Why Minnesota?
Public records point to differences in how Minnesota designed and watched over these programs. Those differences may help explain why the fraud was so large here. They do not prove a single cause, and they do not prove the same thing did not happen elsewhere.
The same program in neighboring states
Cash payments from the Child and Adult Care Food Program (CACFP), the federal child-nutrition program Feeding Our Future claimed money from, taken from the same U.S. Department of Agriculture (USDA) table for Minnesota and all four states that border it. The comparison uses every state that borders Minnesota, so the neighbors were not hand-picked.
CACFP payments in Minnesota and every neighboring state
In the same federal program, Minnesota's CACFP payments ran roughly seven times Wisconsin's in FY2021 and FY2022, then collapsed by about two-thirds after the January 2022 raids and sponsor terminations. Every neighboring state's series is close to flat across all five years.
| State | FY2021 | FY2022 | FY2023 | FY2024 | FY2025* |
|---|---|---|---|---|---|
| Minnesota | 219,013,945 | 200,484,134 | 62,303,438 | 64,638,169 | 59,163,788 |
| Wisconsin | 30,896,606 | 28,979,840 | 31,753,518 | 30,907,465 | 31,335,761 |
| Iowa | 22,314,706 | 23,861,466 | 26,327,012 | 25,530,136 | 25,748,222 |
| North Dakota | 8,457,544 | 9,893,357 | 10,442,202 | 9,492,728 | 9,405,130 |
| South Dakota | 5,852,543 | 6,346,637 | 6,773,417 | 6,498,351 | 6,480,191 |
The comparison uses all four states that share a land border with Minnesota, from the same USDA table with the same as-of date. *FY2025 is preliminary. All years are federal fiscal years. Primary source:USDA payment tables
This comparison uses CACFP and not the Summer Food Service Program because the FY2021 SFSP spike was nationwide under COVID-19 waivers (Iowa's FY2021 SFSP alone was$215.5M). SFSP by itself is not Minnesota-specific evidence.
This comparison does not control for how many children each state serves, sponsor counts and sponsor mix, each state's waiver adoption choices, or claim timing. It raises the question but does not answer it.
Minnesota's own payment records show which sponsors the budget-year 2021 surge flowed through, on what the state paid. The March 2021 payment stop and the decision to start paying again are on the oversight record.
What was different
Each card states a difference the record documents, then a possible explanation, then the alternative. The record supports the differences, but it does not prove a cause.
A first-in-nation benefit design
Minnesota was the first state to cover housing-related services as a standard Medicaid benefit, under a part of federal Medicaid law (section 1915(i)) that lets a state add a benefit for everyone eligible rather than run it as a limited experiment. The program launched July 20, 2020. The state projected roughly $2.6 million per year. The program paid out $302 million over about 4.5 years, and the state ended it on October 31, 2025.
A brand-new benefit that any provider could join, with payments far past projections, outgrew the controls designed for it. The record supports that the controls were not sized to the benefit's risk. It does not support that no screening existed.
Rapid growth partly reflects real unmet need in a program serving people who are homeless or at risk. Growth alone is not fraud, and the charged conduct is a subset of the spending.
That the section 1915(i) design caused the fraud, or that running the benefit as a limited experiment (a waiver) would have prevented it.
The flagged services were paid claim by claim
The 14 Medicaid service types Minnesota flagged as high risk were paid fee-for-service, meaning the state paid each billed claim directly. Optum, the reviewer the state hired, examined only those fee-for-service claims. Its own scope statement says claims paid through managed care (health plans paid a set amount per person to manage care) were not part of the scope. The state's Medicaid director said managed-care organizations handled about $6 billion of the roughly $18 billion billed in those programs since 2018.
Fee-for-service claims met thinner gatekeeping. The state paid billed claims directly, and checks on these claims before payment were only added in December 2025.
Managed-care claims were outside the review's dataset, so fraud there may be less visible rather than less present. What the review covered shapes what is known.
That fraud concentrated in fee-for-service because of the payment model, or any comparison of actual fraud levels between the two systems.
