Is taking a train something you do every day, rivaling a car? Something flexible, you just get on and go?

Or is it something you do less regularly? Something where you plan, that you know long ahead where and when you will travel? More similar to how most people take scheduled flights.

The answer to that of course, is it is a bit of both.

How a passenger behaves in what they do on an everyday basis over 20km will differ massively from what they do occasionally for a trip of 500km. What compromises I, the passenger, am willing to stomach will likewise be different in each case. I might be willing to stand for 20km, but I would rather avoid that for 500km. I might likewise be OK to reschedule a long trip a bit so as to increase my comfort.

Railways have to devise systems to cope with all of this, and also cope with customer behaviour patterns – peaks at commuter hours, peaks to escape the city at weekends etc.

The question then arises: how far should a railway go to shape that demand?

Train tickets were traditionally calculated by the distance travelled, and passengers could flexibly take any train. Some railway systems – like SBB in Switzerland – still operate this way. But giving people the complete freedom to choose when to travel results in very large morning and evening peaks each day, and coping with those peaks means maintaining extra trains to only be used then – see the massive fleets of P trains SNCB parks up at Schaerbeek outside of Brussels for this purpose.

So, instead of maintaining extra trains for the peaks, can you flatten out the demand with price incentives? NS in the Netherlands runs a peak and off peak pricing system, giving those passengers that can travel off peak a price incentive to do so, but with flexibility to take any train within broad time bands.

But how do you manage to do that in countries with trips much longer than the Netherlands, or indeed even internationally? Here is what the Germans call Zugbindung comes into play, as applied by Deutsche Bahn – a ticket that allows you to travel only on the train stated on your ticket. Pretty much all railways in larger European countries operate a version of this, as do most international routes. Take a low demand train, compromise on flexibility, and get to your destination for a low price.

Both of these cases are examples of yield management – using price incentives to shape customer behaviour.

However in both the Dutch and German cases, if I pay enough I have flexibility still – if I do need to depart right away, I can still purchase a flexible ticket, valid at any time on any train. If the train is overloaded I might have to stand, but I will still get to my destination.

But what then happens when you go one step further – and introduce compulsory reservation? In other words, every passenger travelling has to have a ticket for that specific train. This makes the train the same as an aeroplane – no ticket for that specific service, well tough, you cannot go. And a train can literally be sold out.

This might not be such a problem in Sweden or Italy where there are often non-reservation alternatives to the compulsory reservation high speed trains, but in France (SNCF) and Spain (Renfe) the only way between many cities is on compulsory reservation services.

Faced with that, how does a railway company behave? Nudge passengers onto off-peak services with cheap tickets for those? Or ditch the off peak services and hike up the prices for the peak services? SNCF was proud that 40% of its TGVs were sold out last year – but to me that means 40% of trains no one could take if their plans change at the last minute. And if you look at how thin TGV timetables are at off-peak, SNCF is pursuing a profit-maximisation rather than a capacity-maximisation strategy. Yield management has crossed from being a way to nudge passengers onto off peak services, and has become purely a means to maximise profit. And the consequence is severe restrictions on passengers’ flexibility, and rail not being able to adequately boost its modal share.

Also please do not try to tell me that this compulsory reservation malarkey is all for safety reasons. Anyone who has travelled on one of the few TGVs that runs during French train strikes can tell you that running these trains full to the rafters is possible on those days.

As if that were not enough, this approach messes up international journeys with connections. If my Deutsche Bahn train Frankfurt-Strasbourg is delayed, and I miss my SNCF TGV Strasbourg-Lyon, there may well be no way to take the next one – because it is full, and I would not be allowed to stand on it even if I wanted to. Different railways have until now largely failed to facilitate systems to combine yield managed tickets they each separately offer.

With regard to all of this, railways differ from airlines – because in air there is competition. If Lufthansa does not offer me a cheap off-peak ticket to a destination, then Ryanair or Easyjet will be only too happy to do so. But – in France at least, and on Paris-Bruxelles, Paris-Switzerland and Paris-Barcelona – there is no Ryanair-style operator. There is OUIGO on some routes – but that is run by SNCF as a sop to politicians to show the company is vaguely doing something for low income groups. In Italy, Spain and Sweden competition is at least part present, but for the moment this is a rare thing on international routes. And the difference of service pattern in Spain on lines with and without competition is stark.

