On the African Industrial Revolution

Over my travels, I've acquired some first-hand wisdom I believe should be shared. Sometimes I hear people from the West ask about Africa's development and assume that there is no industrialization on the continent. But this is absolutely not the case, the continent is actively industrializing as we speak. Quickly.

Africa has no shortage of capital OR labor. Africa adds about ten to twelve million people to its working-age population every year. The number will keep climbing for the next two decades while it falls everywhere else. When you have the labor capacity, you just add in skills. But the thing is, across the board, labor capacity ALWAYS preceded skills. Shenzhen did not have a trained electronics workforce in 1985, it had a lot of people and strong reason and capability to train them.

I also am being serious about the capital. Africa's major problem is not capital, the capital is there, and there is a lot. It's just sitting in low risk instruments and not being utilized for catalyzing industrialization.

Kenya alone is a good illustration. The SACCO movement holds something over a trillion shillings across more than fifteen million members, which is real savings from real working people, and most of it recycles into member loans for land, school fees, and vehicles. The pension funds hold hundreds of billions more and park the majority in treasury bills because that's what the allocation rules and the available instruments permit. Add remittances, which run near a hundred billion dollars a year continent-wide and exceed foreign direct investment and aid combined, and you're looking at an enormous pool of patient domestic money that has nowhere productive to go.

Why the capital sits still

Underwriting a transaction costs roughly the same whether the transaction is fifty thousand dollars or fifty million. You need the legal review, the site visit, the financial model, the collateral search, the credit committee, the ongoing reporting. Call it thirty to eighty thousand dollars of professional time for anything an institution will touch, and that's my own estimate from watching deals get built. On a fifty million dollar facility it's a rounding error. On a fifty thousand dollar equipment lease it's the entire economics, so the lease never happens and the workshop never buys the machine.

Institutional money can only reach transactions large enough to absorb the cost of finding out whether they're real, and almost nothing being built on the continent is that large yet. So the money goes into government paper, the workshop stays a workshop, and everybody concludes there's a capital shortage.

Labor becomes capital

There's an obvious objection to all of this, which is that robotics will be perfected long before Africa finishes industrializing, and a continent whose advantage is people will find that its advantage has been automated away.

I think this gets the relationship backwards. Robots are capital that performs labor, and what they actually do is multiply the output of the person operating them. A worker with a robotic cell produces what six workers used to produce, and the six workers were never the scarce input in the first place. Automation raises the ceiling on what a small operation can attempt, which is precisely what a continent full of small operations needs.

Robots also won't improve inefficiencies or spearhead change. They will just strengthen those that do. A badly run business with robots is a badly run business that loses money faster. Every real gain from automation still requires someone to identify the bottleneck, redesign the process, train the operators, and hold the quality line, and those are human jobs that grow rather than shrink as the machines get better.

Everyone starts small

The countries that industrialized most recently did not start by building finished products. They started by making one component well and selling it to somebody who made the finished product elsewhere.

Turkey is the cleanest example. It didn't design a Turkish car, it entered joint ventures with foreign manufacturers in the late sixties, Tofaş with Fiat and Oyak with Renault, and spent thirty years building a component supply base underneath them. Today Turkish suppliers make wiring harnesses, castings, seats, and drivetrain parts for most of European automotive, and the whole sector grew out of licensed assembly. Vietnam did the same thing a generation later, starting with cut-and-sew garments and moving into electronics assembly for Samsung before it began making the components it once imported.

Take servo motors, a servo is a laminated steel stator, copper windings, a magnet rotor, a machined shaft, precision bearings, an aluminum housing, an encoder, and a drive board. A Kenyan operation could credibly do housing casting and machining, shaft machining, lamination stacking, winding and impregnation, magnetization, balancing, assembly, and end-of-line test inside two years, which is somewhere around a third of the value of the finished unit. Stamping, magnet wire, and die casting follow in the next few years. Electrical steel, sintered magnets, precision bearings, and semiconductors are imported for a very long time, and that's completely fine. Nobody makes all of those anywhere.

The green part

Kenya generates the overwhelming majority of its electricity from renewables, and the geothermal fields in the Rift Valley are the largest in Africa and among the largest anywhere. The resource that's been drilled is a fraction of what's estimated to be down there. Geothermal is also firm baseload, which is the thing solar and wind can't give a factory, and there's already surplus at night that nobody has found a use for.

