Development financing is tightening from every direction at once: bilateral aid is down, philanthropic portfolios are being re-scoped, and every allocation that survives has to answer a harder question than “what did we fund?”. It has to answer “what did it achieve — and how do we know?”
Impact accounting answers it without over-claiming: how the sector moved, what the funder plausibly contributed, and what it cost — with an evidence grade on every line.
The Squeeze, in Numbers
Official development assistance from members of the OECD’s Development Assistance Committee fell by 23.1% in real terms in 2025, to USD 174.3 billion. Preliminary OECD data call it the largest annual contraction on record. Bilateral aid to sub-Saharan Africa fell by about a quarter, and the OECD projects a further 5.8% decline in 2026.
Philanthropy cannot fill a gap of that size, and it is not trying to. Funders are choosing instead: which programs to protect, which to hand over, which to close. Each of those choices rests on a claim about results.
Allocation Is Not Impact
Even money that is allocated does not always move. In Nigeria, states and the Federal Capital Territory accessed ₦45.4 billion of the ₦131.5 billion allocated to the Basic Health Care Provision Fund in 2024, according to BudgIT and national agency data reported by BusinessDay.
The lesson for funders is uncomfortable but useful. A portfolio report that counts disbursements, trainings and activities describes effort. It does not describe what changed for people — and when budgets tighten, effort alone is hard to defend.
“Contributed to” is a stronger sentence when it survives scrutiny than “caused” when it does not.
What a Defensible Account Looks Like
Impact accounting asks three questions, in this order.
- How did the sector move — with or without us? National surveys and routine data set the baseline. When sources disagree, a defensible account shows both and says why, rather than choosing the convenient number.
- What did our investments plausibly contribute? Contribution analysis, set out by the evaluator John Mayne, builds a reasoned case for contribution where strict attribution is out of reach. Each claim carries an evidence grade: independently evidenced, verified by the program officer, or self-reported.
- What did it cost? Linking spending to the results it supports turns “value for money” from a slogan into a line in the account.
Every contribution claim carries one of three evidence grades, from independently evidenced to self-reported.
Source: Mayne (2008), Contribution analysis: an approach to exploring cause and effect, ILAC Brief 16.
Looking forward, the same account can carry scenarios — conservative, central and best — to a long horizon, so leaders see a range rather than a single promise.
None of this replaces an evaluation. An evaluation tests whether one program caused one outcome. An account shows what a whole portfolio is contributing to, every year, at a cost a country office can sustain. It is built for the people who decide: a country director preparing a strategy review, a program officer defending a renewal, a board asking what a decade of grants achieved.
Four Disciplines That Keep It Honest
In our work with foundation country offices in Africa, four rules have mattered more than any tool.
- Separate evidence from expectation. Evidence, working hypothesis, expectation and recommendation each get their own label, so readers can tell which is which.
- Grade every claim. A self-reported result is still useful — as long as it is marked as self-reported.
- Show disagreement. Survey and administrative figures often differ. Hiding that weakens the account; showing it builds trust.
- Never claim attribution you cannot prove. “Contributed to” is a stronger sentence when it survives scrutiny than “caused” when it does not.
AI Changes the Cost, Not the Standard of Proof
Our first accounts took the better part of a year per cycle. The 60 grant-level accounts we completed by hand now train the models that draft the next ones: rules-based automation refreshes the data, AI writes the first pass, and a named consultant reviews every claim before it reaches the client.
That changes the economics of accountability. Accounting for a whole portfolio every year — not a sample, not once a decade — becomes affordable. What does not change is the standard: the evidence grade, the visible disagreement, the refusal to over-claim.
What This Means for You
Five questions to put to your own portfolio before the next budget conversation:
- Can we say how each priority sector moved this year, from sources others would accept?
- Which investments plausibly contributed — and on what grade of evidence?
- Where do our sources disagree, and do we show it?
- What did each result cost, and what would we stop funding if we had to?
- Could a program officer defend the account, line by line, without us in the room?
Defensible is not the same as positive. In a tighter funding era, an honest account of what did not work is what protects the investments that did.
Talk to One of Our Experts
Get in touch today to find out about how Evalueserve can help you improve your processes, making you better, faster and more efficient.


