The line at a Philadelphia grocery store moves slowly. A mother, still in the fluorescent vest from her shift at a big-box warehouse, sets milk, bread, and a carton of eggs on the belt. When the cashier announces the total, she swipes an EBT card—the debit card used to access federal food assistance through the Supplemental Nutrition Assistance Program, or SNAP. She anxiously waits for the register’s approval beep.
She is one of the nearly one in three Philadelphians who rely on SNAP to help pay for groceries. Her benefits averages about six dollars a day—meant to cover every meal her family will eat. At the checkout counter, she can pay for the milk, bread, and eggs in front of her. By the fourth week of the month, the benefits are often gone. What remains is the long stretch of days before the next SNAP deposit, when getting even one solid meal on the table can be a struggle.
A working mother stretching groceries at the end of the month is rarely the image invoked in the national conversation about welfare beneficiaries. Instead, the political imagination returns to a different figure: a woman portrayed as gaming the system—collecting benefits she doesn’t deserve, living comfortably at public expense. This is the image of the “welfare queen,” a trope popularized in the Reagan era that has long stood in for a broader suspicion: that poor people, especially women, are dishonest recipients of aid.
But the trope was never confined to a single caricature. In practice, it has functioned as a visual shorthand for suspicion—one that attaches most readily to those already marked as marginal. Decades of research show that when Americans are asked to picture a “typical” welfare recipient, they disproportionately imagine a Black single mother, even though public assistance has always served a far more diverse population. The image reflects repetition more than reality—shaped by political messaging, news coverage, and campaign ads that pair poverty with racialized cues: urban settings, large families, idleness.
Over time, that association extends beyond its original frame. It travels easily, attaching itself to new groups cast as undeserving: immigrant families, non-English speakers, communities already viewed as outsiders. The specifics change, but the underlying logic remains the same. Certain people are rendered visible as potential fraudsters long before any evidence is considered.
That pattern of suspicion resurfaced in Donald Trump’s 2026 State of the Union address, where Minnesota was cast as the epicenter of a supposed national fraud crisis. Trump pointed to prosecutions tied to the state’s Somali community, invoking images of widespread abuse of public programs and repeating an unverified claim that as much as $19 billion had been taken.
Building on that claim, Trump announced a new federal “war on fraud,” placing Vice President JD Vance in charge and elevating what had largely been state-level investigations into a centerpiece of his administration’s agenda.
The next day, at a press conference in Washington, Vance—standing alongside Centers for Medicare and Medicaid Services administrator Dr. Mehmet Oz—announced that the administration would halt roughly $259 million in Medicaid reimbursements to Minnesota until the state demonstrated “good stewardship” of federal funds.
Minnesota had already been under scrutiny over the Feeding Our Future case, a sweeping federal prosecution involving nonprofit meal providers operating within Somali immigrant communities in the Twin Cities. In late December 2025, the issue exploded into the national spotlight after a viral video by YouTuber Nick Shirley accused Somali-run daycare centers of widespread fraud.
The administration seized that moment to launch a immigration enforcement operation, deploying ICE agents across Minneapolis and St. Paul and tying the operation to welfare-fraud investigations. However the results of this operation diverged sharply from its stated justification: thousands were detained, while investigators revealed little to no connection between those arrests and the fraud cases used to justify the sweep.
The pattern illustrates how the welfare-fraud narrative operates. It casts suspicion on people in poverty—particularly immigrants, people of color, and mothers—while reshaping programs that millions of working families rely on to survive, including SNAP, the nation’s largest anti-hunger program. The mother in the grocery line becomes the backdrop to a political story that treats people in poverty less as workers struggling to afford food than as potential suspects.
At Murphy’s Giving Market in Upper Darby, PA, pantry founder and director Desiree LaMarr Murphy says the people coming through the doors look very different from the caricatures that dominate national debates about welfare. “The stereotype of the ‘welfare queen’ has always been deeply upsetting because it’s built on a false narrative,” Murphy said. “Many people imagine SNAP recipients as people who don’t want to work. But what we’re seeing are working families, older adults, and people who simply can’t make the math work anymore.”
