Galana-Kulalu and the Myth of Mega-Farm Salvation

A million acres in Tana River and Kilifi.

Two decades of presidential commissions, master plans, and ribbon cuttings. A grand harvest that mostly stayed on PowerPoint.

Galana-Kulalu was supposed to end Kenya's food insecurity. Instead, it became Kenya's most expensive lesson in why mega-farms don't feed countries.

It is also the lesson we keep refusing to learn — because the next mega-farm proposal is already in someone's drawer.

Table of Contents

• What Galana-Kulalu Was Meant To Do
• Why Mega-Farms Keep Failing
• Why Smallholders Feed Kenya
• What Better Public Investment Looks Like
• The Pattern To Watch
• Why I Wrote This

What Galana-Kulalu Was Meant To Do

The Galana-Kulalu Food Security Project, launched in 2014, allocated about 1.75 million acres for irrigated agriculture across two counties. The pitch: turn semi-arid land into a maize bowl. Use the Tana River. Apply mechanisation. Solve the national food deficit in a decade.

The numbers were always astonishing. Eighty bags of maize per acre. A target of 40,000 acres under maize at full scale. National food security "achieved." Cabinet decks made the math sing.

Twelve years on, the project has produced a fraction of the projected harvest. The model farm has changed hands repeatedly. The original contractor's role has been investigated and re-investigated. Reports have been commissioned. The project has cost the public purse in the billions.

The question is not whether Galana-Kulalu underperformed. It did, on every measure that matters. The question is why this particular mode of food security thinking keeps failing — in Kenya and across Africa — and why it keeps getting funded.

Why Mega-Farms Keep Failing

Five reasons. None of them are about the soil.

1. The economics of smallholder agriculture are not what cabinet decks assume.

Smallholder farmers in Kenya — the people who actually feed Kenya — produce 70–80% of the country's domestic food. They do this on average plots of 0.4–2 hectares. They do it imperfectly, but they do it under the constraint that any single household failure does not bring down the system. Mega-farms have no such resilience. One bad season, one management dispute, one pest outbreak, one funding gap, and 40,000 acres go to weeds.

2. The water is borrowed.

The Galana scheme draws from the Tana. The Tana is also drinking water for Mombasa, Garissa, and millions downstream. Mega-irrigation in semi-arid land is a hydrological transfer dressed up as agriculture. It is not creating water. It is moving it. The downstream cost is real and rarely costed.

3. The labour model is wrong.

Mega-farms need workers — but workers without ownership are workers without long-term care. The same families who farm one hectare of their own land productively often produce poorly on a hundred hectares of someone else's. The land tenure model embedded in mega-farm thinking has failed every time it has been tried at scale on the African continent.

4. The pest and disease pressure scales non-linearly.

A million acres of monoculture maize is a million acres of MLN target. Smallholder mosaics — different crops, different varieties, staggered planting — provide natural firebreaks. Mega-monoculture invites the kind of pest collapse that smallholder agriculture mostly avoids.

5. The political fragility is built in.

Mega-projects need consistent political ownership. Kenya's political cycle does not provide this. Each new administration questions, restructures, or quietly defunds the previous administration's flagship. By year five, the project is in legal limbo. By year ten, the topsoil is gone and the contracts are in court.

Why Smallholders Feed Kenya

The opposite truth: Kenyan food security has held — barely, but held — because of three million smallholder farmers operating at the household scale. They are not romantic figures. Their yields per hectare are far below modern potential. They are under-served by extension, under-financed, and chronically vulnerable to weather.

But they are resilient. When one farmer fails, the next field still produces. When one crop collapses, the next plot has something else. When prices crash, the household absorbs by eating more of what they grew. When the rains return, planting resumes within hours, not after a procurement cycle.

This is not an argument for keeping smallholders poor. It is an argument for the obvious public policy implication: the most cost-effective food security investment in Kenya, in 2026, is to make smallholder agriculture more productive, more market-connected, and more resilient — not to bypass it.

We have known this for forty years of development economics. The IFPRI literature has said it. The Lipton thesis from 1977 said it. The CGIAR has said it repeatedly. Kenyan cabinet papers have ignored it for at least three administrations because mega-projects make better photo opportunities than agricultural extension does.

What Better Public Investment Looks Like

If you took the public funds spent on Galana-Kulalu over the last decade and asked: what would have moved the food security needle more? The list is not exotic.

1. Smallholder irrigation at the village scale.

Sand dams, micro-irrigation, low-cost drip systems. For the cost of one Galana mega-pump, you can fund hundreds of village-scale sand dams. Each one feeds a small community for forty years.

2. Seed and variety access.

Drought-tolerant maize, orange-fleshed sweet potato, climate-adapted bean varieties. The cost-per-impact ratio is dramatically better than mega-irrigation, and the gains compound across millions of farms.

3. Storage and post-harvest loss.

Kenya loses an estimated 20–30% of its maize harvest to post-harvest mismanagement — pests, mould, poor storage. Halving that loss would do more for national food security than tripling Galana's yield. Hermetic bags. PICS bags. Solar-cooled village stores. Boring. Effective.

4. Extension services that work.

The number of working extension officers per farmer in Kenya has fallen for two decades. Smallholder productivity gains require human contact — not master plans.

5. Markets that absorb surplus reliably.

When smallholders produce more and there is no buyer, prices crash and the lesson is "produce less next year." The aggregator-cooperative-warehouse-receipt cluster of investments is unsexy. It also works.

None of this gets a ribbon cutting. All of it would have fed Kenya better than Galana-Kulalu.

The Pattern To Watch

A new mega-farm proposal lands every few years. The latest will be along the Tana, or in Galana-adjacent acreage, or in a new SEZ. The deck will be slick. The contractor will be politically connected. The yield projections will be familiar. The water source will be vague.

When you see the next one, ask the five questions:

• Where does the water come from, and who downstream loses it?
• What is the labour model, and who owns the land?
• What is the disaster scenario at full scale (pest, drought, fire), and what's the contingency?
• What's the political durability across a change of administration?
• For the same money, what would 5,000 smallholder interventions deliver?

If the answer to any of these is "we'll figure it out," the project is Galana-Kulalu in a new costume. Which means we will spend the money and have nothing to show in fifteen years.

Why I Wrote This

Because food security in Kenya is being held together by three million people who have never been to a national policy meeting. They deserve more honest investment than the next mega-farm fantasy.

For the related conversation on what real food sovereignty looks like, see the price of a plate and traditional foods and modern nutrition.

Mega-farms are not the answer. They are the question we keep failing to retire.