- Design a proportionate internal control system (COSO, three lines model, segregation of duties) for a multi-donor NGO.
- Prevent, detect and handle fraud and corruption, including aid diversion, with a rigorous investigation procedure.
- Prepare for a donor audit, understand ineligible costs and manage risks linked to sanctions and counter-terrorism.
Keep the course at home: free PDF handout with exercises, key takeaways and quiz answers.
1Internal control: principles and architecture
The COSO framework (2013) defines internal control as a process that provides reasonable assurance on three objectives: effective operations, reliable financial information, and compliance with laws and contracts. It rests on five components: control environment (tone at the top, ethics, competence), risk assessment, control activities, information and communication, monitoring.
The Institute of Internal Auditors' three lines model (2020) divides the roles: the first line (operations: programmes, logistics, field finance) manages risks day to day; the second line (compliance, risk management, financial control) sets the rules and checks them; the third line (internal audit) provides independent assurance to the board.
Key controls for a field-based NGO:
- Segregation of duties: the person who requests a purchase is not the one who approves it, receives the goods or pays.
- A written delegation of authority, with signature thresholds by amount and by type of expenditure.
- Monthly bank reconciliations signed by someone who handles neither cash nor the bank.
- Surprise cash counts, physical inventories and stock control.
- Verification of beneficiaries and payment lists, preferably by a team independent from targeting.
In a small base, full segregation is impossible. Compensate with detective controls: monthly review of supporting documents by the next level up, job rotation, surprise remote checks.
2Understanding fraud: triangle, schemes and red flags
The fraud triangle (Cressey) combines three conditions: pressure (debts, family pressure, unrealistic targets), opportunity (weak controls) and rationalisation ("I'm underpaid", "everyone does it"). Management acts mainly on opportunity, but also on rationalisation through culture and example.
The most frequent schemes in humanitarian work:
| Area | Scheme | Red flags |
|---|---|---|
| Procurement | Collusion between suppliers, splitting orders to stay under thresholds, fictitious suppliers, bribes | Always the same bidders, bids with very close prices, identical addresses or phone numbers, orders just below the threshold |
| Beneficiaries | Fictitious or duplicate beneficiaries, inclusion in exchange for payment, taxation of aid by intermediaries | Duplicate names or phone numbers, complaints from excluded households, gaps between lists and actual attendance |
| Human resources | Ghost employees, fictitious overtime, jobs in exchange for payment | Staff never seen, shared bank accounts, complaints from candidates |
| Stock and fuel | Theft, diversion, abnormal consumption | Repeated inventory discrepancies, abnormal consumption per kilometre |
Aid diversion by armed groups or authorities (taxation at checkpoints, imposed lists, levies on beneficiaries) is a major risk in conflicts. It is managed through risk analysis, independent verification, suspension where necessary, and coordination with other organisations so as not to give in separately.
3Handling a suspicion: reporting, investigation and sanctions
An anti-fraud policy must provide for: several reporting channels (including one outside the line management and anonymous), protection of whistleblowers against retaliation, a rule of zero tolerance for inaction (every credible suspicion is handled), and the obligation to notify donors within contractual deadlines, often very short (for example within a few working days of becoming aware of a credible suspicion).
Steps of an internal investigation:
- Initial assessment: is the report credible and within the scope of the policy? Who must be informed? Is there a risk to anyone's safety?
- Preserving evidence: copying data, securing documents, suspending access if necessary, without alerting the person implicated.
- Investigation by trained people, independent from the line management concerned, with an investigation plan, interviews with witnesses, then with the person implicated, who has the right to respond.
- Report based on facts, with an explicit standard of proof (in disciplinary matters, generally the balance of probabilities).
- Decisions: disciplinary sanctions, recovery of losses, reporting to the authorities if the law requires it or if appropriate, and above all fixing the control weaknesses.
Serious mistake: entrusting the investigation to the suspected person's direct manager, or starting interviews before securing the evidence. The first compromises independence; the second makes documents disappear.
4Donor audits and ineligible costs
A project audit checks that the expenditure declared is real, eligible (provided for in the budget, incurred during the period, compliant with the rules), justified by supporting evidence, recorded in the accounts and compliant with procurement and visibility rules. Any expenditure that does not meet these conditions becomes an ineligible cost that the organisation must repay from its own funds.
The most frequent causes of ineligible costs: procurement procedure not followed (number of quotations, advertising, thresholds), missing or non-original documents, expenditure outside the period, unjustified allocation keys, wrongly calculated exchange differences, and assets that cannot be found in the inventory.
Preparing for an audit in six points: a file per contract with all the documents in the order of the procedure; reconciliation between the financial report and the general ledger; a list of staff charged with contracts, timesheets and payslips; an inventory of tagged assets; a register of exchange differences; a note explaining the allocation keys.