Most families arrive after work, still in uniforms, hoping to stretch what they bought at the store. The rise in working family food insecurity has skyrocketed since 2020. Since then, food prices have climbed roughly 25% nationwide and rents about 20% while wages for many lower-income workers have barely moved. Paychecks that once covered groceries for a week now run out days sooner.
Federal SNAP data mirrors what Murphy sees in the food pantry line. Many SNAP households include working adults whose paychecks do not cover the rising costs of food, housing, transportation, and childcare. Another two-thirds of SNAP recipients include children, the elderly, and people with disabilities.
Murphy also sees families who do not appear in those statistics: households that earn too much to qualify for SNAP but still cannot afford enough food. That’s because SNAP eligibility is tied to the federal poverty line—a formula created in the early 1960s, when economists assumed families spent about one-third of their income on food. More than sixty years later, that formula still anchors federal policy, even though housing, childcare, and healthcare now consume far larger shares of household budgets.
In Philadelphia, median rents approach $1,700 a month, before utilities, transportation, childcare, or groceries. A family earning $35,000 a year is therefore officially considered above poverty. But by most realistic cost-of-living measures, that income does not come close to covering basic expenses. The MIT Living Wage estimates a single adult in Philadelphia needs roughly $40,000 a year to get by, while a family with children requires closer to $80,000.
Local officials highlight the mismatch between federal poverty metrics and Philadelphia’s growing rates of housing and food insecurity. Philadelphia City Councilmember Kendra Brooks said the decades-old formula used to measure poverty fails to capture the financial reality many families face. “The federal poverty guidelines don’t account for the extremely high cost of housing in a city like Philadelphia,” Brooks said. “More than half of Philadelphia renters are cost-burdened, spending over 30 percent of their income on rent, and we’re facing a severe shortage of affordable housing. Federal cuts to programs like SNAP and Medicaid are forcing working families to make impossible choices between housing, healthcare, and food.”
Just maintaining those benefits often requires navigating administrative hurdles that can be difficult for many households. SNAP recipients must regularly submit pay stubs, respond to verification letters, complete eligibility interviews, and recertify their income and household information.
In Philadelphia, where roughly 30 percent of households do not own a vehicle, even routine paperwork can become a logistical puzzle. A worker with an irregular schedule may have to find time between shifts to ride two buses to a benefits office, upload documents from a phone with limited data, or track down a printer to send pay stubs before a deadline. Missing a notice or failing to submit a document on time can mean the benefits stop—forcing families to begin the process again. Research on administrative burdens across benefit programs shows the consequences: between 15 and 30 percent of eligible participants experience procedural terminations, meaning benefits end not because income has risen but because paperwork was incomplete or late.
The same administrative process that cuts off eligible households is also where most SNAP fraud is detected. Typically, the violations are small: a household may underreport income, fail to list a working adult in the home, or misstate household size to qualify for a slightly larger benefit.
For decades, accusations that poor families and marginalized communities are defrauding SNAP and the American taxpayer have generated headlines and congressional speeches. In reality, federal estimates suggest only about 1 to 2 percent of SNAP benefits are lost to fraud—roughly $1 to $2 billion annually.
The real scandal lies elsewhere. When welfare fraud does occur, it rarely looks like our image of the welfare queen- or even the family misreporting income on a SNAP application. It operates through corporations managing publicly funded welfare programs.
That misdirection has deep roots. The political fixation on the Reagan-era “welfare queen” myth helped justify cutting public programs and shifting large parts of the welfare system into private hands. In the decades that followed, core government functions were carved into contracts and handed to private firms.
Today, nearly all welfare spending flows through those institutions rather than directly to families. In 2021, 97% of public welfare spending went toward operational costs—payments to Medicaid providers, private contractors, nonprofit service organizations, and the administrative systems that manage them. The largest share—more than $700 billion—went to vendor payments for medical care alone.
Only a small fraction remains. Roughly three cents of every dollar reach families as direct cash assistance through programs like SNAP, Temporary Assistance for Needy Families, or TANF, Supplemental Security Income, and the Federal Low Income Home Energy Assistance Program.
