Feeding Our Future submitted reporting documents.
The documents listed children, counted meals, recorded food purchases and supported reimbursement claims.
On paper, the programme appeared to be feeding millions of children.
The records looked complete.
The underlying activity was not real.
In March 2025, a federal jury convicted Feeding Our Future founder Aimee Bock of wire fraud, conspiracy and bribery offences. In May 2026, she was sentenced to 500 months in prison, equal to more than 41 years.
The United States Department of Justice said Feeding Our Future and the sites under its sponsorship fraudulently obtained and distributed more than $240 million from federal child-nutrition programmes.
The scheme relied on reporting records such as:
- fake attendance rosters;
- false meal counts;
- fabricated supplier invoices;
- invented children;
- false ages;
- shell companies;
- and claims from sites that could not realistically have served the reported number of meals.
The fraud did not succeed because nobody submitted reports.
It succeeded because fabricated reports were treated as evidence that funded work had happened.
That is the central lesson:
> A completed reporting template is not proof of delivery. The underlying people, activities, suppliers and transactions still need to be checked.
This article is part of the Process Failure Case Studies series, which examines what happens when information, review and reporting processes fail.
Who this guide is for
This case is relevant to:
- donor-funded programme teams;
- nonprofits and NGOs;
- grant-making organisations;
- government funding agencies;
- programme managers;
- monitoring and evaluation teams;
- donor-reporting officers;
- finance and grants teams;
- and organisations paying partners based on reported delivery.
Most organisations will never deal with fraud on the scale of Feeding Our Future.
They can still face the same underlying risk when:
- partner reports are accepted without verification;
- beneficiary lists are easy to fabricate;
- activity totals are not compared with capacity;
- invoices do not link to delivery evidence;
- rapid growth is treated as success rather than a reason to check;
- and the same organisation collects, approves and submits its own evidence.
What was Feeding Our Future?
Feeding Our Future was a Minnesota nonprofit that acted as a sponsor in two federally funded child-nutrition programmes:
- the Child and Adult Care Food Program;
- and the Summer Food Service Program.
The programmes were funded by the United States Department of Agriculture.
In Minnesota, the Minnesota Department of Education administered the programmes.
Meals were served through local sites. A site could be a childcare provider, community organisation, restaurant or other approved food-distribution location.
Each site needed a sponsor.
The sponsor was responsible for:
- applying for sites;
- overseeing the sites;
- monitoring programme compliance;
- collecting meal records;
- preparing reimbursement claims;
- submitting the claims to the state;
- and distributing approved funds to the sites.
Feeding Our Future performed that sponsor role.
This placed the organisation in the middle of the funding route:
Meal site
→ Feeding Our Future
→ Minnesota Department of Education
→ Federal reimbursement
→ Feeding Our Future
→ Payment to meal site
Feeding Our Future could retain part of the funding as an administrative fee.
That arrangement created a direct link between the volume of reported meals and the money moving through the organisation.
What happened?
During the COVID-19 pandemic, federal rules were temporarily changed to make it easier to distribute food while schools and normal programme sites were disrupted.
For-profit restaurants were allowed to participate, and meals could be distributed away from normal educational settings.
Those changes were intended to help children receive food during an emergency.
They also reduced some of the normal barriers and monitoring arrangements.
According to the Department of Justice’s account of the scheme, Feeding Our Future employees recruited people and organisations to open food-programme sites across Minnesota.
Many of the sites claimed to begin serving thousands of children a day within days or weeks of being formed.
Feeding Our Future eventually sponsored more than 250 sites.
The amount moving through the organisation grew rapidly:
| Year | Federal funding received and distributed |
|---|---|
| 2019 | Approximately $3.4 million |
| 2021 | Nearly $200 million |
The Department of Justice said the organisation and participating sites fraudulently obtained and distributed more than $240 million.
At Bock’s trial, prosecutors proved that the organisation falsely claimed to have served 91 million meals.
Why was the data fabricated?
The programme reimbursed sites based on meals reportedly served to children.
More reported meals meant larger reimbursement claims.
