Evidence and data
Rwanda's poultry statistics disagree with each other by a factor of six, and the 2026 press figures for egg output turn out to be the government's own planning targets reported as achievements. So nothing here is asserted. Each figure names its survey, strategy document or breed manual, and where the sources conflict we say which one we chose and why.
Discipline
Applied to every figure
The national record
Two columns: the figure the Rwandan record actually reports, and the figure the genetics or the manual are rated for. The distance between them is not an opinion about farming — it is the distance between two published documents.
The national record
| Population | Head | Share | Source |
|---|---|---|---|
| Pigs | 1,006,890 | — | NISR, Agricultural Household Survey 2024 |
| Layers | 1,272,783 | 33% of the flock | NISR, AHS 2024, Table 45 |
| Broilers | 240,711 | 6.2% of the flock | NISR, AHS 2024, Table 45 |
| National chicken flock | 3,860,000 | — | NISR, AHS 2024 |
| Eggs produced, survey year Sixty-eight per cent were sold, so the constraint is not demand. | 112,856,811 | 68% sold | NISR, AHS 2024 |
| Measure | Rwanda today | The standard | Gap |
|---|---|---|---|
| Eggs per hen, to end of cycle Rwandan yield is measured to 90 weeks and the breed objective to 100, so the gap understates nothing — those ten weeks eat feed too. | 360ISA Brown in Rwanda · Netherlands 2019 | 476Hendrix Genetics / ISA objective | −116 a hen |
| Layer flock mortality A national planning baseline against a breed liveability figure. We use the kinder end of the manual's 5–7% range. | 20%MINAGRI Livestock Development Strategy 2024–2029 | 5–7%Lohmann Brown-Classic, liveability in lay 93–95% | 13 points |
| Broiler slaughter age | 42–45 days2.4–2.5 kg at a commercial integrator · Netherlands 2019 | 35 daysCobb 500 at 2,521 g, cumulative FCR 1.441 | 7–10 days |
| Cumulative feed per broiler | 5.100 kgCobb 500 table, day 42 | 3.635 kgCobb 500 table, day 35 | +1.465 kg a bird |
| Broiler stocking density | 5–10 / m²Netherlands 2019 | 4–5 / m²CASA recommendation | up to 2× |
| Pig farms keeping any record | 38%Mbuza et al., Int. J. Livestock Production, 2016 | 90% by 2028/29PSTA 5 compulsory registration target | 52 points |
Market signals
Rwandan food inflation is the mildest major component in the basket. Everything a farm buys is not. You are absorbing energy at six times the rate you can pass on at the farm gate.
NISR consumer price index, June 2026 · exchange rate from a daily public feed · rings drawn proportionally on a 0–50% scale · curated 1 August 2026
Rwandan maize is rising 14.5% year on year while the world price falls 8.8% month on month. The local feed market is not tracking the global one, so a hedge read off an international benchmark is a hedge against the wrong thing.
RWF 560 per kilo, national average across 26 markets · FEWS NET / e-Soko, May 2026
The practical consequence: the only reliable forward view of your feed cost is your own purchase record, priced per ration, per house, per week.
| Signal | Latest | Change | Source |
|---|---|---|---|
| Maize, national average across 26 markets The single biggest line in a poultry ration, and the one that sets your feed invoice. | RWF 560/kg | ▲ 14.5% y/y | FEWS NET / e-Soko, May 2026 |
| Maize, five-year move | — | ▲ 139% | FEWS NET / e-Soko |
| Maize, world price Falling, while Rwanda's rises. The local feed market is not tracking the global one. | USD 197.1/mt | ▼ 8.8% m/m | World Bank Pink Sheet |
| Soybean meal, world price The protein side of the ration, and it is still climbing while maize falls. | USD 425/mt | ▲ rising | World Bank Pink Sheet |
| Rwandan franc against the dollar Premix, vaccines, day-old parent stock and equipment are all imported and priced in dollars. | 1,472.39 | ▲ 51.4% since 2020 | Market mid-rate, 1 Aug 2026 · not the BNR official rate |
| Agriculture's share of all bank lending Nineteen per cent of GDP, financed by under three per cent of the credit. A lender who cannot read your record cannot price your risk. | 2.7% | — of the loan book | NISR / BNR |
One number is missing here on purpose. There is no current published Rwandan egg price. The WFP series ends in 2015 and FEWS NET carries no eggs at all — so the thing a layer farm actually sells is, nationally, unmeasured. Every farm we have worked with knows its own gate price to the franc and can produce it in seconds; the country cannot. That asymmetry is the whole argument for keeping the record, and it is why we would rather show you an absence than an invented figure.
What we do not claim
We went looking for evidence that record-keeping software raises output. The best-controlled study we could find says it does not — and we would rather publish that and argue the other side of it than quote a number we cannot stand behind.
912 dairy herds, software users against non-users. First calving came two and a half months earlier, and the modelled marginal return on the software was 163%. It found no significant difference in yield.
Journal of Dairy Science, Costa RicaRecords do not make an animal produce more. They make the losses visible — and in African smallholder poultry those run to half the flock, with three-quarters of brooding chicks lost. A Kenyan study of 1,908 households found 70% of all off-take was involuntary loss, and 12% had reached a vet.
FAO · Kenya household surveyReading a photographed tally sheet, mapping an unfamiliar spreadsheet column, and answering a question in plain language. It never invents a figure: every answer opens the screen and the rows it came from. We publish no claim that AI raises output, because we found no evidence for one.
A bird that dies has eaten its feed and produced nothing — the emissions of the whole flock, divided across fewer eggs. Commercial layers convert 2.3 kg of dry matter per kg of egg; backyard flocks need 9.2. Cutting loss is the cheapest decarbonisation a farm has, and it happens to be the same work as cutting cost.
FAO GLEAM · Poore & Nemecek 2018Method
Every figure traces to a survey, a strategy document, a breed manual or a public price feed. If we could not find one, the figure is not on the site — which is why some obvious-looking numbers are absent.
Rwanda's poultry statistics disagree by a factor of six depending on the source. Where we had to choose, we say which we took and why — and we take the conservative side of any range that flatters our argument.
Two inputs in the gap calculator are ours rather than cited — the feed price and the days of downtime between broiler placements. Both are marked, both are editable, and the legend names them.
The 2026 press figures for Rwandan egg output turn out to be the government's own planning targets, reported as results. They are not used anywhere on this site.
Market figures are curated, not live, and say so. Where a figure can be refreshed from a public feed on a schedule, it will be — and it will still carry the date it was last read.
| Source | Cadence |
|---|---|
| Exchange rate | Daily public feed |
| World Bank indicators | Weekly |
| World Bank Pink Sheet | Monthly |
| FEWS NET market prices | Monthly |
| NISR surveys and CPI | As published |
| Breed manuals | On revision — rarely |
We would rather correct a number than defend it. If you hold a source that contradicts something on this page — particularly a current Rwandan egg price, which we could not find — we want to see it.