This structure shapes what the system produces—and how fraud is understood. Even if fraud among welfare recipients were widespread, it would necessarily involve only a small share of total spending. Most public welfare money is absorbed elsewhere.
Those institutions do not simply “administer” benefits—they build entire systems around them. Consultants redesign eligibility rules, write policy manuals, and advise states on how to reduce caseloads. Technology contractors build platforms where people apply for aid, upload documents, and receive approvals or denials. Call centers field questions. Case management companies run job-training and work programs, tracking attendance and monitoring compliance.
Each step generates revenue. Software platforms require ongoing contracts. Call centers bill per interaction. Compliance systems generate fees for monitoring and reporting. Consultants are paid to evaluate the system, then paid again to redesign it.
Private equity firms often sit one layer above, acquiring these vendors and consolidating them into larger platforms. A company that runs call centers may be merged with one that builds eligibility software, allowing a single firm to hold multiple contracts within the same program. Revenue flows not just from delivering aid, but from managing every step around it.
That management requires rules that are detailed, shifting, and often difficult to navigate. Contractors help design those rules: how many hours a parent must log, what counts as qualifying work, how often eligibility must be recertified, and what documentation must be uploaded. Each requirement creates a new task. Each task creates a new layer of administration.
A parent may be required to attend a job-readiness workshop, log hours in an online portal, and respond to automated notices. Missing a session, uploading a document incorrectly, or misunderstanding a rule can trigger a sanction. Benefits are reduced or terminated.
But the system around that decision continues to be paid. The company running the workshop is paid to host it. The firm managing the digital portal is paid to maintain it. The contractor overseeing compliance is paid to track attendance and produce reports. When rules change—as they often do—consultants are brought in to redesign the system, generating new rounds of contracts.
Over time, the system grows more complex and more dependent on the companies that built it. Rules multiply. Guidance shifts. Agencies rely on the same contractors to interpret the policies those contractors helped design. What appears, on paper, as “program integrity” becomes a continuous stream of administrative activity—work created to manage the rules, and more work created to manage their consequences.
A similar structure shapes publicly financed healthcare– one of the largest welfare systems in the country. Medicaid and Medicare are funded by taxpayers, but much of the care is delivered through private managed-care companies. States pay these companies a fixed monthly rate for each person enrolled. The basic equation is straightforward: more enrollees mean more revenue; profit depends on how much of that payment is spent on care.
What determines whether someone stays enrolled, however, is often far removed from medical need.
Between 2023 and 2024, 16.2 million people—including 4 million children—lost Medicaid coverage. By September 2023, roughly 73 percent of those disenrollments were procedural. People were not necessarily found ineligible; they were cut off because renewal notices never reached them, paperwork wasn’t processed in time, or forms were incomplete.
In practice, coverage can hinge on small administrative moments. A renewal packet is mailed to an old address after a family moves. A pay stub uploaded from a cracked phone screen never registers in the system. A caseworker misses a deadline, and the file auto-closes. A child’s coverage lapses while a form sits marked “pending.” The loss of insurance begins not with a diagnosis, but with an envelope, a portal, a missing signature.
Yet the companies administering Medicaid did not shrink alongside enrollment. Georgetown researchers found that the five largest Medicaid managed-care companies lost 4.3 million enrollees between March and December 2023. Even so, several reported higher Medicaid revenue that year. UnitedHealth’s Medicaid revenue rose from $63.8 billion in 2022 to $75.0 billion in 2023. Centene’s increased from $84.1 billion to $86.8 billion. Molina’s grew from $24.8 billion to $26.3 billion.
Executive compensation followed a similar pattern. UnitedHealth CEO Andrew Witty’s total pay increased from about $20.9 million in 2022 to $23.5 million in 2023. Centene CEO Sarah London’s compensation rose from roughly $13.2 million to $18.6 million over the same period.
There is no evidence that executives are paid more because individual people lose coverage. The connection is more indirect. Compensation is tied to revenue, margins, and stock performance—metrics that can remain strong even as enrollment falls, particularly if higher-cost patients are removed or care is tightly managed.
