it a "debunked conspiracy theory that COVID-19 was created in a lab," and the rating said, "The claim is inaccurate and ridiculous."⁸²

What later records showed: Yan's claim was that the virus was engineered, which is a separate claim from a lab accident. In 2021 most U.S. intelligence agencies assessed "with low confidence that SARS-CoV-2 probably was not genetically engineered," and two said the evidence did not allow an assessment either way.⁶⁷ In 2025 the WHO advisory group said claims of intentional manipulation "are not supported by accurate science."⁷⁸

Correction status: withdrawn with a note. On May 17, 2021 PolitiFact wrote that its sources had "asserted the SARS-CoV-2 virus could not have been manipulated. That assertion is now more widely disputed," removed the rating from its database, and said, "Currently, we consider the claim to be unsupported by evidence and in dispute."⁸² That description matches the later record. The original "Pants on Fire" and "debunked" were more certain than the evidence allowed on engineering, where two agencies later declined to judge. On a lab accident, the claim Yan did not make, the record does not support them at all.

Where the origins question stands (September 2026)

No investigation has determined how SARS-CoV-2 first infected people. The two main hypotheses, a spillover from animals and an accident linked to research in Wuhan, are both still open in the official record.

U.S. intelligence is split, and no agency has reached high confidence. In August 2021 every agency assessed both hypotheses as plausible. Four agencies and the National Intelligence Council leaned natural at low confidence, one agency leaned lab at moderate confidence, and three could not decide. The intelligence community also judged that the virus "was not developed as a biological weapon."⁶⁷ The June 2023 declassified report named the Department of Energy and the FBI as the agencies favoring "a laboratory-associated incident ... although for different reasons."³⁸ CNN reported the Energy Department's view as a low-confidence assessment and the FBI's as moderate.⁸³ In January 2025 the CIA said it "assesses with low confidence that a research-related origin of the COVID-19 pandemic is more likely than a natural origin," and that "both research-related and natural origin scenarios of the COVID-19 pandemic remain plausible." The Associated Press reported that the finding was not based on new intelligence.⁶⁸ In June 2026 the Director of National Intelligence released internal documents on how the assessments were made. Her office's release says the Wuhan research is "now widely viewed as the source" of the pandemic. The first part of the documents is correspondence, not a new graded assessment.⁸⁴

In March 2025, Die Zeit and Süddeutsche Zeitung reported that Germany's foreign intelligence service had put the likelihood of a lab accident at 80 to 90 percent in 2020. That assessment was never published.⁸⁵

The World Health Organization's advisory group reported in June 2025 that "the weight of available evidence ... suggests zoonotic spillover." It said it lacked the information to assess a lab origin and was "not in a position to rule this out as a possibility." It said claims of intentional manipulation "are not supported by accurate science."⁷⁸

In December 2024 the Republican majority of the House Select Subcommittee on the Coronavirus Pandemic concluded that the virus "most likely emerged from a laboratory in Wuhan, China."⁸⁶ The Democratic members filed their own final report. The ranking member said the two sides were "leaving with different impressions of what we did or did not find."⁸⁷

Most peer-reviewed analyses favor a market-linked spillover: two 2022 papers in Science placed the early cases around the Huanan market, and a 2024 paper in Cell traced animal and viral genetic material there.⁸⁸ Fewer papers lean the other way. A 2024 risk analysis scored the evidence as pointing to "a greater likelihood of an unnatural than natural origin" and concluded that "the possibility of a laboratory origin cannot be easily dismissed."⁸⁹ The March 2020 Nature Medicine paper that much early coverage relied on said, "we do not believe that any type of laboratory-based scenario is plausible."⁹⁰ The House majority's report says that paper was "prompted by Dr. Fauci to push the preferred narrative."⁸⁶

In 2021 the intelligence community said it could not give a more definitive answer unless new information showed "the specific pathway for initial natural contact with an animal" or "that a laboratory in Wuhan was handling SARS-CoV-2 or a close progenitor virus before COVID-19 emerged."⁶⁷ The WHO group's 2025 report said it still lacked that information.⁷⁸

The best reporters had the best nouns

EDD. Bank of America. Kabbage. Fake farms. Borrower pipeline. Lender of record. Processor. Potential fraud. Likely fraud. Convicted fraud. False Claims Act settlement. EUA. Mandate. Aerosol. Ventilation. Hypothesis. Laboratory-associated incident. Bioweapon. FOIA. Standing. Agency speech. Platform policy.