The reported activity therefore had a direct financial effect:
More children reported
→ More meals claimed
→ Larger reimbursement
→ More money paid to the sponsor and sites
Feeding Our Future also received administrative fees linked to the claims submitted by the sites it sponsored.
The Department of Justice said the organisation received more than $18 million in administrative fees to which it was not entitled.
Employees also received bribes and kickbacks from participating sites. Some of these payments were disguised as consulting fees paid through shell companies.
The false reporting had a clear purpose.
It made sites appear to be delivering far more food than they actually delivered, which allowed them to claim larger reimbursements.
The proceeds were used for:
- luxury vehicles;
- residential property;
- commercial property;
- real estate in the United States and abroad;
- and international travel.
How did the reimbursement process work?
The reporting route was simple enough to understand.
A participating site would record:
- how many children attended;
- how many meals were served;
- which days meals were provided;
- what food had been purchased;
- and which suppliers had provided it.
Those records supported the reimbursement claim.
The sponsor compiled or submitted the claims to the Minnesota Department of Education.
The state then received federal reimbursement funding and passed the money through the sponsor to the participating sites.
In a functioning programme, the reporting chain should have looked like this:
Real children attend
→ Meals are prepared or purchased
→ Meals are distributed
→ Attendance and meal records are completed
→ Supplier invoices support the claimed volume
→ Sponsor checks the evidence
→ State reviews the claim
→ Reimbursement is paid
In the fraudulent scheme, the route was different:
Sites claim children attended
→ Names and ages are fabricated
→ Meal counts are inflated
→ Invoices are created to support the false volume
→ Sponsor submits the claims
→ Reimbursement is paid
→ Money is diverted
The reporting documents were not describing the programme.
They were being used to manufacture the appearance of a programme.
What records were falsified?
The scheme relied on several linked types of false evidence.
1. Fake attendance rosters
Attendance rosters were supposed to identify the children receiving meals.
The records contained names and ages.
Prosecutors found that some lists had been created using names taken from a random-name website.
In another part of the scheme, an Excel formula inserted random ages between seven and 17 into a roster.
A spreadsheet containing names and ages can look structured.
That does not mean the children exist.
A beneficiary list becomes useful evidence only when the organisation can verify:
- that the person exists;
- that the person was eligible;
- that the person attended;
- that the activity occurred on the recorded date;
- and that duplicate or invented records have been removed.
2. False meal counts
Sites submitted forms stating how many meals had been served.
Some claimed to be serving thousands of children every day shortly after opening.
One site connected to Safari Restaurant claimed to be serving 5,000 children per day, seven days a week, only a few months after entering the programme.
Between April 2020 and November 2021, the site claimed to have served more than 3.9 million meals.
The programme did not only need a total.
It needed to test whether the total was physically plausible.
3. Fabricated invoices
Invoices were submitted to show that food had been purchased.
An invoice can look like supporting evidence because it contains:
- a supplier name;
- a date;
- items;
- quantities;
- and an amount.
The invoice still needs to be verified.
The reviewer should be able to establish:
- whether the supplier exists;
- whether the supplier had the capacity to provide the goods;
- whether the goods were delivered;
- whether the quantities match the programme claim;
- whether payment was made;
- and whether the supplier is connected to the person submitting the claim.
A PDF invoice is not proof that a transaction happened.
4. Shell companies
Shell companies were created to enrol as meal sites, receive funds and move the proceeds.
Other shell companies were used to disguise kickbacks as consulting fees.
This made the reporting structure appear more complicated and more legitimate.
A payment to a registered company can look safer than a transfer to an individual.
The company still needs to be checked.
The core reporting failure
The Feeding Our Future case is often described as pandemic fraud or nonprofit fraud.
For donor and programme teams, it is also a reporting failure.
The system received:
- beneficiary lists;
- attendance records;
- activity totals;
- invoices;
- sponsor assurances;
- and reimbursement claims.
The problem was that the records were not reliably connected to real delivery.
The process placed too much weight on whether a document had been submitted and not enough weight on whether the document could be verified.
