All abroad!
In Switzerland, they had brotherly love, and they had 500 years of democracy and peace. And what did that produce? The best way of running a railway in the world.
Everybody agrees that the railways are stuck. Partly this is because of decades of under-investment; partly, because a diagram of their ownership structure looks like it was drawn by a spider on an acid trip. But a third problem, which gets much less attention, is that the railways do not know what they should be optimising for.
This matters because railways are good for the country. Aside from their environmental benefits, railways are much more efficient than cars. One track of the Victoria Line moves more than 40,000 people per hour; given that the average car in the rush hour has 1.2 occupants, one lane of the M8 in Glasgow, our most efficient road, can only move about 2,000 people per hour. At their best, trains are the most land-efficient form of transport, enabling faster journeys that are not subject to congestion, without bulldozing town and city centres to build roads. It is in the national interest that the railways are run well.
The railways should be optimising for connectivity: making as many journeys from A to B as fast and straightforward as possible. The only way to beat the car is to provide a quicker, easier, more pleasant journey. This is possible if A and B are linked by a fast direct train. But very few of the 3.3 million possible train journeys in Britain have a direct train, let alone a fast one; the rest require a change of trains. People will be less likely to make an A <> B journey, changing at C, either if they need to run from one side of a station to the other, or if they need to hang around for an hour.
To make connections easy, we should copy Switzerland. At Zurich station, nearly all long-distance trains arrive just before the xx.00 or xx.30 minute, and depart just after. This means that every train makes a good connection with every other train. It’s easy, say, to get from Chur in the south-eastern Alps to Yverdon in the west of the country: the train from Chur gets into Zurich at xx.22, and the train to Yverdon leaves 8 minutes later at xx.30: long enough to get from one train to another without rushing, but short enough to avoid waiting around.
Well-designed connections are not restricted to Zurich. Suppose you want to get from Grafenried, a village near Bern, to Le Châble in the Alps. In Britain, even railfans would think twice about making a journey that involves changing trains three times. In Switzerland, however, passengers get a 14 minute wait at Bern, 9 minutes at Lausanne, and 7 minutes at Martigny.
This isn’t because the Swiss have designed their network around Le Châble and Grafenried. It is instead because as many A <> B connections as possible are easy, which gives them the A <> Z connections for free.1
The Swiss approach works: pre-pandemic, 20% of all kilometres travelled used rail, the highest modal split in western Europe, in spite of the high fares. In the canton of Zurich, 43% of commutes use public transport. Swiss-style planning has been adopted by Austria and is being implemented in Germany.
The Swiss approach, however, is about more than timetabling: it is also a paradigm for designing infrastructure. In the late 1980s, Switzerland began calling all trains at Zurich around the xx.00 and xx.30 minutes. However, the trick could not be repeated at Bern, the most important node in the west of the country, because the journey from Zurich took 1h09.
At the same time, the line between Bern and Zurich was becoming congested. This situation would be familiar to Britain in the late noughties, France in the late 70s, and Japan in the late 50s: the main line connecting your most important cities is full, so it’s time to build a high-speed bypass. The Swiss could have built a very fast line, reducing the Zurich <> Bern travel time by half an hour, to 39 minutes: useful for people travelling between Zurich and Bern, but it wouldn’t have made the connections at Bern any easier.
Instead, Switzerland built a slightly slower line, with the goal of getting Zurich <> Bern travel times to just under an hour. This means that trains leave Zurich just after xx.00/30, and arrive at Bern just before xx.00/30: both nodes get quick connections.
Two things stand out about this approach. Firstly, the line was built to be as fast as necessary, not as fast as possible.2 The speed of a particular service is not useful in itself: speed is useful to the extent it makes journeys faster. Speeding up the line between A and B by 15 minutes, but in doing so breaking the connections at B, is counterproductive. Secondly, infrastructure was built to operate a specific service. This gets things the right way round: infrastructure is a means to an end, not an end in itself.
If Britain adopted Switzerland’s railway planning, we would begin by redesigning the timetable so as many connections are as easy as possible. Birmingham would be our equivalent of Zurich, due to its central location and direct trains to nearly every other city, all of which would now connect with one another.
In itself, this would be as transformative as HS2. HS2 aimed to make a small number of direct journeys a lot faster. The Swiss approach speeds up the very large number of journeys that require a change of train, by cutting waiting times at interchange stations. Redesigning the timetable to focus on connectivity requires spending no money, pouring no concrete and appeasing no NIMBYs. It just requires a change of mindset in the way the railways are run.
Following the Swiss, we would then start to build new lines. There would be a programme of upgrades aiming to get journey times between important nodes to just under a multiple of 30 minutes. Obvious candidates would include: 60 minute journeys between Manchester and Birmingham, and Birmingham and Reading; 30 minutes between each of Manchester, Sheffield and Leeds, and between Sheffield and York; 60 minutes between Peterborough and York, and Newcastle and Edinburgh.
Obviously, Britain’s bigger problem than building the wrong infrastructure is our inability to build any infrastructure. But these kinds of schemes are more robust to an environment of institutional sclerosis, for two reasons. Firstly, Swiss-style planning is easier to sell to a sceptical public, because it involves designing services first, then building the infrastructure to support them: the benefits are clear from the word ‘go’. This won’t necessarily be enough to defeat the NIMBYs, but it will be better than the HS2 approach, where the benefits were left vague – nobody has published a timetable for the services made possible by HS2 – but the costs were certain.
Secondly, there will be no ‘big bang’ projects, like HS2, alleged to fix all the country’s problems in one go. Big bangs are vulnerable to huge cost overruns, and to being cancelled for political reasons. But the opponents had a point: HS2 was trying to solve the wrong problem, using the wrong methods, using a wrong model of the world. To avoid making the same mistake again, our railways need to start by recognising that they are a network for moving passengers around, and plan accordingly.
This is my entry for the TxP Progress Prize, a prize for blogposts written by people early in their careers that identify solutions to Britain’s malaise. It’s also inspired me finally to start publishing stuff I’ve written on Substack.
Thumbnail image by Fraknö on Wikimedia Commons.
.jpg){kind=link}
The Swiss approach is called in German « integraler Taktfahrplan », which doesn’t translate well into English: ‘integrated clock-face timetable’ is the usual translation, although I quite like ‘coordinated pulse network’.
The German slogan is, « Nicht so schnell wie möglich, sondern so rasch als nötig ».