Europe has more rail network per square kilometre than almost anywhere else, and it moves a smaller share of its freight by rail than it did a generation ago. Both facts are true at once, and the contradiction between them is the whole subject.
The intuitive explanation is that the track is congested by passenger services, which is partly right and mostly a distraction. Where freight has been studied closely, the binding constraint turns out to sit at the two ends rather than in the middle: the terminal where a container is lifted on, and the terminal where it comes off. A route with capacity at both ends will carry freight. A route with capacity at neither will not, however much track sits between them.
The unglamorous asset nobody photographs
A railhead is a slab, a crane or a reach stacker, a few sidings long enough to hold a full train without fouling the running line, and a road connection that can absorb a day's lorries. None of it is technically hard. All of it is expensive, land-hungry, planning-intensive and invisible in a way that makes it politically unrewarding. Ministers open tunnels. Nobody cuts a ribbon on a siding extension.
The consequence is a network with long stretches of well-maintained line feeding terminals that are too short, too few, or in the wrong place relative to where goods now originate. Trains are limited to the length the shortest siding on the route can hold, which in practice means a great deal of European freight runs at well under the length the track could carry, and the economics of a short train against a lorry are not favourable.
What has changed is that the money has started following the constraint. A meaningful share of current European rail investment is going into terminal capacity, siding lengthening and the electrification of last miles into ports and industrial estates, rather than into headline new lines. It is a shift from building the network to finishing it, and it is being driven by operators who can demonstrate exactly which movements they cannot run and why — the same evidence-first turn that moved nearshoring from conference talk to freight manifest.
There is a second constraint that money addresses more slowly. Freight wagons and locomotives are certified nationally, and although the paperwork has been harmonised in principle for years, in practice a train crossing several borders still requires equipment approved in each and drivers qualified for each. The cost of that friction lands entirely on the operator, and it is why an intermodal service that is competitive on a single-country run stops being competitive the moment it crosses two frontiers.
None of this is unique to Europe, and the American version is instructive precisely because the failure is at a different point. There, the ports have largely solved their own throughput problem and discovered that the roads and yards behind them did not, which is the same lesson arriving from the opposite direction — capacity is a property of the whole chain, and the chain reports the capacity of its worst link rather than its best.
The load-bearing detail is that terminal investment pays off only if the freight it is built for actually shows up, and freight shows up on the basis of reliability rather than price. Shippers will pay a premium for a service that arrives when it says it will and will not accept a discount for one that mostly does. That makes the first years of a new terminal a chicken-and-egg problem which public money is unusually well suited to solving and unusually reluctant to, for reasons that rhyme with the gap between a funded project and a finished one.
It also explains why the projects moving fastest are the ones attached to a single large shipper willing to commit volume in advance. Those get built. The speculative ones, which are the ones the network actually needs to become dense enough to compete, mostly wait — for the same structural reason a bridge stays posted rather than repaired until somebody with a budget owns the consequence.
Topics worldfreightinfrastructure
