A seed certified, a fertiliser subsidised and a loan disbursed all amount to little if the farmer does not know what to do with them. Extension — the unglamorous business of getting agronomic advice into the field — is the silent determinant of yields, and it is also where Africa’s dual agricultural economy shows most starkly. South Africa proves the point inside its own borders: its extension is strong where the farms are large and commercial, and thin where they are small and poor. The knowledge gap is not between countries first; it is between farmers.
The Anchor: A Tale of Two Farm Sectors
South Africa’s agricultural extension is markedly stronger in the commercial sector than for smallholders — a structural feature that mirrors the country’s broader agricultural divide. Large commercial operations access private agronomists, input-supplier advisers, and well-resourced support; emerging and subsistence farmers rely on a public extension service that is stretched thin across vast rural areas, with advisers responsible for far more farmers than they can meaningfully reach. The advice exists; its distribution is unequal.
That is the diagnostic lesson. Extension does not fail by being absent; it fails by being concentrated where it is least needed.
Takeaway: The problem is rarely a shortage of knowledge — it is a shortage of delivery to the farmers who lack it.
The Comparator: East Africa’s Delivery Experiments
Kenya, Ethiopia and Zambia are each wrestling with the same delivery problem and trying different routes around it. Kenya has been a continental testbed for digital extension — mobile advisory services, SMS agronomy and app-based pest diagnosis aimed at the millions of smallholders no field officer will ever visit in person. Ethiopia took the human-infrastructure route, building one of Africa’s largest networks of development agents and farmer training centres to push a standard package of practices into the countryside at scale. Zambia has leaned on farmer field schools and the lead-farmer model, where trained smallholders demonstrate techniques to their neighbours, multiplying a handful of advisers into a web of peer instruction.
The continental research catalogued by the Food and Agriculture Organization and the Alliance for a Green Revolution in Africa shows each model’s trade-off: digital extension scales cheaply but struggles with trust and the last agronomic mile; human networks build trust but cost a fortune to staff.
Takeaway: Every extension model trades reach against trust — and the field needs both.
The Mechanism: Why Advice Has to Be Local and Repeated
The reason extension is hard is that good agronomic advice is specific — to the soil, the season, the crop variety and the farmer’s cash position — and it must be repeated, because practices change behaviour only when reinforced. A futures price or a fertiliser bag travels well; agronomic judgement does not. The successful systems combine channels: digital tools for timely, low-cost reminders and alerts, layered over human or peer contact that supplies trust and troubleshooting. World Bank rural-development data, accessible through its agriculture and rural development portal, consistently links closing the yield gap less to new technology than to whether farmers actually adopt practices already known to work.
Takeaway: Extension works when it is local, repeated and trusted — not merely when it is delivered.
The Verdict: A Gap No Single Channel Closes
South Africa’s commercial extension is a model worth copying for its professionalism; its smallholder extension is a cautionary tale of under-resourcing that Kenya, Ethiopia and Zambia are not waiting to repeat. None of the comparators has fully solved the problem, but their experiments collectively point the way: blend digital reach with peer-based trust, and resource the public service for the farmers the private one ignores. In smallholder digital extension, in fact, East Africa is ahead of South Africa — another honest inversion.
Takeaway: The future of extension is hybrid — a phone for reach, a neighbour for trust, an agronomist for judgement.
What must be in place is concrete: extension officers resourced to a workable farmer ratio, a digital advisory layer tied to local agronomy rather than generic tips, and a peer-multiplier model that turns lead farmers into the system’s last mile. That is the forward action for any agriculture ministry or agritech founder. And it confirms the series thesis once more — South Africa is the continent’s most complete agricultural template, but on reaching the smallholder with knowledge, it is as much student as teacher.