So my preferred outcome is yield management without compulsory reservation – Deutsche Bahn style. The railway company spreads the demand, while I, the passenger, have flexibility if I need it – albeit at a price. Compulsory reservation can just about work if there are alternatives – either in the form of slower non-compulsory reservation services, or competition against the incumbent operator. In the absence of either of these, the cards are too stacked in favour of the railway company – as the SNCF case shows.


  1. Max Wyss

    It may be a bit extreme, but wherever yield management with Zugbindung exists, it is a sign of lousy service. (and SNCF and Renfe are the proof for that).

    For regional rail (such as S-Bahn), there is always a morning peak, no matter what you do. The only thing to do against that is increase frequencies. The evening peak is way more flattened, and therefore less a problem.

    In Switzerland, not on federal level, but on level of Tarif-/Verkehrsverbund, there is a little bit of an incentive to avoid the morning peak with the various 9-Uhr tickets, available as day pass but also monthly or annual (such as from ZVV). Also in Switzerland, there are some trains where a reservation is recommended, and in some cases, specifically through the alpine base tunnels, a train may not proceed if it is overloaded (I read about 130% capacity), for safety reasons. Tilting trains overloaded by about this level have to run with the tilting mechanism deactivated. But these are special cases.

    Yield management has been introduced by phased out airline managers who have no clue about rail operation.

  2. Another sharp and timely article. I have been meaning to write something about compulsory reservations from a Czech perspective as it became a hot topic last summer, as a result of severe over-crowding, not least on the Prague – Berlin expresses. Thats where the argument for compulsory reservation comes into sharp relief. An overcrowded Berliner short-changes those who did reserve. It was a fight just to get over the legs of people to get to the loo, or to the restaurant car, where you’d find people without reservation nursing a coffee to sustain them all the way to Berlin. That isn’t right, is it? However there was a furious debate on social about it. České drahy responded with a very smart pragmatic decision. They’ve made all reservations free. It´s smart because Czechs are cost conscious and the ticket prices cheap. Reservation charges pushing up the total cost were resented by a proportion of travellers. ČD will probably look and hope for a good take-up of their offer on the Berlin route this summer. We’ll see…

  3. It’s always interesting to read about how railways are managed in Europe. In the UK, of course, we have very much the type of model that you are advocating and, mostly, it works pretty well (how well the rest of it works… well… that’s a matter for discussion – strikes and other issues have certainly led to a noticeable drop in punctuality over the last few years. To the point where I decided to move closer to work in order to avoid the increasingly risky commute).

    Regardless, in the UK, everyone moans about how the Continent does everything so much better / cheaper, so I am finding this series of articles, from such a well-travelled man, extremely interesting. Thanks, Jon!


    • Honestly having now travelled in pretty much every European country that has a railway, Britain’s trains do not look too bad in comparison to many. In fact that the UK does so much with a network that hasn’t received major investment (in terms of new lines or stations) for some decades is remarkable. Many trains are modern, timetables are generally good, service patterns reasonable, and there is no compulsory reservation – which is excellent. Sure, the trains are expensive (which is inevitable given the relatively low subsidy), but all things considered it could be a hell of a lot worse.

  4. Joa Falken

    A better solution would be facultative reservations, with reservation fees dependant on expected demand or occupancy, all seats up for reservation, but no compulsory reservation and option to cancel or transfer a reservation at low fee.

    Zugbindung (“prohibited” change in plans) is a mess in any case.
    It is always combined with late-booking price increases and cancellation fees, otherwise everyone would just wait to book the train until they know which they like best to take, takinginto account price differences

    Thus, Zugbindung is always about forcing passengers to use a different train than they would prefer – though it also occurs that the preferred train (in view of price differences) is still the same at the time of travel as when the ticket was purchased, so less damage takes place.

    Differing prices according to time of day / expected occupancy is fine, but please provide a reason why a passenger with a ticket for a train at peak hour should be prohibited to use it for off-peak trains instead?

    Or, if change of booking or cancellation is allowed with or (in case of change) without fee, why should the passenger pay more for the new off-peak ticket than for the original peak-train ticket – which is usually the case as all tickets tend to become more expensive when the time of travel came closer?

    Zugbindung with limited cancellation and compulsory reservation obligation is even worse, because trains will inevitably not be fully used in consequence, as some ticketowners always will just not use the place.

    It also happens that at some point in time, train 1 is initially less crowded and cheaper than train 2, but closer to the day of travel, price differentials have reserved (maybe just because too many clients were lured to train 1, or maybe because holiday travelers that prefer train 1 can plan earlier that weekend visitors that prefer train 2. Again, switching to the cheaper rainis prohibited.

Leave a Comment

Your email address will not be published. Required fields are marked *