The worst stories let the nouns blur. Lenders, fintechs, processors, and LSPs collapsed into one actor. Flags became fraud. Potential fraud became stolen money. A vendor's estimate became a public number. A staff report became a finding. An open hypothesis became a debunked conspiracy theory. The government-run EIDL program became "a separate" thing.

The correction failure is more damaging than the first-day mistake. Nobody covered these programs with perfect information in 2020, 2021, or 2022. By 2025 the record was better. DOJ had Kabbage's settlement and Blueacorn's convictions. FTC had Womply's advertising case. SBA had its suspension on the record and its Inspector General had the EIDL-PPP split. GAO had a UI fraud estimate. The House had asked ID.me for the $400 billion math. The Womply-Benworth arbitration had tested one lender's counterclaims. The Supreme Court had left the First Amendment merits in Murthy unresolved. ODNI had published a split intelligence community; the CIA later leaned toward a lab and the WHO's advisory group toward spillover, neither with confidence. School-cost evidence had made prolonged closure harder to defend.

The public account did not absorb those records at the volume of the accusations. The Times, Post, NPR, ProPublica, and political-press report-week pieces should add updates distinguishing Blueacorn's convictions, Kabbage's False Claims Act settlement, and Womply's record: the SBA suspension, the FTC advertising settlement, and the arbitration award.

Detailed methodology and sources

What was graded

One beat was the economic rescue: PPP, EIDL, unemployment insurance, stimulus checks, ERC, RRF, SVOG, provider relief, airline aid, Fed facilities, tax credits, loan forgiveness, fraud enforcement, and the argument over whether speed was a virtue or a design flaw. The other was the health response: testing, vaccines, EUA law, therapeutics, masks, school closures, mandates, origins, EcoHealth, public-health speech, platform moderation, and the courts that spent years trying to separate emergency policy from emergency power.

This scorecard grades the published record, not private motive. "Bias" here means a visible reporting pattern: source bias, selection bias, denominator bias, risk bias, access bias, or correction bias. A reporter can quote every sentence accurately and still leave readers with the wrong map.

The ranking is weighted toward the subjects this site covers: economic-relief programs, pandemic companies, FDA/EUA records, origins and FOIA, mandates, schools, and the speech cases. It grades the twenty reporters and teams whose work shaped the public record on those subjects.

Some pandemic policies were undercovered but do not have one bad byline to rank. RRF, SVOG, Main Street Lending, Fed 13(3) facilities, Provider Relief Fund, airline payroll support, PREP Act litigation, business-interruption insurance, and many mandate cases got trade-press, local, or legal coverage rather than sustained national accountability reporting.

How the grades work

Method: each entry rests on named pieces, listed in the notes, read against the later official record: inspector-general reports, GAO estimates, court dockets, agency statements and a congressional committee's own releases. The research cutoff is September 30, 2026.

Five things decide a grade. First, documented factual errors in the reviewed pieces, weighted by where they sat: a wrong number in a headline or opening paragraph counts more than one in a caption. Second, whether the outlet corrected them. Third, how the piece handled the source hierarchy: an estimate against an audit, a staff report against a court record, a lender against a processor. Fourth, whether the reporter or outlet updated when the record changed. Fifth, whether the piece claimed more certainty than the evidence allowed on the day it ran.

On origins the fifth test is judged by what was known at the time. A reporter who told readers in 2020 or 2021 that the lab-origin hypothesis was "debunked," "disproved" or a "conspiracy theory" overstated the record: no investigation had ruled it out then, and none has since. A reporter who presented a lab origin, or an engineered virus, as established fact overstated it too. The accident hypothesis and the bioweapon claim are graded separately. U.S. intelligence agencies judged in 2021 that the virus "was not developed as a biological weapon," so a piece that rejected the bioweapon claim is not marked down for it; a piece that used that rejection to dismiss a laboratory accident is.⁶⁷

The bands follow from that. A and A- go to reporters with no located error across a multi-year body of work and either independent recognition of that work or a documented call that agencies or auditors later adopted. B+ goes to reporters with no located error where the reviewed record is one to three pieces, or where the later record has not yet tested the central claims. B and B- cover corrected non-central errors, frames that needed an update and did not get one, and entries that rest partly on secondary sources. The C band is a piece whose caveats were inside but whose headline or frame outran them, with no correction. The D band is a headline or lead figure later found unsupported by an official inquiry and left uncorrected, or repeated large errors in the same direction.

