The Government Declared War on Fraud... But They Don't Know What Fraud Even Means

Natalie Bulger
2 days ago
17 min read

I guess the dictionaries, Google searches and AI were all unavailable when the EO was written.

On March 16, 2026, President Trump signed an executive order establishing the Task Force to Eliminate Fraud — a whole-of-government mobilization, chaired by Vice President JD Vance, targeting federal benefit programs. Housing. Food. Medical care. Cash assistance. Title sounds great right? We all want less fraud. But wait, housing, food, medical care?

The White House fact sheet laid out the stakes: the government’s “vast benefits system” was being exploited. States were embracing “loopholes.” People were self-certifying their eligibility. The error rate on food stamps was nearly 9%! There was a Somali immigrant ring! There was a childcare scam! So. Much. Shock. And. Awe.

To be clear: some of those things are real. Fraud is real. Fraud is bad. Nobody — certainly not a former federal compliance officer — is arguing that the government should throw money into the void and shrug.

But here’s the thing about a “war on fraud”: the enemy you choose to fight says everything about who you actually want to protect — and who you’re willing to sacrifice.

So let’s look at the numbers. The real ones. From boring, government-issued, credentialed sources that absolutely nobody in the White House seems to have read. All of which are cited at the end of this article. Numbers that the analysts, compliance professionals, and accountants have dedicated lives to tabulating and understanding only to be told, go back to your basement office and be quiet - we got this.

First, Let's Define What We're Talking About

Before we can fight fraud, we have to know what it is. And right now, “fraud” is being used as a rhetorical all-purpose seasoning — sprinkled over any number that sounds big enough to be alarming. This is dangerous and will take years to correct in the public, if we’re even able to.

Here’s the actual legal definition: fraud is intentional deception or misrepresentation for unauthorized gain. The keyword is intentional. It’s not a mistake. It’s not a paperwork gap. It’s not a state agency that forgot to check a box. It’s deliberate lying to obtain money you aren’t entitled to. There are requirements that government agencies and leaders be trained on what fraud is, clearly, a lot of people skipped that orientation day.

There are three broad categories of government fraud that actually matter for this conversation:

1. Benefit fraud (recipient fraud): A program enrollee lies about their eligibility, income, or household to receive benefits they don’t qualify for. This is what the EO is focused on.

2. Provider/billing fraud: A doctor, hospital, nursing home, pharmacy, or other vendor bills the government for services they didn’t provide, inflates codes, gives kickbacks, prescribes unnecessary treatments, or otherwise games reimbursement systems. This is the single largest source of healthcare fraud in the United States.

3. Contract and procurement fraud: A government contractor misrepresents their capabilities, charges for work not done, substitutes inferior materials, inflates costs, or engages in bid-rigging. This category includes defense contractors, IT vendors, construction companies, and consultants billing the federal government — among others — for things that either don’t exist or didn’t perform as advertised.

Now let’s look at what each of these categories actually costs. Full disclosure - AI was not too busy for me this weekend, and we worked together to pull and validate the numbers and statistics that follow.

What the Data Actually Shows

Benefit Fraud: The Villain in the EO

The White House fact sheet cites a roughly 9% “error rate” for SNAP (food stamps) as evidence of rampant fraud. The actual FY 2024 national payment error rate is 10.93%, representing roughly $10+ billion in improper payments across $90+ billion in outlays.

Sounds alarming. Until you read the footnotes.

USDA itself states that the National Payment Error Rate is not a measure of fraud. The agency says these errors are “largely unintentional” — a mix of recipient errors, state agency errors, and eligibility miscalculations. In FY 2021, the last year with a detailed breakdown, confirmed recipient trafficking and application fraud accounted for $54 million in overpayments out of a nearly $80 billion program. State agency errors accounted for $85 million. Recipient errors (not fraud — mistakes) accounted for $352 million.

In other words, the White House is describing an error rate as if it’s a crime rate. These are not the same thing. Calling an administrative miscalculation “fraud” is like calling a wrong answer on your tax return a federal offense. Sometimes people get things wrong. Sometimes states fail to verify paperwork properly. That is a program administration problem.

It is not an army of con artists gaming the welfare system.

