A drought season is arriving across southern Africa while East Africa braces for heavy short rains, and the question for every cooperative and ministry is the same: what do we do when we cannot predict exactly how the year will unfold? Two pieces you shared offer complementary answers. IFAD’s new report on food systems under stress, released on the opening day of COP31, finds that farmers lose less income when inputs, markets and finance arrive together. And political economist Yuen Yuen Ang argues that development rarely follows a copied template: it works when central direction is combined with local adaptation and the scaling up of what works. This article puts the two side by side and tests them against African evidence, from Zimbabwe’s small plot farming programme to Kenya’s county climate grants.
Read this first. We read the Down To Earth report on the IFAD publication in full but not the IFAD report itself, so its figures are as reported there. The Ang interview is a conversation about China, the United States and Africa; we use only her general argument and her remarks on Africa, and we draw the African examples ourselves. The Bank for International Settlements release you pasted is a 2017 macroeconomic statement with no direct bearing on African farming, so we set it aside. Other evidence comes from published African studies, mostly before 2026. Research stops at 13 November 2026, and the El Niño season is still developing.
Six findings
- The season is a stress test. IFAD says El Niño could peak between October 2026 and January 2027, with southern Africa facing a more than 50% probability of agricultural drought in five countries.
- Combinations beat single fixes. IFAD’s modelling cuts farm income losses by 18.7% when inputs, markets and finance are combined, against 9.8% for finance alone.
- The last drought was severe. In 2024 Zimbabwe’s maize output was projected 55% below normal and Malawi’s 45%.
- Directed programmes can work and still disappoint. Zimbabwe’s small plot scheme raised yields in studies, but with waterlogging, labour burdens and mixed site results.
- Local adaptation needs money and rules. Kenya’s counties choose climate projects under performance based grants.
- Metrics shape what we fix. Counting fertiliser kilograms misses that 63% of sampled Kenyan soils were too acid.
The season in view
IFAD and Türkiye’s agriculture ministry published Food Systems Under Stress: Building Resilience from the First Mile on 9 November. It says the 2026 to 27 El Niño develops against rising temperatures and follows an energy price shock linked to the Strait of Hormuz crisis, which pushed up fertiliser and transport costs for small scale farmers within a single planting season. Botswana, Namibia, South Africa, Zambia and Zimbabwe face a probability above 50% of agricultural drought between October 2026 and March 2027, threatening maize, pasture and water. The report cites World Food Programme estimates of 49 million more people acutely food insecure by the end of 2027, and a global income loss of about $11 trillion over five to six years.
Months, October 2026 to December 2027. Windows from the IFAD report as summarised by Down To Earth; Kenya’s short and long rains as in the national outlook and the forecast pilot; dashed line marks the start of 2027
What the last drought did
Southern Africa has just lived through this. In 2024, Malawi, Zambia and Zimbabwe declared disasters. Zimbabwe projected maize output 55% below normal and Malawi 45%. In Zambia, according to Dialogue Earth, 84 of 116 districts were affected and almost half of the 2.2 million hectares of maize was destroyed. In South Africa, which accounts for about 70% of regional maize, USDA estimated 2023/24 production at 14.5 million tonnes, 15% down, with yields 17% below expected, and exports falling 44.4%. Estimates of the people affected vary: SADC’s executive secretary spoke of 68 million at risk of hunger, while another report put the number affected at more than 61 million. World Weather Attribution found the drought was driven mostly by El Niño rather than climate change.
Per cent change. Zimbabwe and Malawi: 2024 production projections reported by The Outlier. Zambia: share of planted maize destroyed (Dialogue Earth), drawn at 50% for “almost half”. South Africa: USDA via IFPRI, 2023/24. Bases differ, so read each bar on its own
Why combinations beat single fixes
IFAD’s modelling looks at how drought and rising input costs interact for an individual small scale farmer. Insurance and other financial measures alone cut income losses by 9.8%, while combining support for inputs, market access and finance cuts them by 18.7%. IFAD’s president says investments that connect the three have increased incomes by more than 40%, with technology adoption rising by about three quarters, which are IFAD’s own figures from its programmes.