A correction made counts for the reporter. A correction owed and not made counts against.

Two limits. For several reporters (Reinicke, Egan, Salmon, the NPR team) the reviewed record is one piece or one story cycle, and the entry says so. For the Twitter Files team, the underlying pieces could not be re-read for this revision, and that entry carries the earlier assessment with that caveat.

Notes

1. The Pulitzer Prizes, "Ed Yong of The Atlantic," 2021 Explanatory Reporting citation. Our copy. ↩
2. STAT, "Helen Branswell, STAT's Acclaimed Infectious Disease and Global Health Reporter, Wins George Polk Award For Public Service For Coverage of Covid-19 Pandemic," February 2021. Our copy. ↩
3. Stephen Council, "Freeze first, verify second: Unemployed Californians get a fright from EDD," CalMatters, Oct. 7, 2020. Our copy. ↩
4. Lauren Hepler and Stephen Council, "How Bank of America helped fuel California's unemployment meltdown," CalMatters, Nov. 20, 2020. Our copy. ↩
5. California State Auditor, High Risk Audit Report 2020-628.2 on the Employment Development Department's fraud prevention, Jan. 28, 2021. Our copy. ↩
6. UNC-Chapel Hill, "'Getting the big things right.'" Our copy. ↩
7. Zeynep Tufekci author archive, The Atlantic. Our copy. ↩
8. Zeynep Tufekci, "We Were Badly Misled About the Event That Changed Our Lives," The New York Times, March 2025 (syndicated by The New York Times Company, March 19, 2025). Our copy. ↩
9. Derek Willis and Lydia DePillis, "Hundreds of PPP Loans Went to Fake Farms in Absurd Places," ProPublica, May 18, 2021. Our copy. ↩
10. Derek Willis and Lydia DePillis, "An Online Lender Gave Hundreds of PPP Loans to Fake Farms. Now Congress Is Investigating," ProPublica, May 28, 2021. Our copy. ↩₁ ↩₂ ↩₃
11. U.S. Department of Justice, "Kabbage Agrees to Pay up to $120 Million to Resolve Allegations that it Defrauded the Paycheck Protection Program," May 13, 2024. Our copy. ↩₁ ↩₂ ↩₃
12. Carmen Reinicke, "Small business owners missed out on thousands of dollars in loans when PPP funding ran out early," CNBC, May 16, 2021. Our copy. ↩₁ ↩₂
13. Annie Nova and Greg Iacurci, "How Innovation Refunds cashed in on the employee retention credit," CNBC, Oct. 27, 2023. Our copy. ↩
14. Emily Flitter and Stacy Cowley, "Banks Gave Richest Clients 'Concierge Treatment' for Pandemic Aid," The New York Times, April 22, 2020. Our copy. ↩
15. Stacy Cowley and Ella Koeze, "How Two Start-Ups Reaped Billions in Fees on Small Business Relief Loans," The New York Times, June 27, 2021. Our copy. ↩₁ ↩₂
16. Stacy Cowley, "15% of Paycheck Protection Program loans could be fraudulent, study says," The New York Times, via The Seattle Times, Aug. 17, 2021. Our copy. ↩₁ ↩₂
17. Oto Analytics, Inc. d/b/a Womply v. Capital Plus Financial, LLC, N.D. Tex. No. 3:21-cv-02636: notice of removal, Oct. 25, 2021, ¶ 1 (state-court petition filed Sept. 9, 2021); electronic order of Nov. 14, 2022 (parties reported settlement); order of dismissal with prejudice, Jan. 5, 2023. Our copy. ↩
18. Stacy Cowley, "Two Fintechs Fueled Extensive Pandemic Relief Fraud, House Report Finds," The New York Times, Dec. 1, 2022. Our copy. ↩
19. Order to Show Cause for Expedited Temporary Injunctive Relief, same case, Oct. 7, 2021 (Judge Susan Stacy), ¶ 2. Our copy. ↩
20. Notice of Voluntary Dismissal (with prejudice), same case, Nov. 18, 2022. Our copy. Full docket: our copy. ↩