The distinction is simple:
| Reporting completion | Delivery verification |
|---|---|
| Attendance sheet submitted | Participants independently confirmed |
| Meal total entered | Volume tested against site capacity |
| Invoice attached | Supplier and delivery verified |
| Sponsor approved claim | Sponsor independence and monitoring checked |
| Claim fits template | Data makes sense across other records |
| Report submitted on time | Reported activity actually occurred |
An organisation can meet every reporting deadline while still providing false information.
The sponsor was supposed to provide oversight
Feeding Our Future was not merely another meal site.
It was the sponsor responsible for overseeing its sites.
This meant it was supposed to:
- monitor the sites;
- check compliance;
- prepare claims;
- and provide assurance to the state.
But the sponsor received administrative fees based on the claims moving through the system.
According to the Department of Justice, Feeding Our Future also received bribes and kickbacks from sites it sponsored.
The organisation that was supposed to check the evidence benefited from approving more activity.
That is a conflict built into the information route.
A sponsor can legitimately receive an administration fee.
The funder then needs controls that do not depend only on the sponsor’s own assurance.
What the Minnesota oversight review found
In June 2024, Minnesota’s Office of the Legislative Auditor published a review of the Minnesota Department of Education’s oversight of Feeding Our Future.
The auditor concluded that inadequate oversight created opportunities for fraud.
The official review summary found that the department:
- approved applications despite unresolved concerns;
- did not always verify statements made by Feeding Our Future;
- failed to follow up properly after a 2018 review identified serious findings;
- conducted limited offsite monitoring during the pandemic;
- received at least 30 complaints involving Feeding Our Future or its sites;
- failed to investigate some complaints adequately;
- asked Feeding Our Future to investigate complaints about itself;
- and deferred serious deficiencies without reasonable assurance that corrective action had been completed.
The Minnesota Department of Education disputed the auditor’s description of its oversight as inadequate.
The department said it had met applicable standards, identified concerns, made referrals to law enforcement and later introduced stronger controls.
Those changes reportedly included:
- creating an Office of Inspector General;
- adding a General Counsel’s Office;
- updating fraud-reporting procedures;
- training staff;
- and commissioning financial reviews of selected sponsors.
The disagreement matters.
The people committing the fraud were responsible for the fraud.
That does not remove the need to examine why the controls failed to detect or stop it earlier.
Where the process failed
The case exposes weaknesses at several stages.
1. Site approval did not establish whether the sites could deliver
A site should not be approved only because its application contains the correct fields.
The funder or programme administrator needs to understand:
- who owns the site;
- who manages it;
- where it operates;
- how meals will be prepared;
- where food will be stored;
- how many people it can serve;
- which suppliers it will use;
- and whether the proposed volume is plausible.
A new restaurant claiming to serve several thousand children every day should trigger a deeper review.
The application is the start of verification.
It is not the end.
2. Reported growth was not treated as an exception
Feeding Our Future grew from approximately $3.4 million in programme funds in 2019 to nearly $200 million in 2021.
Growth can indicate that a programme is meeting a real need.
It can also indicate error or fraud.
A controlled system should flag:
- rapid increases in beneficiaries;
- rapid increases in expenditure;
- newly created sites claiming large volumes;
- claims far above comparable sites;
- repeated maximum-value claims;
- and sudden changes without matching staffing or infrastructure.
The system should not automatically reject the claim.
It should require an independent review before more money is released.
3. Beneficiary records were easy to fabricate
A name and age were enough to make a roster look complete.
A stronger system could have checked for:
- duplicate names;
- duplicate children across sites;
- repeated address or contact details;
- unusual age patterns;
- impossible attendance across locations;
- names copied across reporting periods;
- and identical attendance numbers over long periods.
For programmes involving vulnerable people, verification must also protect privacy.
The answer is not to publish children’s personal information.
It is to maintain controlled records that authorised reviewers can check.
4. Meal totals were not reconciled with operational capacity
A meal site has physical limits.