This matters commercially and not only environmentally. Europe's carbon border adjustment mechanism will start pricing the embedded emissions in imported goods, which means a motor wound in Naivasha on geothermal power carries a structurally lower landed cost into the EU than the same motor made on a coal grid. Africa is the only continent that can industrialize green from the beginning, because it has no installed base of dirty capacity to retire first. That's a genuine head start and I don't think it's being priced yet.

The private sector is the engine

I don't think we can seriously compare Africa and China. Chinese prosperity was created by the Chinese state, through directed credit, state-owned enterprise, industrial policy, and enormous public infrastructure spending executed by a government with the capacity to execute it. People look at Africa, look for the equivalent state capacity, don't find it, and conclude industrialization won't happen.

What I see every day is an African private sector that will do far more than any African government is going to do, and that is already doing it without much help. The traders, the workshops, the SACCOs, the informal manufacturers, the cross-border wholesalers. The organizing capability is already present at community scale. What has never existed is a platform that lets that capability aggregate capital at industrial scale.

Kenya has actually already proven the model works at industrial scale, and almost nobody outside the country talks about it. The Kenya Tea Development Agency runs more than sixty processing factories owned by around six hundred thousand smallholder tea farmers, organized as factory companies that the growers hold shares in, managed professionally under an agency agreement. It has operated since 1964, it processes a large share of the world's black tea exports, and it built real industrial capacity out of farms averaging well under an acre each. Co-operative Bank of Kenya did the same thing in banking, majority owned by thousands of co-operative societies through a holding co-operative and now one of the largest banks in the country. Farmer-owned and member-owned industry is not a thought experiment here. It's the biggest thing in the economy that Western observers consistently miss.

What KTDA could not solve is the part I care about. Building it took decades, outside capital cannot get into it, and a farmer's stake is locked up for a generation with no way to sell it or borrow against it. Those three constraints are all liquidity and infrastructure problems, and they are exactly what a shared ledger fixes.

So imagine a SACCO with twelve thousand members and a blockchain-based capital layer underneath it. Members subscribe to a tokenized fund the way they already subscribe to shares, except the fund is financing a joint venture with a foreign manufacturer to assemble motors, and the members who take the training get the jobs. The fund holds equipment and receivables, the units are transferable so nobody's savings get locked up for a decade, the diligence is cheap enough that a fifty thousand dollar machine purchase is a financeable transaction, and the returns come back to the same members who put the money in and work the line. KTDA took thirty years to reach industrial scale. Five years from a standing start is not an unreasonable timeline for the same thing now, and it needs no ministry to approve it.

The pieces all exist independently. The savings exist, the members exist, the foreign manufacturers looking for market access exist, and the demand exists. What's been missing is the plumbing to hold them together, and that's now a solved problem.

The low hanging fruit

The industries worth attacking first share three properties: real domestic demand today, moderate technical complexity, and heavy import dependence.

Electric motors and drives sit at the top of that list, since everything from a maize mill to an irrigation pump to a conveyor runs on one and every single one is imported. Solar inverters and mounting systems are next, and so are battery packs, which are assembly problems more than chemistry problems. Agricultural processing equipment, cold chain units, transformers and switchgear, wire and cable, and electric two and three-wheelers all qualify. None of these require a semiconductor fab. All of them require a competent machine shop, a trained workforce, and someone willing to finance the first thirty units.

The hard parts don't vanish though. Power costs what it costs outside the geothermal corridor, ports are as congested as they are, and enforcement is the piece I'm least confident about, because a tokenized claim still requires somebody to physically repossess a machine when the claim goes bad. Better collateral registries help and I don't know whether they help enough. Financing also creates nothing on its own; fund a plant with no customers and cheaper capital just distributes the loss more efficiently :)

But those are problems you earn the right to have. Right now the financing problem stops the conversation before it starts, and it stops it for reasons that are entirely mechanical and entirely fixable.

If you want to see the machinery, go read ERC-3643 and ERC-7943. They're short, and they'll show you exactly how eligibility and transfer restrictions get enforced inside the instrument itself. Then go find out what your local SACCO or pension fund is currently permitted to hold, and go ask your regulator why the list is that short.