21. Complaint, Fountainhead SBF v. OTO Analytics, Inc. d/b/a Womply, Fla. 18th Cir. Ct. (Seminole County) No. 2021CA002143, Aug. 27, 2021, ¶ 7 n.1 and ¶ 19. Our copy. ↩
22. Affidavit of Alexander L. Cheney (Womply's counsel), same case, Sept. 13, 2021, ¶¶ 6–8. Our copy. ↩
23. House Select Subcommittee on the Coronavirus Crisis, staff report, "We Are Not the Fraud Police": How Fintechs Facilitated Fraud in the Paycheck Protection Program, December 2022. Our copy. ↩
24. Order No. 6, Order re: Motion to Admit Congressional Staff Report, Oto Analytics, Inc. d/b/a Womply v. Benworth Capital Partners, LLC, JAMS No. 1210038203, May 2, 2023 (Arbitrator Alexander L. Brainerd), filed as Exhibit 12, N.D. Cal. No. 4:24-cv-03975, Dkt. 41-14. Our copy. ↩
25. U.S. Department of Justice, "Co-Founder of Paycheck Protection Program Lender Service Provider Sentenced for $63M COVID-19 Relief Fraud Scheme," Nov. 21, 2025. Our copy. ↩₁ ↩₂ ↩₃
26. U.S. Department of Justice, "Co-Founder of Paycheck Protection Program Lender Service Provider Sentenced for $65M COVID-19 Relief Fraud Scheme," Dec. 18, 2025. Our copy. ↩₁ ↩₂
27. Corrected Final Award, Oto Analytics, Inc. d/b/a Womply v. Benworth Capital Partners, LLC, JAMS No. 1210038203, June 26, 2024, §VII. Our copy. ↩₁ ↩₂ ↩₃
28. Zachary Warmbrodt, "Banks warn of chaotic launch of small business lending program," Politico, April 2, 2020 (our copy); "Bankers plead with Trump: Fix small business rescue plan," April 7, 2020 (our copy); "Banks brace for new wave of small businesses at troubled loan program," April 9, 2020 (our copy). ↩
29. Zachary Warmbrodt, "Wall Street titans lobby for a cut of small business bailout," Politico, April 1, 2020, as syndicated by Yahoo News. Our copy. ↩
30. Zachary Warmbrodt, "Banks warn of chaotic launch of small business lending program," Politico, April 2, 2020 (our copy); "Massive small business rescue gets off to stormy start," April 3, 2020 (our copy); "Justice Department launches probe of small business loans," April 30, 2020 (our copy). ↩
31. Zachary Warmbrodt, "Watchdog calls for fraud crackdown at small business lending program," Politico, June 25, 2020 (our copy); "Watchdog sees massive potential loan fraud at SBA, sparking fight with agency," Oct. 28, 2020 (our copy). ↩
32. SBA Office of Inspector General, Report 22-22, "Follow-up Inspection of SBA's Internal Controls to Prevent COVID-19 EIDLs to Ineligible Applicants," Sept. 29, 2022. Our copy. ↩
33. U.S. Government Accountability Office, "COVID-19: Opportunities to Improve Federal Response and Recovery Efforts," GAO-20-625, June 2020. Our copy. ↩₁ ↩₂
34. Katherine Eban and Jeff Kao, "COVID-19 Origins: Investigating a 'Complex and Grave Situation' Inside a Wuhan Lab," Vanity Fair / ProPublica, Oct. 28, 2022, with the editor's note appended. Our copy. ↩₁ ↩₂
35. Katherine Eban, "The Lab-Leak Theory: Inside the Fight to Uncover COVID-19's Origins," Vanity Fair, June 3, 2021. Our copy. ↩
36. Office of the Director of National Intelligence, "Report on Potential Links Between the Wuhan Institute of Virology and the Origins of COVID-19," June 23, 2023. Our copy. ↩₁ ↩₂ ↩₃
37. Alexander Nazaryan, "Who's to Blame for the Catastrophe of COVID School Closures?," New York, May 4, 2025. Our copy. ↩
38. The 74, "Journalist David Zweig Calls COVID School Closures 'A False Story About Medical Consensus,'" April 17, 2025. Our copy. ↩