It has:
- kitchen space;
- staff;
- food-storage capacity;
- opening hours;
- transport routes;
- suppliers;
- equipment;
- and a local population.
A claim of 5,000 meals a day should be tested against those limits.
A capacity check could ask:
- How many meals can the kitchen prepare per hour?
- How many staff members are present?
- How much food was purchased?
- Where was it stored?
- How was it distributed?
- How many vehicles were used?
- How many children live or attend programmes nearby?
- Does the site have evidence matching the claimed scale?
This is where qualitative judgement and structured data need to work together.
5. Invoices were not sufficiently connected to delivery
An invoice should not sit in a reporting folder as a separate attachment.
It should link to:
- the supplier;
- the purchase order;
- the programme site;
- the reporting period;
- the payment;
- the goods received;
- and the activity being claimed.
The funder should be able to compare:
Food purchased
→ Meals that could be produced
→ Meals reportedly served
→ Children reportedly present
A large difference should create an exception.
6. The organisation was asked to investigate complaints about itself
The legislative auditor found that the Minnesota Department of Education sometimes asked Feeding Our Future to investigate complaints involving Feeding Our Future.
That is not independent fact-finding.
The subject of a complaint may provide information.
It should not control the investigation.
A complaint process should include:
- protected intake;
- named ownership;
- risk classification;
- evidence preservation;
- an independent investigator;
- limited disclosure to the subject;
- findings;
- actions;
- and follow-up.
The investigator needs access to records outside the accused organisation’s control.
7. Corrective actions were recorded but not properly followed up
A review can identify problems.
A corrective-action plan can list what must change.
Neither matters if nobody verifies completion.
A follow-up process should show:
| Field | Purpose |
|---|---|
| Finding | What was wrong |
| Required action | What must change |
| Responsible person | Who owns it |
| Due date | When it must be complete |
| Evidence required | What proves completion |
| Reviewer | Who checks it |
| Status | Open, submitted, verified or rejected |
| Follow-up date | When the next check occurs |
“Completed” should mean the evidence has been reviewed.
It should not mean the subject said the action was complete.
The wrong lesson is to add more reporting forms
The Feeding Our Future scheme already produced large volumes of documentation.
The fraud depended on documents.
Adding another form would not necessarily have solved the problem.
A longer reporting template could have created more false fields to complete.
The stronger response is to improve:
- verification;
- reconciliation;
- risk scoring;
- independent review;
- exception handling;
- and payment controls.
Reporting should collect the information required for review.
It should not create paperwork for its own sake.
The wrong lesson is to assume fraud detection can be automated completely
Several parts of this scheme could have been flagged by data rules.
For example:
- a new site reporting thousands of daily meals;
- rapid funding growth;
- repeated names;
- random age patterns;
- claims exceeding site capacity;
- identical attendance totals;
- suppliers connected to site operators;
- and invoices inconsistent with reported delivery.
Automation could have surfaced these patterns.
It could not decide the full case.
A person would still need to:
- inspect the site;
- speak to suppliers;
- confirm beneficiaries;
- examine bank records;
- assess explanations;
- protect complainants;
- and decide whether payments should continue.
The better workflow is:
System flags risk
→ Independent reviewer investigates
→ Evidence is recorded
→ Claim is approved, held, reduced or rejected
What a stronger grant-reporting workflow could look like
A safer process would connect the organisation, site, beneficiary records, expenditure, delivery evidence and payment decision.
| Stage | Human responsibility | System support |
|---|---|---|
| Sponsor approval | Assess governance, experience and conflicts | Store verification records and risk score |
| Site approval | Confirm location, staff, facilities and realistic capacity | Compare application volumes with site profile |
| Beneficiary intake | Record eligible participants appropriately | Check duplicates and unusual patterns |
| Activity delivery | Confirm meals or services took place | Record dates, locations, staff and evidence |
| Supplier purchase | Confirm goods were ordered and received | Link supplier, invoice, payment and delivery |
| Partner reporting | Submit the activity and cost records | Require complete linked fields |
| Sponsor review | Check site evidence and investigate exceptions | Record reviewer, findings and decision |
| Independent monitoring | Test selected sites and high-risk claims | Maintain inspection and sample records |
| Reimbursement | Pay only supported claims | Hold or route flagged claims for approval |
| Donor reporting | Report verified activity, spending and exceptions | Produce status-based reporting views |
| Corrective action | Confirm identified problems were fixed | Track evidence and independent verification |
The system provides structure.