For Medicaid and Medicare, the picture is even more instructive. In FY 2024, Medicare reported $54.3 billion in improper payments, and Medicaid reported $31.1 billion — a combined $85+ billion that sounds catastrophic until you learn that CMS says approximately 79% of those improper payments resulted from insufficient or missing documentation, not fraud. The payments went to the right providers for the right services — the paperwork just didn’t meet regulatory standards. Technically improper. Not theft. This is what is defined as waste due to rework, and is what most risk, compliance and financial performance groups would have the biggest bang for their buck with if it were the focus.

And here’s the kicker: when Medicaid fraud is actually fraud, it is almost never committed by the person holding the benefits card.

According to Georgetown University’s Center for Children and Families, beneficiary fraud represents approximately 2% of reported convictions in Medicaid fraud cases, and just one-tenth of one percent of financial recoveries. The money is being stolen by providers — doctors, labs, pharmacies, nursing homes, and durable medical equipment (DME) suppliers (think wheelchairs, oxygen tanks, back braces) — billing for services never rendered, inflating codes, running kickback schemes, and prescribing unnecessary treatments.

For Social Security disability (SSI/SSDI), the SSA’s own Office of Inspector General has stated that “most improper payments we detect do not involve any evidence of intent to commit fraud.” Improper payments exist — the SSI error rate runs around 10.62% — but they largely result from beneficiaries failing to report changes in circumstances (a job, a marriage, a death), not from organized schemes to steal government money.

For VA benefits, the story being pitched is particularly deceiving — and honestly, embarrassing for the administration’s framing.

The administration has referenced a VA OIG finding that 69% of sampled claims showed “one or more indicators of fraud risk.” That number is real. What’s less real is calling it a fraud rate. By that logic, a veteran who filled out paperwork slightly incorrectly is a fraudster. Those are indicators — flags for further review — not confirmed fraud, and VA officials themselves have pushed back on that characterization.

What the OIG has actually confirmed about VA fraud? Most of it is committed by the providers and the vendors, not the veterans.

The Veterans Community Care Program — the VA’s external and contracted care network allowing veterans to see private-sector providers — has become a documented magnet for provider billing fraud. According to VA OIG data, 38% of the 218,000 providers in the VA’s Community Care Network engaged in “upcoding” — the fraudulent practice of assigning higher billing codes to procedures than what was actually performed in order to inflate reimbursement. This is not a rounding error. More than a third of participating providers were potentially bilking the program for veterans’ care.

The VA process, instituted after providers began refusing to see veteran patients due to long payment lags, is to pay providers before proof of treatment is even received — then, if deemed necessary, chase recoupments after the fact. A third-party administrator layer sits between the VA and those providers rake in millions while adding administrative density and opacity to an already broken system - all it takes is a look at the current Sam.Gov contract listing for the services. With all the technology innovation of recent years, no one has seriously pushed the question of whether we now have the capability to pay only once services are validated, like every other payer in the market.

In FY 2020 alone, 45,600 community care providers received $37.8 million for evaluation and management services that should have been bundled into their surgery fees — meaning surgeons charged separately for activities that were already included in their payments.

And then the VA’s own fraud detection capability. A capability that was split apart into different components and placed within a variety of offices and functions that rarely talked to each other. When progress was made in the last two years to consolidate these efforts, the administration change left it understaffed, underfunded, slashing support contracts and freezing critical hiring. The tool used for detection was determined to be defective and required a rebuild, a rebuild that was tasked to another department and hit barrier after barrier.

Since the start of FY 2024 alone, the VA OIG has issued 21 separate reports documenting challenges in administering community care.

The VA’s fraud problem is not veterans lying about service-connected injuries. It is an exploding outsourced provider network with documented, systemic billing fraud — and a fraud detection tool the agency broke, then couldn’t fix for over a year. The EO doesn’t mention any of this. It mentions benefit recipients.

Contract Fraud: The Villain Nobody's Naming

So where is the real story? Well it's in the contracts - those giant boring documents that hide the real fraud risks.

The Department of Defense — which accounts for roughly 71% of all federal contracting dollars, or about $431 billion per year — reported approximately $10.8 billion in confirmed fraud just between 2017 and 2024. Defense procurement fraud ranges from product substitution (selling the military inferior parts) to cybersecurity fraud to inflated contracts to outright billing for work that was never done.