Nonprofit sponsors at unusual scale, thinly overseen
The Minnesota Department of Education (MDE) distributed about $845 million to nonprofit sponsors across fiscal years 2018 through 2022, against about $545 million to public-entity sponsors. The Office of the Legislative Auditor (OLA) found MDE's oversight of Feeding Our Future inadequate. It found that MDE relied on child-nutrition program staff rather than trained investigators. And, in its words, MDE's choices made pandemic waivers a source of reduced oversight, not the waivers themselves.
A sponsor model moving hundreds of millions through nonprofits demanded monitoring capacity the agency did not have.
Every state operated under the same federal waivers. The OLA attributes the gap to MDE's choices, and other states' sponsor oversight is not documented here.
That any particular monitoring step would have detected the scheme earlier, or how neighboring states' sponsor oversight compared.
Legal and investigative capacity arrived late
The Legislative Auditor's exact sentence reads, “MDE did not have a general counsel until January 2022 to advise the department on how to interpret and implement complex federal regulations and identify policy changes to bolster the department's regulatory authority.” A general counsel is an agency's in-house lawyer. The statewide Office of Inspector General (OIG) was created by a law enacted in May 2026, with its main provisions taking effect January 1, 2027.
Without in-house counsel, the agency underused authority it already had. The OLA said MDE had the authority it needed to act against the sponsor.
MDE created its own internal inspector general and general counsel's office after the raids, before the statewide OIG law. Capacity was added in stages, not absent throughout.
That earlier legal capacity would have changed the outcome, or that Minnesota was unusual among states on this dimension.
How other states built it
These cards show how other states built similar benefits. They are for comparison only, and nothing here proves another state avoided the same problem.
Housing benefits run as limited experiments elsewhere
Other states used federal demonstration waivers (section 1115), which let a state run a benefit as a limited experiment, to cover housing-related services. Minnesota instead made them a standard benefit in its state Medicaid plan. Per the program director at the Minnesota Department of Human Services (DHS), that let the state assist more people while avoiding the federal requirement that demonstrations be revenue neutral.
Whether states that used waivers experienced less fraud in comparable services. No such comparison is published in the records this site has reviewed.
Checking claims before paying them, the control Minnesota added afterward
In December 2025 the state began checking claims before paying them for the 14 high-risk services. A vendor verifies that billed services were necessary, correctly documented, and provided before DHS pays. This applies to fee-for-service claims only. DHS describes it as a permanent new business process.
What other states' before-payment checks looked like during the growth years. This documents, by contrast, what was not in place in Minnesota before December 2025.
Losses in other states
The waiver rules were national, but the choices were made state by state.
The U.S. Department of Agriculture (USDA)’s onsite-monitoring waivers were nationwide. The same relief was available to every state agency that chose to use it, and desk audits were encouraged "to the maximum extent practicable." Waivers were optional, so state choices differed. The Legislative Auditor’s conclusion about Minnesota’s choices is on the COVID page.Primary source:USDA Food and Nutrition Service
In the federal audits and national reporting this site has reviewed (listed with this finding), no other state shows a directly comparable single-sponsor loss as of October 11, 2026. The largest other-state single-sponsor case in those sources is a Missouri nonprofit executive convicted of stealing $19.7 million from a child-nutrition program, per the USDA Inspector General. That is a statement about those sources, not a finding about what happened in other states.
Records reviewed: The published reports of the U.S. Department of Agriculture’s Office of Inspector General (USDA OIG) on pandemic-era child-nutrition program oversight (oig.usda.gov), the Government Accountability Office (GAO)’s published pandemic-relief oversight reporting on the child-nutrition programs (gao.gov), and the Associated Press national coverage of the Minnesota program-fraud cases cited in this section, as reviewed for this site’s research brief. As of October 11, 2026.
Announcing the first indictments in September 2022, Attorney General Merrick Garland described them as “alleging the largest pandemic relief fraud scheme charged to date.” That is a statement about the size of Minnesota’s case, not a claim that other states had zero fraud. Sen. Chamberlain’s other-states claim appears on the public statements page as an attributed, dated quote, never as a site finding.Primary source:Justice Department
Prosecutions, state by state
A separate question from the ones above. These are the statistics the U.S. Department of Health and Human Services Office of Inspector General (HHS-OIG) publishes for all 53 Medicaid Fraud Control Units, the state-level units that investigate and prosecute Medicaid fraud. They cannot explain Feeding Our Future, and they measure enforcement output, not how much fraud exists.