Worked example. An auditor tests a sample of USD 120,000 of expenditure out of a total declared of USD 1,000,000 and finds USD 9,000 of errors (7.5%). Depending on the agreed method, some auditors extrapolate this rate to the whole population: USD 75,000 of potentially ineligible costs. A systematic error, such as an allocation key, therefore costs far more than its amount in the sample.
5Sanctions, counter-terrorism and partner vetting
NGOs working in areas controlled by designated groups must comply with sanctions regimes (UN Security Council, European Union, United States, United Kingdom) and with the counter-terrorism clauses in their contracts. These clauses prohibit making funds or resources available to listed persons or entities, directly or indirectly.
Security Council resolution 2664 (2022) introduced a cross-cutting humanitarian exemption to UN asset-freeze regimes for humanitarian assistance and activities that support basic human needs, and several jurisdictions have adopted comparable exemptions. But these exemptions do not remove the need for due diligence or the donors' contractual clauses.
Due diligence includes: screening partners, suppliers and key staff against sanctions lists before contracting and periodically, a diversion risk analysis, controls on payments (no cash payments to unvetted intermediaries), and a written policy for managing the contacts with armed groups that are necessary for access.
Ethical point of attention: according to most donors and sector positions, screening does not apply to beneficiaries, because aid is based on need. Screening beneficiaries against lists would violate the principle of impartiality; in case of contractual doubt, the question must be settled in writing with the donor.
Exercise: the red file
Situation. During a review, the country finance manager notices that over 14 months, the same building materials supplier has won 11 out of 12 contracts at the Kalumbi base, that 7 orders are between USD 4,800 and 4,990 while the threshold for a restricted tender is USD 5,000, and that two of the three competing quotations carry the same phone number. The base logistician approved all the evaluation minutes. The main donor requires notification within 5 working days of any credible suspicion.
Task. 1) Identify the red flags and the probable scheme. 2) List, in order, your first ten actions over 72 hours. 3) State who must investigate and who must not. 4) Write the notification to the donor (8 lines, factual, without accusing anyone). 5) Propose three control improvements.
Answer key. Probable scheme: order splitting and collusion or fake quotations. Actions: secure the procurement files and data (copies), discreetly restrict the logistician's procurement rights, inform the anti-fraud focal point and country management, assess the security risk, notify the donor within the deadline, appoint an investigation team independent from the base. Improvements: automatic control of cumulative orders per supplier and per period, supplier vetting (address, registry, phone), a bid opening committee with at least one member from outside the base.
Key points
- COSO: control environment, risks, control activities, information, monitoring; three lines: operations, second line, internal audit.
- Fraud triangle: pressure, opportunity, rationalisation.
- Investigation: secure the evidence before any interview; independent investigators; notify the donor within the deadline.
- A systematic error can be extrapolated by the auditor to the whole population of expenditure.
- Sanctions: due diligence on partners and suppliers; no screening of beneficiaries against lists.
💼 Programme management jobs
Put this course into practice: the latest jobs in this field.
- Consultant- Technical Expert (STEM)Centre for microFinance · India
- ПерекладачДанська Рада у справах біженців · Ukraine
- Фахівець/чиня із залучення спонсорства / National Sponsorship CoordinatorSOS Children's Villages Ukraine · Ukraine
- Social WorkerRazi Social Development Organization (RSDO) · Afghanistan
- Assistant Manager - ImpactMagic Bus India Foundation · India
Further reading: course references
- ALNAP (M. Buchanan-Smith and K. Scriven) (2011). Leadership in Action: Leading Effectively in Humanitarian Operations. www.alnap.org
- PM4NGOs (2020). Project Management for Development Professionals (PMD Pro). pm4ngos.org
- CHS Alliance, Groupe URD and the Sphere Project (2024). Core Humanitarian Standard on Quality and Accountability (CHS). www.corehumanitarianstandard.org
- OECD DAC (2019). Better Criteria for Better Evaluation: revised evaluation criteria. www.oecd.org
- COSO (2013). Internal Control – Integrated Framework. www.coso.org
- Institute of Internal Auditors (2020). The IIA's Three Lines Model. www.theiia.org
- GISF (2020). Security to Go: a risk management toolkit for humanitarian aid agencies. www.gisf.ngo
- IASC (2019). Six Core Principles relating to Sexual Exploitation and Abuse. interagencystandingcommittee.org
- ICRC and Brussels Privacy Hub (2020). Handbook on Data Protection in Humanitarian Action, 2nd edition. www.icrc.org
- R. Fisher, W. Ury and B. Patton (1981). Getting to Yes: Negotiating Agreement Without Giving In. www.pon.harvard.edu
- CCHN (2019). CCHN Field Manual on Frontline Humanitarian Negotiation. frontline-negotiations.org
- R. Rumelt (2011). Good Strategy, Bad Strategy. www.goodbadstrategy.com
- J. P. Kotter (1996). Leading Change. www.kotterinc.com
- IASC (2016). The Grand Bargain. interagencystandingcommittee.org/grand-bargain
Weekly quiz
5 demanding questions to check your learning. Answer them all to complete the week.