39. Ken Schwencke, "Fintechs Made 'Massive Profits' on PPP Loans and Sometimes Engaged in Fraud, House Committee Report Finds," ProPublica, Dec. 1, 2022. Our copy. ↩
40. Meghan Bobrowsky and Ben Wieder, "Quickie lender Kabbage doled out billions in PPP loans. A number of borrowers raised red flags," Miami Herald, Sept. 10, 2020. Our copy. ↩
41. Kevin G. Hall, Ben Wieder and Nicholas Nehamas, "Big banks generated billions in fees from Paycheck Protection Program," Miami Herald / McClatchy, December 2020. Our copy. ↩
42. Ben Wieder, Shirsho Dasgupta, Christina Saint Louis and David Szakonyi, "A felon, an alleged drug dealer and a comic: How some who may be ineligible got PPP loans," Miami Herald and Anti-Corruption Data Collective, Dec. 16, 2020. Our copy. ↩
43. Ben Wieder, "PPP loans earned a Miami lender $680 million. Now it's being sued by the Federal Reserve," Miami Herald, via AOL, July 24, 2024. Our copy. ↩₁ ↩₂ ↩₃
44. Ben Wieder, "Awaiting ambassador confirmation, Bernie Navarro settles with Federal Reserve bank," Miami Herald, Sept. 12, 2025, as syndicated by InsuranceNewsNet (our copy) and AOL (our copy). ↩₁ ↩₂ ↩₃ ↩₄
45. Ben Wieder, "Miami lender close to Rubio got senator's support as it fought over PPP loan spoils," Miami Herald, Aug. 6, 2024, via AOL (our copy); the correction appears on the Herald's own page (our copy). ↩₁ ↩₂ ↩₃ ↩₄
46. Ben Wieder, "Top SBA official oversaw PPP, then did work for one of its worst offenders," Miami Herald, May 4, 2023. Our copy. ↩₁ ↩₂
47. Arbitrator's Order re: Motion to Admit Congressional Staff Report, Oto Analytics, Inc. d/b/a Womply v. Benworth Capital Partners, LLC, JAMS No. 1210038203, May 2, 2023, filed as N.D. Cal. No. 4:24-cv-03975, Dkt. 1-13. Our copy. ↩
48. House Select Subcommittee on the Coronavirus Crisis, "Select Subcommittee Expands Investigation into Role of FinTech Industry in PPP Fraud," Nov. 23, 2021. Our copy. ↩₁ ↩₂
49. Murthy v. Missouri, No. 23-411 (U.S. June 26, 2024). Our copy. ↩
50. Richard Lardner, Jennifer McDermott and Aaron Kessler, "The Great Grift: How billions in COVID-19 relief aid was stolen or wasted," Associated Press, June 12, 2023. Our copy. ↩₁ ↩₂
51. SBA Office of Inspector General, "COVID-19 Pandemic EIDL and PPP Loan Fraud Landscape," Report 23-09, June 27, 2023. Our copy. ↩₁ ↩₂
52. Austin Fast and Sacha Pfeiffer, "Fintechs fueled widespread PPP loan fraud, according to a report from Congress," NPR, Dec. 6, 2022, with the Dec. 12, 2022 correction. Our copy (archived June 25, 2026). ↩₁ ↩₂ ↩₃
53. Anna Hrushka, "PayPal, Square, Intuit approved to participate in coronavirus relief program," Banking Dive, April 13, 2020 (our copy); House Select Subcommittee on the Coronavirus Crisis, "We Are Not the Fraud Police": How Fintechs Facilitated Fraud in the Paycheck Protection Program, Dec. 1, 2022, pp. 16-17. Our copy. ↩
54. Fatima Hussein, "House panel says lax screening helped facilitate PPP fraud," Associated Press, via The Business Journal, Dec. 1, 2022. Our copy. ↩
55. Tony Romm, "'The more you submit, the more we get paid': How fintech fueled covid aid fraud," The Washington Post, Dec. 1, 2022. Our copy. ↩₁ ↩₂
56. Ken Dilanian and Laura Strickler, "'Biggest fraud in a generation': The looting of the Covid relief plan known as PPP," NBC News, March 28, 2022. Our copy. ↩₁ ↩₂ ↩₃