People remain responsible for the decisions.
What funders should verify before paying a claim
A grant or reimbursement team should not ask only whether the report was submitted.
It should ask:
Does the organisation exist and operate as described?
Check:
- registration;
- directors or office bearers;
- physical address;
- bank account ownership;
- related entities;
- conflicts of interest;
- and prior programme history.
Could the site physically deliver the claimed activity?
Check:
- staffing;
- facilities;
- storage;
- operating hours;
- equipment;
- supply chain;
- local demand;
- and historical volume.
Do the records agree with each other?
Compare:
- beneficiary list;
- attendance;
- activity totals;
- invoices;
- payments;
- supplier capacity;
- staff records;
- and site inspections.
Does the growth make sense?
Review:
- month-to-month growth;
- comparison with similar sites;
- changes in staffing;
- changes in facilities;
- and changes in supplier volume.
Has anyone checked independently?
Use:
- site visits;
- unannounced inspections;
- participant sampling;
- supplier confirmation;
- bank verification;
- third-party evidence;
- and risk-based audits.
A practical reporting database
A donor-funded or grant-funded programme could maintain linked tables for the following records.
Organisation table
- organisation ID;
- registration;
- office bearers;
- ownership;
- bank account;
- address;
- risk rating;
- conflict declarations;
- verification date;
- review status.
Site table
- site ID;
- location;
- programme;
- capacity;
- staff;
- equipment;
- operating dates;
- approved daily volume;
- inspection history;
- risk status.
Beneficiary or participant table
- controlled participant ID;
- eligibility status;
- programme;
- location;
- reporting period;
- attendance status;
- duplicate check;
- verification status.
Activity table
- activity ID;
- site;
- date;
- service delivered;
- reported quantity;
- responsible staff;
- source evidence;
- reviewer;
- exception status.
Supplier and expenditure table
- supplier ID;
- invoice;
- goods;
- quantity;
- amount;
- payment;
- delivery evidence;
- related-party status;
- programme site;
- reporting period.
Claim table
- claim ID;
- site;
- period;
- reported meals or services;
- amount requested;
- evidence status;
- validation results;
- review decision;
- payment status.
Complaint and exception table
- issue ID;
- source;
- date;
- risk level;
- organisation involved;
- evidence;
- assigned investigator;
- action;
- status;
- resolution.
The aim is not to create the largest possible database.
It is to connect the records that need to agree.
Exception reporting should be visible
A credible programme report should distinguish between:
- submitted claims;
- automatically checked claims;
- independently verified claims;
- claims awaiting review;
- disputed claims;
- rejected claims;
- and corrected claims.
Those statuses should not be merged into one total.
For example, a dashboard could show:
| Claim status | Amount |
|---|---|
| Submitted | $4,200,000 |
| Passed basic validation | $3,600,000 |
| Independently verified | $2,900,000 |
| Under review | $450,000 |
| Held due to exceptions | $180,000 |
| Rejected | $70,000 |
A funder should know whether a number represents a claim or a verified result.
Human review has to be specific
“Reviewed by the programme team” is too vague.
A reviewer needs a defined task.
For example:
- confirm the organisation;
- check the site capacity;
- review the beneficiary records;
- reconcile invoices with reported delivery;
- investigate anomalies;
- check conflicts;
- record the decision;
- and state which evidence was used.
The reviewer should not merely confirm that the form is complete.
They should test whether the claim makes sense.
Could a better data system have prevented Feeding Our Future?
Not by itself.
This was a deliberate fraud involving false documents, shell companies, kickbacks and coordinated misconduct.
A reporting system cannot make dishonest people honest.