Reduction in farm income losses from drought and higher input costs. IFAD modelling as reported by Down To Earth, 2026
Earlier African trials point the same way, though not uniformly. A randomised experiment in Ethiopia found that only a package of insurance, credit and inputs significantly raised investment in modern technology. A preliminary trial of One Acre Fund’s programme, which bundles input loans, insurance and training, found yields up 32% and profits up 23%, and that relieving information constraints alone was not enough. But an evaluation in Senegal found no evidence that insurance bundled with loans raised investment or incomes overall. Combinations help when each piece fits local conditions, which is Ang’s point about direction without a template.
Direction and improvisation: Zimbabwe’s small plots
Ang describes China’s growth as directed improvisation: government sets a direction, defines incentives and scales up what works, rather than planning every detail. She is not arguing that Africa should copy China; she says copying is the wrong lesson and that today’s world of weaker globalisation and more unknowns calls for policy built around uncertainty. A Zimbabwean programme shows both the promise and the strain of direction from above. Pfumvudza (or Intwasa in Ndebele) was launched in 2020/21, giving farmers free maize seed, fertiliser and lime if they establish small conservation plots with minimal soil disturbance, mulching and legume rotation.
The evidence is mixed. A June 2021 evaluation for FAO and the Indaba Agricultural Policy Research Institute estimated that the approach raised maize yields by more than 1,500 kg per hectare, but found that waterlogging under mulch cut yields and that more plots per household lowered yields. Ian Scoones’s field comparisons at four sites, which he calls very approximate, show Pfumvudza plots ahead at two sites and behind at two. A 2024 survey of 408 households in Mashonaland East found beneficiaries less likely to grow labour hungry cash crops, and reported that income effects were limited. A national survey found 86% of participating households satisfied, though satisfaction is not yield.
Approximate maize yield, tonnes per hectare, one season at four sites. Ian Scoones, Zimbabweland blog; figures described by the author as very approximate. Pfumvudza plots are 0.06 ha, so total output is not comparable
Improvisation at scale: Kenya’s counties
Kenya offers a different version. Its county climate programme gives performance based grants to counties that build climate funds, units and ward level risk assessments, then lets wards choose local projects, and the Treasury reports it has reached 1,137 of 1,450 wards. The Kenya Meteorological Service Authority’s forecast pilot, with the Alliance of Bioversity and CIAT and AIM for Scale, uses the short rains as a proof of concept before wider dissemination in the March, May 2027 long rains. In both, the centre sets direction and funds experiments, and what works is meant to scale. We map them onto Ang’s loop below; the mapping is ours, not hers.
Our mapping of Zimbabwe’s Pfumvudza, Kenya’s county climate grants and its forecast to farmer pilot onto Ang’s “directed improvisation”; this is an interpretation, not a finding of the studies
Using what you have: informal channels and groups
Ang closes her book with Nollywood, which grew in a country with weak institutions by working with what existed, including pirates who doubled as informal distributors. The agricultural parallel is the informal trader. The World Bank’s September 2026 Kenya diagnostic reports that about 80% of mango output flows through informal brokers and aggregators, with postharvest losses up to 35%, so informal channels are both the route to market and a loss point. Savings groups, which CARE says involve around 500 million people in Africa and Asia, are another asset. The pragmatic question is how to make these existing channels more reliable, with records and shared tools, instead of replacing them. Ang says the instinct to copy comes partly from a colonial history of imposed models, and that “using what you have” resonates strongly among Nigerian readers.
Unknowns, not winners
Ang warns that in a world of unknowns planners cannot pre select winners, and notes that some of China’s most successful sectors, e commerce and fintech, were supported after they emerged. She says African economies have seen leapfrogging in services that is neither a Chinese nor a Western story. Farming has examples. M-PESA, which nobody planned as an agricultural tool, lifted an estimated 194,000 Kenyan households out of poverty, according to a Science study. A meta analysis of African intercropping found an average land equivalent ratio of 1.45, meaning mixed plots could produce the same output with about 45% less land than separate crops need by our reading. The lesson for a drought year is to hold options open, with several crops, several finance routes and several advisory channels, and to expand what proves itself.
Who defines the metric
Ang argues that categories and indices are shaped by power and carry biases we should question. African agriculture shows how. Africa’s fertiliser use of about 18 kg of nutrients per hectare, against a 50 kg Abuja target, frames the problem as a shortage of volume. Yet a KALRO survey of 231 Kenyan sites, as reported by the Nation, found 63% below pH 5.5, where nutrient uptake falters, so more fertiliser on acid soil may do little. SIANI’s review of agrobiologicals makes a similar point: biological products are often judged by registration frameworks designed for chemical pesticides, adding cost and delay. Who sets the measure shapes what gets funded.
