Every large African infrastructure announcement carries a hidden claim: that a model has been found which others can copy. The danger is treating a single deal as a blueprint before its assumptions have been tested. On 23 May 2024 Kenya offered a case worth examining on exactly those terms. The Kenya National Highways Authority and the investment manager Everstrong Capital said they had agreed to develop a 440-kilometre Nairobi-Mombasa expressway at a stated US$3.6bn, funded by a blended pool of international and domestic capital. The interesting question for an analytical reader is not the road but the framework it represents, and whether that framework travels.

The Model: Financing-led Infrastructure

The structure on display is a shift in who originates a road. Traditionally a government designs a highway and finds a contractor to build it against a budget line; here an investment manager arrives with a financing structure and a delivery plan for a nationally critical route. That inverts the usual sequence, putting the capital stack and the revenue model at the centre and treating the asphalt as the means to a cash flow rather than the object of the exercise. The logic is that essential corridors with broad, durable demand can be underwritten by private and institutional money rather than the sovereign alone, and that discipline about revenue, imposed at the design stage, produces assets that are financed to last rather than built to a ribbon-cutting. It is a way of importing the rigour of capital markets into public works.

The takeaway: the innovation is not the highway but the financing-first way it is being assembled.

The Local Assumptions: What Might Not Travel

A model is only transferable if its assumptions hold elsewhere, and several here are distinctly Kenyan. The Nairobi-Mombasa route has an unusually strong demand base, because it is the coastal leg of the Northern Corridor serving Uganda, Rwanda, South Sudan and eastern DRC; few African routes carry comparable regional freight, and demand is what makes a financing-led model bankable in the first place. Kenya also has relatively deep domestic capital markets, a functioning highways authority in KeNHA, and prior experience structuring expressway concessions, which lowers the cost and risk of assembling the deal. A country without that traffic base, that institutional depth or that pool of domestic savings might reach a very different answer with the same template, and could find that the structure that de-risks the road in Kenya simply shifts risk onto the state elsewhere.

The takeaway: the corridor’s demand and Kenya’s institutions are local variables, not universal ones.

The Second-order Questions: Governance, Data and Control

A financing-led corridor raises questions that outlast construction. Who owns the asset and sets the tolls, and under what regulatory oversight, determines whether a public good becomes a private rent extracted from captive users. Cashless tolling generates a continuous stream of mobility and payment data, which raises governance questions about who holds it, who may use it and under what rules, in a country already building out its data-protection regime. And the concession’s length and exit terms decide who controls a critical national artery decades from now, and on what basis it returns to public hands. These are strategic choices embedded in the fine print, and they are far harder to renegotiate once the asset is built and financed.

The takeaway: the durable stakes are ownership, data and control, not the opening date.

So What: Study the Template, Test the Assumptions

For an African operator or policymaker, the value of 23 May 2024 is as a case study, not a recipe. The financing-led model is worth understanding, because closing the infrastructure gap will require drawing private capital into essential assets rather than waiting for budgets that will not stretch. The discipline is to name the assumptions before copying it: is there a demand base as strong as the Northern Corridor, institutions as capable as KeNHA, and domestic capital deep enough to share the stack. Where those hold, the template may transfer. Where they do not, the same structure could deliver a very different, and less benign, result.