The case required:
- active state oversight;
- site inspections;
- financial investigation;
- complaint protection;
- independent verification;
- payment holds;
- audit powers;
- law enforcement;
- and people willing to act on warning signs.
A stronger data and reporting process could have made the scheme harder to scale.
It could have:
- flagged impossible growth;
- identified duplicate or fabricated records;
- connected invoices to supplier capacity;
- highlighted related entities;
- preserved complaint histories;
- tracked unresolved deficiencies;
- separated submitted from verified claims;
- and prevented automatic payment of high-risk claims.
A system supports oversight.
It does not replace it.
Where my work fits
I would not position this as a case that a reporting-workflow consultant could have solved alone.
Fraud investigation, forensic accounting, regulatory enforcement, site inspection and criminal prosecution require the relevant specialists and authorities.
My work fits around the information and reporting route used by legitimate programme teams, nonprofits, contractors and funders.
It can help ensure that:
- reporting inputs are structured;
- source records stay connected;
- exceptions remain visible;
- human review is recorded;
- and reports distinguish between claims and verified delivery.
Data Collection & Intake Systems
A Data Collection & Intake System could help define:
- organisation and site records;
- participant or beneficiary IDs;
- standard activity fields;
- capacity information;
- supplier records;
- required uploads;
- validation rules;
- complaint intake;
- review statuses;
- and risk flags.
The objective is not only to collect a claim.
It is to collect enough context to test whether the claim is plausible and ready for review.
Traceable Evidence Workflow Support
Traceable Evidence Workflow Support could help connect:
- organisation records;
- site applications;
- beneficiary lists;
- activity totals;
- supplier invoices;
- payments;
- monitoring visits;
- complaint records;
- review notes;
- corrective actions;
- and donor-reporting claims.
The Source Traceability Risk Checker can help identify where the route from reported result to source evidence is weakest.
The central question is:
> Can the reported activity be traced back to independent evidence that someone has actually reviewed?
Data Use, Reporting & Communication Systems
A Data Use, Reporting & Communication System could turn structured records into:
- grant dashboards;
- donor reports;
- payment-control views;
- capacity comparisons;
- exception registers;
- unresolved-evidence reports;
- partner performance views;
- corrective-action trackers;
- and board or management packs.
The output should show whether a number is:
- reported;
- validated;
- verified;
- disputed;
- rejected;
- or corrected.
A clean report should not make an unchecked claim look settled.
A practical evidence-workflow example
The UNICEF Zambia evidence-workflow case study involved qualitative research rather than grant reimbursement.
The same source-control principle applied.
A primary contractor needed to process 120 narrative case studies consistently.
The workflow connected:
- case IDs;
- source material;
- coded themes;
- supporting quotations;
- review rules;
- reporting tables;
- and final claims.
A finding could not be treated as supported simply because it appeared in an AI summary or spreadsheet cell.
It needed evidence.
Grant and donor reporting need the same discipline.
A beneficiary total should link to participant records.
An activity total should link to delivery evidence.
An expenditure claim should link to an invoice, payment and real programme activity.
A related nonprofit reporting failure
The Modest Needs Foundation fraud concerned false governance and personal spending from charitable funds.
The two cases failed at different points.
At Modest Needs:
- the organisation presented a board that did not provide real oversight;
- personal spending was hidden within charity finances;
- and public records created the appearance of governance.
At Feeding Our Future:
- fake activity records were used to claim programme reimbursement;
- the sponsor submitted false evidence;
- and weak verification allowed the claims to scale.
Both cases show why a report cannot be separated from the records and people behind it.
What donor-funded teams should learn from this case
The lesson is not that partner reporting cannot be trusted.
The lesson is that trust should be supported by verification.
A stronger programme should be able to answer:
- Does the partner organisation exist and operate as described?
- Can the site deliver the reported volume?
- Do the beneficiaries or participants exist?
- Did the activity take place?
- Do supplier purchases support the claimed delivery?
- Do the records agree across systems?
- Is rapid growth being investigated?
- Are related parties visible?
- Can complaints be made safely?
- Is the subject of a complaint prevented from controlling the investigation?
