The return of the night train

AFor about thirty years the sleeper train looked like a museum piece. Services across Europe were withdrawn one after another through the nineteen nineties and two thousands, and the reasoning was not sentimental in either direction: a budget airline could move somebody between two capitals for less than the cost of the bed linen, and the night train lost on price in a market where price was the only thing being compared.

BThe economics were worse than they looked. A sleeping car carries perhaps thirty passengers where a seated carriage carries eighty, so the vehicle earns less per journey while costing more to clean, staff and maintain. It also earns nothing during the day, because it arrives in the morning and cannot usefully be sent anywhere until evening. A seated train makes four journeys a day; a sleeper makes one.

CWhat changed was not a technical breakthrough but a change in what passengers were counting. As the climate cost of flying became a public question, a proportion of travellers began treating the journey as a choice with consequences rather than as a price to be minimised. Several operators report that demand returned first on routes where the alternative flight was short, which is the opposite of what a purely financial model predicts, and is explained once you notice that a short flight is the one where the environmental argument is strongest and the time saving smallest.

DThe second change was in what was being sold. The withdrawn services had offered a bunk in a shared compartment with a shared corridor washroom, which is a difficult product to sell to anybody who has a choice. The new services separated the offer into tiers: a seat, a shared couchette, a private cabin, and in some cases a cabin with its own shower. The private cabin is the one that pays, and it is bought by people who are not choosing between the train and a budget flight but between the train and a flight plus a hotel night.

EThat comparison is the heart of the current argument. Counted as a journey, the night train is expensive. Counted as a journey plus a night's accommodation plus the airport transfers at both ends plus the evening and morning that a flight consumes in waiting, it is frequently cheaper, and this is before any environmental figure is introduced. Whether a passenger performs that calculation depends almost entirely on how the price is displayed, which is why the operators now advertise the fare alongside the cost of a hotel room rather than alongside the air fare.

FThe obstacles that remain are administrative rather than physical. A train crossing four countries needs approval, a path and a crew arrangement in each, and the rules differ at every border; the technical requirements for signalling equipment alone have defeated several proposed routes. Rolling stock is the other constraint, since almost no sleeping cars were built for thirty years and the surviving fleet is old. New vehicles have been ordered, but a sleeping car takes years to deliver, and an operator must commit to a route before knowing whether the demand will hold.

GThe honest forecast is therefore narrow. Night trains will not replace short-haul aviation, and the operators making the strongest claims are not helping, because a service that is oversold and then found to be slow and expensive loses the argument permanently. What the evidence supports is a particular niche: city pairs between about six and twelve hours apart, where the train departs late enough to allow an evening and arrives early enough to allow a working day. Inside that band the night train is not a nostalgic indulgence but the fastest useful option, and outside it, it is a very slow way to travel.