Read the recoveries column with care. HHS-OIG's own report says, “A large portion of the FY 2025 criminal recoveries was attributed to a case investigated by the Virginia MFCU, totaling $650M in criminal recoveries (52 percent of FY 2025 reported criminal recoveries).” The report also notes recoveries “vary from year to year, depending on the types of cases that MFCUs investigate and when those cases are adjudicated.” One year of this table ranks single-case timing as much as anything else.
7th most of 53 units
These are the state Attorney General's Medicaid Fraud Control Unit cases, such as the 2023 MN Professional personal-care prosecution. None of this site's selected defendants is an MFCU case. Minnesota's child-nutrition prosecutions (Feeding Our Future and related) are federal cases brought by the U.S. Attorney and the FBI, largely outside these statistics, so this site's list will not “match” this table.
| Rank | Unit | Fraud convictions | Fraud indictments | Open fraud investigations | Total recoveries | Recoveries per $1 billion Medicaid spend |
|---|---|---|---|---|---|---|
| 1 | Pennsylvania | 107 | 109 | 535 | $41.1M | $829,919 |
| 2 | Texas | 75 | 156 | 601 | $151.9M | $2.8M |
| 3 | Florida | 68 | 35 | 546 | $107.6M | $2.7M |
| 4 | Ohio | 64 | 93 | 797 | $25.8M | $667,331 |
| 5 | Louisiana | 44 | 13 | 472 | $17.1M | $1.1M |
| 6 | Arizona | 43 | 66 | 121 | $49.3M | $2.2M |
| 7 | Minnesota | 36 | 49 | 187 | $24.5M | $1.1M |
| 8 | Nevada | 35 | 25 | 314 | $18.5M | $2.5M |
| 9 | California | 33 | 56 | 786 | $102.1M | $588,954 |
| 10 | Missouri | 33 | 25 | 354 | $4.5M | $263,378 |
| 11 | Virginiasingle-case outlier in recoveries | 29 | 55 | 262 | $671.4M | $29.2M |
| 12 | Indiana | 28 | 38 | 633 | $144.6M | $7.1M |
| 13 | New York | 25 | 7 | 354 | $110.5M | $1.1M |
| 14 | Illinois | 20 | 27 | 276 | $18.2M | $504,236 |
| 15 | Arkansas | 19 | 23 | 163 | $3.5M | $397,942 |
| 16 | Massachusetts | 19 | 26 | 449 | $139.9M | $4.9M |
| 17 | Tennessee | 18 | 18 | 162 | $16.7M | $941,277 |
All 53 units, ranked
| Rank | Unit | Fraud convictions | Fraud indictments | Open fraud investigations | Total recoveries | Recoveries per $1 billion Medicaid spend |
|---|---|---|---|---|---|---|
| 1 | Pennsylvania | 107 | 109 | 535 | $41.1M | $829,919 |
| 2 | Texas | 75 | 156 | 601 | $151.9M | $2.8M |
| 3 | Florida | 68 | 35 | 546 | $107.6M | $2.7M |
| 4 | Ohio | 64 | 93 | 797 | $25.8M | $667,331 |
| 5 | Louisiana | 44 | 13 | 472 | $17.1M | $1.1M |
| 6 | Arizona | 43 | 66 | 121 | $49.3M | $2.2M |
| 7 | Minnesota | 36 | 49 | 187 | $24.5M | $1.1M |
| 8 | Nevada | 35 | 25 | 314 | $18.5M | $2.5M |
| 9 | California | 33 | 56 | 786 | $102.1M | $588,954 |
| 10 | Missouri | 33 | 25 | 354 | $4.5M | $263,378 |
| 11 | Virginiasingle-case outlier in recoveries | 29 | 55 | 262 | $671.4M | $29.2M |
| 12 | Indiana | 28 | 38 | 633 | $144.6M | $7.1M |
| 13 | New York | 25 | 7 | 354 | $110.5M | $1.1M |
| 14 | Illinois | 20 | 27 | 276 | $18.2M | $504,236 |
| 15 | Arkansas | 19 | 23 | 163 | $3.5M | $397,942 |