- Are corrective actions independently verified?
- Can high-risk payments be held?
- Does the report distinguish submitted data from verified data?
- Can each material claim be traced back to source evidence?
A reporting deadline should not be the point at which the organisation begins trying to reconstruct the evidence.
The reporting process starts when the organisation, site, beneficiary, activity and expenditure are first recorded.
FAQ
What was the Feeding Our Future fraud?
Feeding Our Future was a Minnesota nonprofit that sponsored sites participating in federally funded child-nutrition programmes.
A federal jury found that the organisation and participating sites used false records to claim reimbursement for meals that had not been served as reported.
The Department of Justice said more than $240 million was fraudulently obtained and distributed.
Who was Aimee Bock?
Aimee Bock was the founder and executive director of Feeding Our Future.
She was convicted in March 2025 of wire fraud, conspiracy and bribery offences.
In May 2026, she was sentenced to 500 months in prison.
How many meals did Feeding Our Future claim?
At trial, prosecutors proved that Bock and her co-defendants falsely claimed to have served 91 million meals.
One site connected to Safari Restaurant claimed more than 3.9 million meals between April 2020 and November 2021.
How were the meal claims falsified?
The scheme used:
- fake attendance rosters;
- invented names;
- random ages;
- inflated meal counts;
- fabricated invoices;
- shell companies;
- and false sponsor assurances.
Some rosters used names from a random-name website. In another example, a spreadsheet formula generated random ages for supposed child beneficiaries.
Why did they create fake meal records?
The programmes reimbursed organisations based on meals reported as served.
Inflating the number of meals increased the amount of reimbursement that could be claimed.
Feeding Our Future also received administrative fees linked to the programme claims, while employees received bribes and kickbacks from participating sites.
Was Feeding Our Future a normal donor-funded charity?
The money at the centre of the case was federal programme funding rather than ordinary public donations.
The reporting lessons still apply to donor-funded projects, nonprofits, grant programmes and contractors paid according to reported delivery.
Did the Minnesota Department of Education know about concerns?
The Office of the Legislative Auditor found that the department had unresolved concerns, received at least 30 complaints involving Feeding Our Future or its sites and did not always investigate or follow up adequately.
The department disputed the auditor’s description of its oversight and said it had met applicable standards, referred concerns to law enforcement and later strengthened its controls.
Would better reporting software have prevented the fraud?
Not by itself.
A stronger system could have flagged rapid growth, duplicate records, implausible capacity, unusual invoices and unresolved complaints.
Preventing and stopping the fraud also required independent inspections, financial investigation, payment controls, law enforcement and accountable decision-makers.
What is the difference between reported delivery and verified delivery?
Reported delivery is what a partner or programme site says happened.
Verified delivery is activity supported by records and independent checks that give the funder reasonable confidence the work actually occurred.
A donor report should make that difference visible.
What should funders check before reimbursing a partner?
Funders should check:
- organisation identity;
- site capacity;
- beneficiary records;
- activity evidence;
- supplier invoices;
- payments;
- related parties;
- unusual growth;
- complaints;
- and independent monitoring results.
High-risk claims should be held until exceptions have been resolved.
A useful next step
A report should not be judged only by whether every field has been completed.
It should be judged by whether the reported activity can be traced back to real delivery.
Map the full route:
- Partner
- site
- beneficiary
- activity
- supplier
- expenditure
- review
- payment
- donor report
Then ask:
- Which records can be fabricated?
- Which totals are never independently tested?
- Which partners are reviewing their own complaints?
- Where does growth trigger additional scrutiny?
- Can a claim be paid while evidence is unresolved?
- Are submitted and verified results separated?
- Can the final report be rebuilt from the original records?
When that route is weak, the organisation may be reporting activity without proving delivery.
If your programme or donor-funded team needs a clearer route from partner reporting to review-ready evidence and donor outputs, you can send me a short project brief.
Sources
Data Collection & Intake Systems
Collect useful, traceable data from the start through forms, fieldwork tools, public submission portals, partner reporting systems, calculators, and intake workflows.