What the Nairobi–Kampala corridor tells you about integration

Regional integration in East Africa is usually measured in signatures. The Customs Union, the Common Market Protocol, the single customs territory, the monetary union that is always five years away. If you wanted a better measure, you could stand at Malaba and count the hours.

The Northern Corridor runs 1,700 kilometres from the port of Mombasa through Nairobi to Kampala and on to Kigali, Bujumbura and eastern Congo. It handles more than 35 million tonnes of cargo a year and over 80% of Kenya's transit trade. It is, physically, the East African Community. Whatever the treaties say, this road is what integration actually consists of.

The border is not the problem

Here is the first thing the corridor tells you. A truck leaving Mombasa for the Ugandan border currently takes 76 to 80 hours, nearly double the target of 36 to 48. The natural assumption is that the delay is at the border, because the border is where two countries meet and integration is supposed to be about countries.

It isn't. The Shippers Council found that border post procedures accounted for about one percent of overall delays, while the biggest single cause was frequent stops on the Kenyan side for police and security checks, at 36%. The one-stop border post at Malaba, the thing built with donor money and cut with a ribbon, works. The bit that doesn't work is the two-day gauntlet of roadblocks, weighbridges and agency checks inside Kenya, before the truck ever sees Uganda.

That inversion is the whole lesson. The international layer of integration, the part that requires two governments to agree, was solved. The domestic layer, the part that requires one government to coordinate its own police, its own weighbridges and its own agencies, was not. And the cost is real: Kenya has lost 8 to 10% of its transit cargo market share over three years as operators divert to Dar es Salaam and the Central Corridor. In April 2026 the government announced it would cut police roadblocks to five or six along the route. That is an integration policy, even though it involves no other country.

Money follows the same shape

Now look at the payment that travels with the goods.

On paper the mobile money layer integrated years ago. Safaricom and MTN interconnected in 2015 so that M-Pesa users could send to Uganda, and the Vodafone–MTN agreement later extended that across seven countries. A trader in Busia can send shillings to a supplier in Jinja from her phone. This is the treaty-level success again: two big operators agreed, and the cross-border leg exists.

But watch what people actually do with it. In Kenya, Uganda, Tanzania and Zambia, users now link their M-Pesa or MTN wallets to crypto accounts and use the mobile money balance only as the on-ramp and off-ramp for a stablecoin transfer. The expensive cross-border leg moves to the stablecoin; the mobile money network keeps the cheap local cash-out at each end. The interconnect exists, and it is being routed around, because the cost of the middle leg sits with whoever owns the corridor between the two wallets and nobody has been made to lower it.

That is the roadblock problem in a different costume. The border, the operator-to-operator handshake, is fine. The domestic stretch, the fee and FX spread on each side, is where the time and money leak, and the leak sits in a part of the system no single party is accountable for.

Integration is a question of who owns the leg

Put the two together and a rule falls out. Integration works where a single party owns an entire leg end to end and can be held responsible for its throughput. It fails where a leg is shared across many parties, each with a reasonable local incentive to stop the truck, add a check, or take a spread.

Malaba works because one border authority owns the crossing. The Mombasa–Malaba road doesn't because six agencies own pieces of it. The M-Pesa–MTN handshake works because two companies signed one contract. The cross-border payment as a whole doesn't because the FX, the fees and the compliance are split across operators, banks and regulators, and every one of them can make the trip slower without being the one who made it fail.

The same pattern shows up at the domestic level, and Kenya has been quietly proving it. When an Airtel subscriber can now top up through an M-Pesa paybill and the funds clear between the two operators' ledgers instantly, that is because the Central Bank of Kenya pushed hard enough for interoperability that the leg got an owner. Nobody signed a treaty for that.

What the rail will and won't fix

The standard gauge railway is the corridor's next promise. Kenya's line already runs Mombasa to Naivasha; the Uganda section from Malaba to Kampala is about 272 kilometres, construction accelerated in 2026, and the core Mombasa–Kampala rail corridor could be complete around 2028. The projected cost of moving a 40-foot container falls from roughly $3,500 by road to about $1,500 by rail.

That is a real gain, and it is the same kind of gain as the roadblock reform: it consolidates a fragmented leg into one owned by a single operator with a single timetable. Rail cannot be stopped thirty times between Mombasa and Eldoret. That, more than the steel, is why it works.

But rail moves the goods, not the money. The payment leg still has to be consolidated by someone, and right now the party doing it is the importer with a USDT wallet, privately building the integration that the region's institutions have not delivered.

The measure I'd use

If you want to know how integrated East Africa is, ignore the protocols and count two things. How many hours from Mombasa to Malaba, and what percentage of a $10,000 supplier payment leaks out between a Kenyan wallet and a Ugandan one. When both numbers fall, integration has happened. Until then it is signatures.