57. U.S. Government Accountability Office, "Unemployment Insurance: Estimated Amount of Fraud During Pandemic Likely Between $100 Billion and $135 Billion," GAO-23-106696, Sept. 12, 2023. Our copy. ↩₁ ↩₂ ↩₃
58. Ken Dilanian, Kit Ramgopal and Chloe Atkins, "'Easy money': How international scam artists pulled off an epic theft of Covid benefits," NBC News, August 2021. Our copy. ↩
59. House Committee on Oversight and Reform and Select Subcommittee on the Coronavirus Crisis, "Chairs Maloney, Clyburn Release Evidence Facial Recognition Company ID.me Downplayed Excessive Wait Times for Americans Seeking Unemployment Relief Funds," Nov. 17, 2022. Our copy. ↩₁ ↩₂
60. Matt Egan, "Fintechs have a fraud problem with their Covid relief loans," CNN Business, Aug. 17, 2021. Our copy. ↩₁ ↩₂
61. Apoorva Mandavilli, "A new vaccine strategy for children: Just one dose, for now," The New York Times, via The Seattle Times, Oct. 6, 2021 (syndicated copy carrying the original figure). Our copy. ↩
62. Fox News, "New York Times issues massive correction after overstating COVID hospitalizations among children," October 2021, reproducing the Times correction. Our copy. ↩
63. Alec Schemmel, "New York Times again misstates COVID-related deaths in children, issues correction," The National Desk / ABC 33/40, May 20, 2022, reproducing the Times correction. Our copy. ↩
64. Lindsey Ellefson, "NY Times COVID Reporter Deletes Tweet Claiming 'Racist Roots' of 'Lab Leak Theory' After Backlash," TheWrap, May 27, 2021. Our copy. ↩
65. Office of the Director of National Intelligence, unclassified summary of the assessment on COVID-19 origins, Aug. 27, 2021 (our copy); declassified "Updated Assessment on COVID-19 Origins," Oct. 29, 2021 (our copy). ↩₁ ↩₂ ↩₃ ↩₄ ↩₅ ↩₆ ↩₇ ↩₈ ↩₉ ↩₁₀
66. CBS News / Associated Press, "CIA now says COVID most likely originated from a lab leak but has 'low confidence' in its assessment," Jan. 25, 2025 (updated Jan. 27, 2025), quoting the CIA statement. Our copy. ↩₁ ↩₂ ↩₃
67. Felix Salmon, "Half of the pandemic's unemployment money may have been stolen," Axios, June 10, 2021. Our copy. ↩₁ ↩₂
68. Eamon Javers and Scott Zamost, "Criminals have stolen nearly $100 billion in Covid relief funds, Secret Service says," CNBC, Dec. 21, 2021. Our copy. ↩₁ ↩₂
69. Jennifer McDermott, "Nearly $100 billion stolen in pandemic relief funds, Secret Service says," Associated Press, via PBS NewsHour, Dec. 22, 2021. Our copy. ↩
70. Paulina Firozi, "Tom Cotton keeps repeating a coronavirus conspiracy theory that was already debunked," The Washington Post, Feb. 17, 2020 (headline as first published). Our copy. ↩
71. The Post's editor's note and revised headline as reproduced in Caroline Downey, "Washington Post Corrects Year-Old Article Calling Lab-Leak Theory 'Debunked,'" National Review via Yahoo News, June 2, 2021 (our copy); Frank E. Lockwood, "Cotton's theory on virus origin no longer dismissed," Arkansas Democrat-Gazette, May 30, 2021 (our copy). ↩
72. Alexandra Stevenson, "Senator Tom Cotton Repeats Fringe Theory of Coronavirus Origins," The New York Times, Feb. 17, 2020, updated Feb. 18, 2020. Our copy. ↩
73. Geoff Brumfiel, interviewed by Ailsa Chang, "Scientists Debunk Lab Accident Theory Of Pandemic Emergence," All Things Considered, NPR, April 22, 2020. Our copy. ↩
74. Geoff Brumfiel and Emily Kwong, "Virus Researchers Cast Doubt On Theory Of Coronavirus Lab Accident," NPR, April 23, 2020. Our copy. ↩