The Small Business Administration’s Office of Inspector General found that roughly $200 billion in COVID-era Paycheck Protection Program and Economic Injury Disaster Loans were potentially fraudulent. Two hundred. Billion. Dollars. Stolen primarily by businesses and organized crime rings — not by families trying to pay rent.

Under the False Claims Act — the government’s primary tool for recovering procurement and healthcare fraud — the DOJ recovered $2.68 billion in FY 2023 and $2.9 billion in FY 2024. Healthcare provider fraud was the single largest category in both years, at approximately $1.8 billion in 2023 alone. Procurement fraud settlements have reached record levels in recent years, with one single settlement in 2023 totaling $377 million.

And this is just what gets caught.

The Government Accountability Office (GAO) — the nonpartisan research arm of Congress — estimated in a landmark 2024 report that the federal government loses between $233 billion and $521 billion annually to fraud across all programs and contract types. That is a range because fraud, almost by definition, is hard to measure precisely. But even at the low end, that is a staggering number, and benefits fraud for individual recipients is a small fraction of it.

The Tax Gap: The Fraud Nobody's Talking About

Now to the thing I learned while working on this article — the tax gap.

The IRS estimated a gross tax gap of $696 billion in tax year 2022. After enforcement and late payments, the net tax gap is approximately $606 billion annually. This is money that was legally owed to the federal government — by individuals and businesses — and simply not paid.

Nearly four-fifths of the tax gap comes from underreporting. Nearly three-fifths comes from business and self-employment income — shell corporations, other business structures, and high-earning individuals who have the resources to underreport because they don’t have wages automatically withheld.

The task force established by the March 2026 EO does not include the IRS. It does not target the tax gap. It does not mention corporate underreporting, contractor overbilling, or defense procurement. It targets housing, food, medical care, and cash assistance.

To put this in perspective: if the IRS collected just one month’s worth of the annual tax gap, it would recover more money than the entire SNAP program’s improper payments for a decade.

What the EO Actually Says (And What It Doesn't)

Read the actual text of the executive order carefully, and the framing goes well beyond fraud prevention.

Section 1 of the EO states that “illegal aliens, criminals, foreign gangs, bureaucrats, State and local officials, non-governmental organizations, and ineligible providers exploit these programs.” It then pivots into a passage about how “self-dealing political actors use such public benefits programs to solidify control over their communities and our political systems,” and alleges that public officials use benefits to attract migrant voters, who then vote illegally to keep those officials in power. Verbatim: “Due to insufficient election integrity measures, some migrants who are not eligible to vote do so anyway, with the same public officials permitting widespread ballot harvesting schemes.”

This is a document ostensibly about benefit fraud. It contains a paragraph about ballot harvesting. In case that wasn’t clear enough about the intended scope of this exercise, the fact sheet includes a direct Trump quote: “The first duty of the American government is to protect American citizens, not illegal aliens.” That’s the stated justification for a task force to audit food stamps.

The operational requirements are tiered: member agencies have 30 days to identify their most fraud-susceptible benefit transactions; the Task Force has 60 days to coordinate minimum anti-fraud requirements; each agency has 90 days to submit a measurable implementation plan. The required anti-fraud measures include identity verification, documentation requirements for services rendered, pre-payment controls, and expanded data sharing — none of which are inherently unreasonable as fraud prevention tools.

But Section 3(a)(ii) authorizes agencies to “determine when ongoing fraud or potential fraud require proactively pausing certain types of funding until such controls can be established.” The Fact Sheet confirms this has already been used — the administration “temporarily paused certain Medicaid funding” to Minnesota while investigations were ongoing. That is not a minor administrative adjustment. That is a mechanism to suspend healthcare funding to a state — and by extension, to the people who depend on it — based on the existence of suspected (not confirmed) fraud.

Section 6 directs the Attorney General to promote False Claims Act suits — but only for “fraud within Federal benefit programs.” The False Claims Act is the government’s primary tool for recovering contractor and provider fraud. This section could have directed FCA enforcement broadly. It was specifically scoped to benefits.

