Follow the money past the budget line.
Others measure what states promise to spend on agriculture. ABX measures what they actually spend, what it buys, and how each state compares with its peers.
One budget line, four numbers, and a field visit
Most commentary reports one number: the allocation. ABX records four for every state, year and line, then checks a sample on the ground.
ABX counts agriculture by function, not by ministry, following the international COFOG standard used for CAADP reporting. Irrigation, rural roads, input agencies and counterpart funds for donor programmes often sit outside the Ministry of Agriculture, so reading the ministry line alone undercounts.
Where agriculture money is lost, who controls it, and what to ask
Tap each stage. Every leak has a responsible office, a document that should reveal it, and a question citizens and lawmakers can put to government.
Three kinds of pressure, because one is not enough
Rankings move the states that already care about their reputation. ABX also works through oversight institutions and through the farmers who bear the cost.
Governors and commissioners
An annual index and zonal league tables, so states compete on delivery as a mark of good governance.
Houses of Assembly and Auditors-General
Evidence packs delivered before each state's budget defence, so spending becomes a standing oversight question.
Farmers, cooperatives and media
Scorecards in English and Hausa, radio, and a role in verification, so farmers can notice and contest non-delivery themselves.
Timing is the mechanism. Evidence delivered six weeks before a budget defence is worth more than the same evidence at a launch event. ABX's publication calendar follows each state's budget calendar. See the budget calendar.
How a number earns a place on this site
The ABX Methodology Paper is published before any ranking and is open for peer review. Anyone should be able to reproduce an ABX figure from the same public documents.
Capture
Record the document, page, and exact words or row. Nothing typed from memory.
Classify
Mark each figure reported or derived, and grade each state-year from A (audited) to E (unavailable).
Derive
Show the formula and inputs for every computed figure. Missing data is named, never estimated.
Reply
Send findings to each state 21 days before publishing. Publish the reply, or the silence.
Five components, scored against fixed standards
Weights are a judgement, so ABX publishes how rankings move when they change. Move the sliders to see for yourself.
| Component | Indicator | Weight | Why it matters |
|---|---|---|---|
| Allocation adequacy | Share of the revised budget going to agriculture | 20% | A state should not get credit for money it cut during the year |
| Release discipline | Capital release rate | 25% | Withheld releases are the failure most directly in government's control |
| Capital execution | Capital spent ÷ revised capital budget | 25% | The core measure of whether investment happened |
| Disclosure | Timeliness and detail of quarterly reports | 15% | Rewards states that make themselves measurable |
| Verified delivery | Share of sampled projects passing all four field tests | 15% | Separates spending from delivery |
Also reported, not scored: the Agriculture Expenditure Share (the Kampala test), budget credibility deviation, counterpart compliance, and real spending per farming household. Request the Methodology Paper.