The result is a system where millions can lose coverage over missed paperwork while the companies running that coverage continue to grow—and where the financial signals at the top remain largely unchanged by the losses below.
We are also paying more for this layer of private management than we would if care were delivered through the public system itself.
The federal advisory commission MedPAC estimates that in 2025, taxpayers paid private Medicare Advantage plans about 20 percent more per enrollee than it would have cost to cover the same patients in traditional Medicare. That difference is not abstract. It adds up to roughly $84 billion in additional spending in a single year. Put another way: for every five dollars spent covering a Medicare Advantage patient, about one dollar reflects the added cost of routing care through private insurers rather than the public program.
Corporate capture of welfare funds does not stop at legal contracting. In some cases, it moves into fraud.
In Medicare Advantage, payments are adjusted based on how sick a patient appears on paper. That creates an incentive to document as many conditions as possible. Teams of coders and consultants review medical charts, searching for diagnoses that can be added or reclassified. A complication is noted. A condition is coded at a higher severity. Each change increases the payment tied to that patient. Across thousands of records, those adjustments can translate into hundreds of millions of dollars.
When those practices cross the line, they rarely resemble the version of fraud emphasized in political speeches. In fiscal year 2025, the U.S. Department of Justice recovered $6.8 billion from fraud involving federal programs. About $5.7 billion—by far the majority—came from healthcare: billing schemes, reimbursement manipulation, and corporate conduct.
Accountability, however, does not land evenly.
In Minnesota’s Feeding Our Future case, defendants accused of inflating meal counts in a federally funded nutrition program face decades in prison; one organizer has already been sentenced to 28 years. On the other hand, Kaiser Permanente affiliates recently agreed to pay $556 million to resolve allegations that physicians were encouraged to add diagnoses that increased federal payments. The case ended without a finding of liability, and the company continued operating in federal programs.
While Donald Trump declares his “war on fraud” in public statements, he has been granting pardons and commutations to individuals convicted in large-scale healthcare fraud cases, including schemes involving hundreds of millions of dollars in Medicare billing. Analyses estimate that these actions reduced or eliminated more than $1 billion in restitution and penalties. Even as those decisions were issued, public messaging continued to emphasize fraud in programs like Medicaid.
Millions of people lose coverage over missed paperwork—renewal forms, documentation gaps, deadlines. Enforcement tightens at the point of entry. At the other end of the system, cases involving far larger sums are resolved through settlements, reduced penalties, or clemency.
For Philadelphia professor and Paying the Price author Sara Goldrick-Rabb, the fixation on fraudulent recipients of SNAP benefits reveals a deeper distortion in how poverty is understood. “The national obsession with welfare fraud isn’t just a distraction—it’s a form of gaslighting directed at the American worker,” she said. “When we see hundreds of billions in taxes go uncollected at the top while scrutinizing the grocery purchases of struggling families, we’re choosing to punish the vulnerable for the ‘crime’ of being poor.”
The fraud narrative makes cuts to direct poverty aid politically possible. Programs portrayed as riddled with abuse become easier to shrink, even when most recipients are working families, children, and seniors. And while the country debates the grocery purchases of struggling parents, far larger systems moving public money remain largely out of view.
This is how millions of Americans—including thousands in Philadelphia—now face the possibility of losing the food assistance they rely on to feed their families. Congress is debating major cuts to anti-poverty programs. Provisions in the administration’s “Big Beautiful Bill” —a sweeping legislative package—tighten work requirements for food assistance and shift more program costs to states—changes analysts say could lead to significant reductions in SNAP participation and push hundreds of thousands of households off the program.
When those cuts arrive, they do not appear in executive salaries. They appear at kitchen tables. Grocery budgets shrink. Meals are skipped. Food pantry lines grow longer. Far from the checkout line, quarterly profits rise and executive bonuses grow. The money does not disappear, it is directed upwards.
But the narrative does not follow the money. It stays fixed on the person at the register—scrutinizing the smallest transaction in the system while the largest transfers happen out of sight.
Leave a Reply