The EO’s evidentiary justification rests almost entirely on Minnesota: Medicaid fraud that “could total billions” (a projection, not a confirmed figure), the Feeding Our Future childcare scam, and “an organized ring of Somali immigrants“ who stole childcare funding and allegedly sent money to a terror group. These cases are real, prosecuted, and reprehensible. They are also the rhetorical foundation for a national task force aimed at every person who receives housing, food, medical care, or cash assistance in America. One state. Two programs. One ethnic group called out by name in the order itself. The implied scope of the threat is not supported by the evidence base provided.

Accountability Theater: The Paper Trail the EO Ignores — And Just Weakened

Here's where the story gets truly remarkable, and requires you to hold two dates in your head simultaneously.

March 10, 2026: MB Director Russell Vought signs a revised OMB Circular A-123 — Management’s Responsibility for Internal Control — the federal government’s foundational framework requiring agencies to assess and certify their internal controls annually.

March 16, 2026: resident Trump signs the Executive Order establishing the Task Force to Eliminate Fraud, blaming the “previous administration” for “adopting policies that weakened the Federal Government’s oversight.”

Six days apart. Read both documents together and a very specific contradiction emerges.

Under A-123 — which has been law since the Federal Managers’ Financial Integrity Act of 1982 (FMFIA) — the head of each executive branch agency is required to annually prepare and sign a Statement of Assurance certifying whether the agency’s internal controls comply with FMFIA. The 2026 A-123 revision lays out three possible statements an agency head can make:

Unmodified (clean): Internal controls "were operating effectively" — full stop.
Modified: Effective, "except for the following material weaknesses reported."
Statement of No Assurance: The agency "is unable to provide assurance that internal control over operations, reporting, and compliance was operating effectively due to the following material weaknesses."

In plain terms — a no-assurance statement is an agency saying out loud, on the record, to Congress: we cannot tell you the money is being protected. It almost never happens. Nobody wants to be the one who signs it.The circular defines a material weakness as a deficiency “significant or severe enough to report outside of the agency” — including any condition that “significantly weakens established safeguards against fraud, waste, loss, unauthorized use, or misappropriation of funds.” Material weaknesses must be reported to OMB and Congress through annual Agency Financial Reports. We’ll talk a lot more about this in Part 3 of this series.

This framework, properly implemented, is exactly the accountability mechanism that should be catching chronic fraud exposure. Agency heads sign their names to it. It goes to Congress. It’s public.

So why doesn’t it work?

GAO has kept Medicare on its High-Risk List since 1990 — 35 years, through nine presidents. CMS had over 100 unimplemented GAO recommendations as of 2024, including 15 specifically on improper payments. Medicare’s ratings for action plan, monitoring, and demonstrated progress on GAO’s high-risk criteria remain only partially met. Medicaid has been on the High-Risk List since 2003. Both programs have been submitting annual Agency Financial Reports with Assurance Statements throughout this entire period.

Which brings us back to March 10, 2026 — and the cover letter Russell Vought attached to the new A-123.

In it, Vought writes that prior versions of Circular A-123 “have overly deferred to direction and priorities of external entities whose views are not binding on the Executive Branch such as the Government Accountability Office. Doing so has failed to prioritize agency internal control processes to adequately protect American taxpayer dollars, leading to documented examples of widespread abuse.”

Read that twice. The same administration that, six days later, would sign an EO attributing widespread benefit fraud to inadequate controls — simultaneously revised the internal controls circular to explicitly sideline GAO, the independent watchdog that has been documenting inadequate controls in these exact programs for three decades. GAO’s standards for internal control (the “Green Book”) are referenced throughout A-123’s history. The 2026 revision demotes their role, framing GAO’s guidance as non-binding opinion from an outside entity.

OMB Circular A-11 — which governs the federal budget process — layers on top of A-123 by requiring agencies to set performance goals and report on program outcomes annually. Together these create a system where agencies are supposed to be held to account for the very control failures now being blamed on every day Americans.

The government’s own internal controls framework, properly applied, would require agencies running programs with chronic improper payment problems to submit modified or no-assurance statements acknowledging material weaknesses — and to show Congress a corrective action plan. Instead, for decades, agencies have submitted annual financial reports while GAO continued to keep their biggest programs on the High-Risk List.

