The White House now has a scoreboard for fraud, and the top line is a headline writer's dream: $260.7 billion in fraud uncovered since January 2025, $72 billion a year stopped, $58.9 billion enforced. The Fraud Ledger at whitehouse.gov/fraud, updated September 24, presents all three figures in matching oversized type, as if they measure the same thing with the same confidence. They do not. A New York Times fact check published September 27 found the headline claim rests on questionable accounting, and you do not have to take the Times' word for it. The ledger's own fine print makes the case.
Start with what each column actually is. Fraud uncovered, the big $260.7 billion, is labeled right on the site as estimated. It represents what agencies think they have identified through data analysis, reported to the task force by the agencies themselves. Fraud stopped is annualized, meaning a projection of what administrative actions might save per year, not money already saved. Fraud enforced is labeled indictments and settlements, which means it counts what prosecutors allege, not what the government has actually clawed back. Three different species of number, one species of presentation.
Georgetown University's Center for Children and Families published a detailed fact check of the ledger on September 21 and reached a blunt conclusion: at best unhelpful and, on Medicaid, fundamentally misleading. Consider the ledger's number one highest-impact action, Medicaid fraud deferrals of over $2 billion from California and over $500 million from Minnesota. Those are federal matching funds CMS is withholding while it decides whether the states' claims are allowable. They are not documented fraud. They are money being held during a review, booked on the scoreboard as though the review already came back guilty. As former state health secretaries from Indiana and North Carolina put it, withholding Medicaid funding does not identify a fraudulent provider or prove a false claim. It simply disrupts the flow of dollars that pay for health care.
Or take the Medicaid fraud war room, which the ledger credits with saving $200 million in its first 90 days. CMS's own July 28 press release described that money as potentially improper payments. Improper and fraudulent are different words for a reason. An improper payment can be a paperwork error. Fraud requires intent and proof. The ledger lets a reader walk away believing the entire sum was criminal.
The pattern holds down the list. Of eleven Medicaid-related items in the White House's August task force report, Georgetown found that four, totaling $2.8 billion, were withheld matching funds rather than proven fraud, five, totaling $676 million, were unproven allegations in pending charges, and only two, totaling $277 million, were actual convictions. The Small Business Administration's eye-popping $122.9 billion in uncovered fraud, the largest single agency figure, is largely pandemic-era loan fraud that occurred in 2020 and 2021 and was first quantified by the SBA's own inspector general back in 2023, whose estimate of roughly $200 billion in potential PPP and EIDL fraud the SBA itself disputed at the time, arguing the likely figure was closer to $36 billion. The fraud was real and enormous. It also happened years before this task force existed, under programs designed for speed over scrutiny, and crediting it to a 2025 initiative is a generous reading of the word uncovered.
Even the hardest column softens under inspection. The ledger counts the SBA's referral of 562,000 borrowers and $22.2 billion in pandemic loans to Treasury for collections. A referral is a request, not a recovery. It counts the DOJ charging 109 defendants with $5.4 billion in intended loss. Intended loss is what a scheme aimed at, not what it took, and a charge is an accusation awaiting its day in court. The ledger's own topline admits it has documented just 25 actions behind all those billions.
The gap between the podium and the paperwork got wider once the numbers left the website. Social media posts promoting the ledger converted the $58.9 billion enforced figure into $58.9 billion recovered, a word the ledger itself does not use. Recovered implies money back in the Treasury. Nothing on the site supports that framing.
None of this means the underlying mission is fake. Pandemic relief fraud was a genuine catastrophe by any estimate. The hospice and home health crackdown in Los Angeles, where CMS eliminated 800 hospice and home health providers and 900 durable medical equipment suppliers it describes as fraudulent, targets a corner of Medicare that investigators have flagged for years. DOJ reports 152 guilty pleas or convictions and 116 defendants sentenced to a collective 4,862 months in prison. Real fraud, real prosecutions, real prison time. An honest dashboard tracking that work would be genuinely useful, which is exactly what Georgetown's reviewers said while calling this one misleading.
The problem is what the inflated total is doing out in the world. Fraud is the stated reason HUD suspended $220 million a year in payments to the Los Angeles Homeless Services Authority pending the outcome of an investigation, and froze payments to the Virgin Islands Housing Finance Authority, which holds $1.9 billion in disaster relief funds, also pending investigation. It is the stated reason Minnesota lost a $260 million Medicaid payment pause in February and California faces $2 billion in deferred federal funds. In each case, the suspension comes first and the proof comes later, and in each case, the suspended money shows up on the ledger as an anti-fraud accomplishment. The accusation generates the savings, and the savings inflate the accusation's credibility. That is a tidy loop.
There is an incentive structure here worth naming. The task force is chaired by the vice president, the figures are self-reported by the agencies the task force oversees, Georgetown found no published methodology explaining how HHS arrived at its numbers, and the totals serve a political purpose: they are the public arithmetic behind program suspensions, payment freezes, and staffing cuts. When the referee keeps the score, supplies the numbers, and benefits from a high score, the scoreboard deserves extra scrutiny, not a press release.
The pandemic taught a hard lesson about what happens when the government moves money faster than it verifies it. The risk now is the mirror-image failure: a government so eager to post fraud numbers that it treats suspects as convicts, estimates as recoveries, and withheld benefits as savings, while the people those benefits were meant for wait on an investigation that may or may not ever prove the case. Real fraud fighters need credible numbers most of all. When the scoreboard inflates, every legitimate conviction on it gets discounted too.
Free Game Takeaway
Learn to read any government or corporate scoreboard by asking one question: which number on this page required proof, and which required only an estimate? On this ledger, the proof column is the small one. If you live in California or Minnesota and use Medicaid, or rely on homeless services in Los Angeles or disaster recovery in the Virgin Islands, the deferrals and suspensions justified by these figures can affect your benefits before any fraud is proven, so watch whether withheld funds are eventually released or converted into documented findings. Watch three signals going forward: whether the White House ever publishes a methodology for the $260.7 billion estimate, whether the $22.2 billion in SBA loans referred to Treasury actually gets collected (referral is not recovery), and whether the charged amounts in pending cases survive to conviction, since Georgetown found only $277 million of the August report's Medicaid items involved actual convictions. Anytime an official says fraud recovered, check whether the source document says estimated, charged, or intended. Those words are where the truth lives.