75. Nicholson Baker, "The Lab-Leak Hypothesis," New York, Jan. 4, 2021. Our copy. ↩
76. World Health Organization, Scientific Advisory Group for the Origins of Novel Pathogens (SAGO), "Independent assessment of the origins of SARS-CoV-2," June 27, 2025. Our copy. ↩₁ ↩₂ ↩₃ ↩₄
77. Chris Cillizza, "Anthony Fauci just crushed Donald Trump's theory on the origins of the coronavirus," CNN, May 5, 2020. Our copy. ↩
78. Chris Cillizza, "CHRIS CRUCIAL: I screwed up," So What (newsletter), Jan. 28, 2025. Our copy. ↩
79. Eliza Barclay, "The conspiracy theories about the origins of the coronavirus, debunked," Vox, March 4, 2020, with editor's notes dated May 24 and May 25, 2021. Our copy; our copy of the Wayback Machine capture of March 10, 2020. ↩₁ ↩₂
80. Daniel Funke, "Tucker Carlson guest airs debunked conspiracy theory that COVID-19 was created in a lab," PolitiFact, Sept. 16, 2020, with editor's note of May 17, 2021. Our copy. ↩₁ ↩₂
81. Jeremy Herb, Natasha Bertrand, Zachary Cohen and Pamela Brown, "Assessment Covid-19 leaked from Chinese lab is a minority view within US intel community, sources say," CNN, Feb. 27, 2023. Our copy. ↩
82. Office of the Director of National Intelligence, News Release No. 11-26, "Fauci Funded Wuhan Lab Research That Sparked COVID," June 18, 2026 (our copy), and the released documents (Part 1) (our copy). ↩
83. Reuters, "German spy agency concluded COVID virus likely leaked from lab, papers say," via U.S. News & World Report, March 12, 2025, reporting on Die Zeit and Süddeutsche Zeitung. Our copy. ↩
84. House Select Subcommittee on the Coronavirus Pandemic, "FINAL REPORT: COVID Select Concludes 2-Year Investigation, Issues 500-Page Final Report on Lessons Learned and the Path Forward," Dec. 2, 2024 (majority release; report "After Action Review of the COVID-19 Pandemic: The Lessons Learned and a Path Forward"). Our copy. ↩₁ ↩₂
85. Rep. Raul Ruiz, ranking member, statement at the Select Subcommittee's final report markup, Dec. 4, 2024. Our copy. ↩
86. Michael Worobey et al., "The Huanan Seafood Wholesale Market in Wuhan was the early epicenter of the COVID-19 pandemic," Science, Aug. 26, 2022, doi:10.1126/science.abp8715 (our copy); Jonathan E. Pekar et al., "The molecular epidemiology of multiple zoonotic origins of SARS-CoV-2," Science, Aug. 26, 2022, doi:10.1126/science.abp8337 (our copy); Alexander Crits-Christoph et al., "Genetic tracing of market wildlife and viruses at the epicenter of the COVID-19 pandemic," Cell, September 2024, doi:10.1016/j.cell.2024.08.010 (our copy); Jonathan E. Pekar et al., "The recency and geographical origins of the bat viruses ancestral to SARS-CoV and SARS-CoV-2," Cell, June 2025, doi:10.1016/j.cell.2025.03.035 (our copy). ↩
87. Xin Chen et al., "Use of a risk assessment tool to determine the origin of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)," Risk Analysis, published March 15, 2024, doi:10.1111/risa.14291 (our copy); Jesse D. Bloom, "Recovery of Deleted Deep Sequencing Data Sheds More Light on the Early Wuhan SARS-CoV-2 Epidemic," Molecular Biology and Evolution, Aug. 16, 2021, doi:10.1093/molbev/msab246 (our copy). ↩
88. Kristian G. Andersen, Andrew Rambaut, W. Ian Lipkin, Edward C. Holmes and Robert F. Garry, "The proximal origin of SARS-CoV-2," Nature Medicine 26, 450-452 (March 17, 2020), doi:10.1038/s41591-020-0820-9. Our copy. ↩