Someone signed those Assurance Statements. They had names. If you’re wondering, I advocated for no assurance statements for two years when I had a seat at the table. But no one wanted the repercussions of being the one to raise it up.

The EO frames fraud as something done to the government by bad actors in the benefit system. The A-123 paper trail suggests a more uncomfortable truth: fraud has persisted in part because agency leaders certified that controls were working when the independent evidence said otherwise — and the administration has now revised the circular to make it harder to hold those leaders to the GAO standard that caught them.

If this task force were genuinely about accountability, it would start by demanding that HHS, USDA, and VA file modified or no-assurance statements for every program on GAO’s High-Risk List. Then it would ask why that didn’t happen for the last thirty-five years. Instead, it is demanding that SNAP applicants upload more documents.

Why This Framing Is a Problem

Here’s what happens when you declare war on “benefit fraud” using inflated statistics and radicalized examples while ignoring the far larger problem of corporate and contractor fraud:

Eligible people stop applying. Research consistently shows that aggressive fraud rhetoric and increased documentation requirements don't primarily catch fraudsters — they cause eligible beneficiaries to give up. Elderly individuals who struggle with paperwork, disabled people who can't navigate complex redetermination processes, and families in crisis who don't have time to gather every document lose access to programs they legally qualify for. The technical term for this is a "chilling effect." The practical term is: people go hungry.
Actual fraud shifts to more sophisticated actors. Organized crime rings and professional fraudsters adapt. They have lawyers, shell companies, and accountants. Increasing identity verification requirements for individuals applying for SNAP does essentially nothing to stop a nursing home chain billing Medicaid for ghost patients.
The misdirection obscures the real money. Every hour of congressional attention on whether a SNAP recipient's self-reported income was $50 too high, is an hour not spent on the DOD contractor that billed $400 million for cybersecurity work that was never performed.
Program integrity gets weaponized as policy. When cries of fraud becomes the justification for cutting programs, the goal stops being fraud prevention and starts being program elimination. Using error rates as a synonym for fraud — when USDA itself disavows that interpretation — is not program integrity work. It is propaganda dressed in a spreadsheet.

What Actual Fraud Fighting Would Look Like

If this administration — or any administration — were genuinely serious about fraud, here’s what it would look like: robust False Claims Act enforcement against defense contractors and healthcare providers, not just benefit recipients. IRS funding sufficient to actually pursue high-income under-reporters, who evade far more money than every SNAP recipient combined. Targeting the organized criminal networks exploiting contracting and provider billing systems — which requires significantly more sophisticated investigation than checking whether someone’s EBT card was used in the right zip code. That means hiring skilled workers and empowering them to act.

It would fund the often ignored back-office employees and functions that aren’t forward-facing but literally build and maintain the control systems keeping everything from collapsing. It would not build a task force that explicitly excludes the programs where the most money is actually being stolen, expect it stood up in a couple of months, and duplicate work that people have been begging for help with for years.

It would not call a 10% error rate a fraud rate. It would not use two high-profile cases involving immigrants to justify mass verification burdens on millions of low-income Americans. It would not build a task force that explicitly excludes the programs and agencies where the most money is actually being stolen and expect that task force to be stood up in a couple of months. A task force that will duplicate and complicate existing work and processes that people have been begging for help with for years.

The Bottom Line

The Task Force to Eliminate Fraud is focused on the part of the federal budget where poor people live. The fraud that actually hemorrhages the most money — provider billing fraud, defense contractor fraud, the $600 billion annual tax gap — didn't make the task force's agenda.

There's a word for choosing to fight fraud that affects people with no lobbying budget while leaving fraud that benefits people with excellent lobbying budgets largely untouched. That word is not "integrity."

Fraud is real. Fighting fraud is legitimate. But fighting the right fraud, honestly, with accurate data? That would require acknowledging that the people stealing the most from American taxpayers are not the ones standing in line at the food pantry, they're the ones in suits and glass offices who pay people to shop for them.

These statements and opinions are my own and based on my experience within and knowledge of the federal government as a GS-15 director level employee. The problems started far before they exploded into the mainstream awareness, but the rhetoric and destruction will be up to me and my peers to clean up, so yes, I'm angry.

Sources cited in this post:

Have thoughts? Receipts? Even angrier statistics? Hit reply.