| 16 | Massachusetts | 19 | 26 | 449 | $139.9M | $4.9M |
| 17 | Tennessee | 18 | 18 | 162 | $16.7M | $941,277 |
| 18 | Washington | 15 | 15 | 474 | $45.1M | $2.0M |
| 19 | Michigan | 10 | 9 | 427 | $17.3M | $634,274 |
| 20 | Oklahoma | 9 | 15 | 316 | $2.9M | $269,615 |
| 21 | Alaska | 8 | 17 | 76 | $1.2M | $366,372 |
| 22 | Iowa | 8 | 20 | 231 | $2.4M | $269,883 |
| 23 | Kentucky | 7 | 18 | 172 | $22.0M | $1.1M |
| 24 | New Jersey | 7 | 9 | 149 | $30.8M | $1.2M |
| 25 | Vermont | 7 | 8 | 45 | $3.1M | $1.3M |
| 26 | Alabama | 6 | 5 | 28 | $2.9M | $350,110 |
| 27 | Connecticut | 6 | 10 | 159 | $6.9M | $586,596 |
| 28 | Georgia | 6 | 8 | 438 | $46.5M | $2.4M |
| 29 | Nebraska | 6 | 0 | 51 | $1.2M | $281,523 |
| 30 | New Hampshire | 6 | 1 | 49 | $356,807.56 | $116,329 |
| 31 | South Carolina | 6 | 19 | 70 | $22.1M | $2.2M |
| 32 | South Dakota | 6 | 1 | 62 | $225,578.04 | $121,023 |
| 33 | Mississippi | 5 | 7 | 113 | $1.2M | $152,015 |
| 34 | North Carolina | 5 | 3 | 360 | $16.6M | $482,398 |
| 35 | Wisconsin | 5 | 13 | 120 | $1.9M | $138,799 |
| 36 | Delaware | 4 | 3 | 204 | $619,300.3 | $182,117 |
| 37 | Maryland | 3 | 5 | 330 | $3.4M | $179,258 |
| 38 | Utah | 3 | 2 | 103 | $2.7M | $512,746 |
| 39 | Colorado | 2 | 0 | 282 | $58.0M | $3.5M |
| 40 | District of Columbia | 2 | 3 | 54 | $289,914.83 | $67,515 |
| 41 | Idaho | 2 | 5 | 143 | $900,756.39 | $216,168 |
| 42 | Kansas | 2 | 7 | 73 | $369,231.23 | $56,810 |
| 43 | New Mexico | 2 | 7 | 331 | $4.9M | $471,830 |
| 44 | North Dakota | 2 | 4 | 59 | $1.9M | $1.1M |
| 45 | Oregon | 2 | 11 | 69 | $13.7M | $723,380 |
| 46 | Puerto Rico | 2 | 2 | 59 | $1.3M | $271,215 |
| 47 | Rhode Island | 2 | 1 | 115 | $1.6M | $368,565 |
| 48 | West Virginia | 2 | 0 | 175 | $1.8M | $311,577 |
| 49 | Montana | 1 | 6 | 48 | $711,088.63 | $270,056 |
| 50 | Wyoming | 1 | 5 | 62 | $1.5M | $1.8M |
| 51 | Hawaii | 0 | 0 | 388 | $453,950.81 | $142,786 |
| 52 | Maine | 0 | 0 | 30 | $343,479.02 | $66,019 |
| 53 | U.S. Virgin Islands | 0 | 0 | 25 | $1,171.06 | $11,186 |
The table counts fraud convictions only (856 of the 1185 nationally). Patient abuse and neglect convictions are excluded. Recoveries are amounts ordered, not cash collected, and the federal report does not split them between fraud and abuse cases. “Recoveries per $1 billion Medicaid spend” divides each unit's reported recoveries by its state's total Medicaid spending from the same report. It measures enforcement output, not a fraud rate.
Units differ in how they are organized and what authority they have. Most sit in state Attorney General offices, and each must have statewide prosecution authority or formal referral procedures. Differences here reflect structure and case timing as much as activity. Data reported to HHS-OIG by the units, current as of February 23, 2026. Primary source:HHS Office of Inspector General
People charged but not convicted are presumed innocent, and amounts labeled alleged come from charging documents, not proven losses. Method and corrections