Picture the meeting. A ministry of transport, a map pinned to the wall, engineers pointing to the obvious route: shorter distance, gentler gradients, fewer river crossings. The financing consortium nods. Then, for reasons that take years to fully understand, the money goes somewhere else entirely.
This happens more often than the infrastructure literature admits. Geographically superior corridors sit dormant for decades while costlier, longer, more technically demanding alternatives attract railways, roads, pipelines, and the bureaucratic machinery that keeps them funded. The question of why is not really a geography question at all.
The anchor freight problem
No corridor sustains investment on projected traffic alone. What it needs is anchor freight: a commodity with enough volume, enough predictability, and enough political weight to justify the fixed costs before the road or rail line has carried a single tonne. Copper moving out of the Copperbelt, cotton from Uzbekistan, landlocked oil from Chad. The anchor commodity does not care about gradient efficiency. It cares about getting to port.
The Northern Corridor connecting Mombasa to Kampala and Kigali developed sustained investment partly because Ugandan and Rwandan import demand, combined with Congolese mineral exports, created a thick, durable freight base. The Central Corridor through Dar es Salaam covers similar geography with arguably better engineering characteristics on certain stretches, yet spent years competing for the same capital. The difference was not topography. It was the point at which anchor shippers committed, and to which port their existing relationships were already tied.
Here is a worked scenario that captures the mechanics precisely. Two landlocked countries, call them A and B, both need a corridor to the sea. Country A has a single large copper mining consortium that signs twenty-year offtake agreements specifying port delivery at the eastern coast. Country B has diversified agricultural exports, fifty mid-sized shippers, no single dominant commodity. Investors looking at both corridors will price Country A's throughput far lower in risk terms, even if Country B's route is topographically cleaner and two hundred kilometres shorter. The consortium in Country A essentially pre-sells the corridor's future. Country B cannot do that, and the capital gap between the two routes reflects exactly that missing guarantee.
That gap tends to compound. Each year without committed freight is another year without the engineering studies, the bilateral agreements, the port slot reservations that would make the next round of financing easier to close.
Sunk costs that bend entire systems
Infrastructure investment is not made in a vacuum. It is made against a background of everything already built, and those prior investments exert gravitational force on every subsequent decision.
The standard-gauge question on the African continent illustrates this with uncomfortable clarity. Metre-gauge rail networks built in the colonial period locked entire regions into a particular technology, not because metre gauge was optimal, but because port facilities, rolling stock workshops, locomotive fleets, and the professional knowledge of thousands of rail workers had accumulated around it. Proposing a new standard-gauge corridor alongside an existing metre-gauge system does not simply require building new track. It requires either duplicating port terminals or accepting a break-of-gauge penalty that can add days and meaningful cost per container. The geographically superior new corridor carries an invisible surcharge: the cost of divorcing the old system.
Think of it as a second mortgage on the future, taken out by decisions made a century ago.
This is the part that trips up clean cost-benefit analyses. A corridor that looks cheaper on paper, measured in construction cost per kilometre, may look far more expensive once you account for the transition costs borne by every shipper currently embedded in the legacy network. Those shippers are not irrational when they stay put. They are reading the full balance sheet, and the full balance sheet includes switching costs that no feasibility study bothers to quantify.
Politics as the real gradient
Geography gives you physical gradient. Politics gives you a second one, and it is often steeper.
A corridor that crosses three sovereign borders faces three sets of customs procedures, three sets of axle-load regulations, three sets of police checkpoints, and three governments whose revenue agencies have learned to treat transit freight as a source of informal income. Each of those friction points is a tax on throughput. Reduce physical distance by thirty percent, add two extra border crossings, and the time-cost calculation can reverse entirely. Shippers are not measuring kilometres. They are measuring days.
The Lapis Lazuli Corridor, connecting Afghanistan through Turkmenistan, Azerbaijan, Georgia, and Turkey, was designed to offer Central Asian exporters an alternative to Russian-controlled routes. The geography is workable. The corridor has attracted diplomatic attention and some investment. But it crosses five countries, each with its own transit fee structure, customs authority, and political relationship with the others. The physical route is competitive. The political gradient is brutal. Progress has been fitful as a result, not because engineers miscalculated, but because transaction costs accumulate at every sovereign boundary.
Contrast that with corridors where one dominant state controls most of the route and has strong domestic reasons to push freight through quickly. The political gradient slopes in one direction only, and investment follows the slope. That is not a coincidence. It is the whole story.
What people get wrong about this
The most common mistake is assuming that a corridor becomes viable once someone proves the business case. Wrong. The business case is almost never the binding constraint.
The binding constraint is coordination. Building a corridor requires simultaneous commitment from shippers, financiers, governments, port operators, and insurance markets, and each party will commit only if they believe the others will. This is a classic coordination trap, which is why corridors can sit analytically viable for twenty years and then get built in three. Nothing changed in the engineering. Something changed in the political economy: a large enough anchor shipper made a credible commitment, or a development bank offered first-loss capital that let private lenders follow, or a regional election produced a government that needed a visible infrastructure win.
Ask yourself this: if the route is so obviously superior, why hasn't a single major shipper signed a ten-year throughput agreement on it?
The answer, almost always, is that no one wants to be first. First movers in corridor development absorb the coordination risk for everyone who follows. That risk does not show up in engineering reports. It shows up in the financing spread, sometimes running 200 to 400 basis points above comparable projects on established routes, a figure that tells you exactly how much the market discounts unresolved coordination problems.
Found a corridor that ticks every geographic box but has attracted almost no capital? Look for the missing anchor. Look for the unresolved coordination problem. Look for the border where informal fees have never been disciplined. The answer will be in one of those places, not in the terrain.
Infrastructure investment is ultimately a search for the most governable route: the one where enough powerful actors have already aligned their interests, locked in enough sunk capital, and created enough shared dependency that no single defection can unravel the whole arrangement. Geography is one input into that calculation. It is rarely the deciding one.
The map on the ministry wall is, in the end, the least important document in the room.