Left: kilograms of fertiliser nutrients per hectare, African Union and summit reporting. Right: soil acidity at 231 Kenyan sites, KALRO survey as reported by the Nation (2026); the 37% share is our arithmetic
Where verified field data fits
Every approach above, whether it is bundled finance, county grants, forecast advice or biologicals, depends on knowing what happens on actual plots. AgriKonnekt is built to check farm plots with sensors and satellites so cooperatives can show results to funders, according to the AgriKonnekt page. We have not seen it used with these programmes, so any role in feeding results back into a direction and adaptation loop is our proposal.
What to do next
- Southern African cooperatives should plan now for a more than 50% drought risk: drought tolerant seed, water and soil conservation, and early finance and insurance arrangements.
- Programme designers should combine inputs, markets and finance, and test the combination locally before scaling, since trials show mixed results.
- Governments that direct farm programmes, like Zimbabwe’s, should publish plot level results, including waterlogging, labour burdens and failures, and adjust the design.
- County and ward planners should use performance based local grants for water, soil and advisory projects, as Kenya’s counties do, and protect the budget lines.
- Cooperatives should record informal channels, from brokers to savings groups, so they can be made more reliable.
- Analysts and funders should question headline metrics, such as fertiliser kilograms, and add diagnostics like soil tests.
What we can and cannot claim
| Statement | Evidence status |
|---|---|
| Combined support cut income losses by 18.7%, against 9.8% for finance alone | IFAD modelling as reported by Down To Earth; report not read |
| Five southern African countries face above 50% drought probability | IFAD report as summarised; forecast could change |
| 2024 maize output fell 55% in Zimbabwe and 45% in Malawi | Projections reported at the time, not final harvests |
| Pfumvudza raised yields by over 1,500 kg per hectare | 2021 evaluation for FAO and IAPRI; Scoones finds mixed site results |
| Kenya’s county programme reached 1,137 of 1,450 wards | Treasury figures as reported by the press; not audited |
| Informal brokers handle about 80% of Kenyan mango | World Bank diagnostic citing ministry data |
| Directed improvisation maps onto these African programmes | Our interpretation of Ang’s concept |
A drought year does not reward certainty so much as the ability to adjust. Ang’s argument that development grows from direction, local improvisation and scaling what works fits what African evidence shows: IFAD’s combined investments cut losses more than finance alone, Zimbabwe’s small plots help in some places and strain in others, and Kenya’s county grants let wards choose what suits them. None of it works without honest, plot level evidence of what happened, shared with the people who farm the land, and without questioning the metrics that decide what counts as a problem. The cooperatives, counties and ministries that keep that evidence, and adjust as the season unfolds, will be best placed to turn the next El Niño from a shock into a managed risk.
Sources & notes
SOURCES: Vimi Wang and The Monitor, Who Gets to Explain Development? Yuen Yuen Ang on how power shapes categories (4 August 2026); Shagun, COP31: El Niño could cause $11 trillion in global income losses, IFAD warns, Down To Earth (11 November 2026); The Outlier, southern Africa maize harvests; Dialogue Earth via PreventionWeb, El Niño triggers food crisis in southern Africa; IFPRI, southern Africa drought and maize; allAfrica on SADC and the 2024 drought; IAPRI and FAO, Impact of Pfumvudza on maize productivity (2021); Ian Scoones, Zimbabweland, on Pfumvudza; The Conversation via allAfrica, Zimbabwe farming subsidies study; The Herald on ZimStat and World Bank survey; Marr et al., Ethiopia trial; Deutschmann et al., One Acre Fund (preliminary); 3ie, Agricultural Insurance Evidence Program; Suri and Jack, Science, 2016; Kenya Treasury, FLLoCA; Alliance of Bioversity and CIAT, forecast to farmer pilot; AgriKonnekt.
Differences between sources: people affected by the 2024 drought are 68 million at risk (SADC) or more than 61 million affected (another report); Zambia’s loss is stated as share of planted area, not tonnes; Pfumvudza yields range from a gain above 1,500 kg per hectare to losses at some sites.
Our own arithmetic: the 37% share of Kenyan soils between pH 5.5 and 7; the reading of the land equivalent ratio as about 45% less land.
Not read: the IFAD report itself, Ang’s books, the BIS annual report. Programme figures are as stated by their sources.