InnoTrans is Souring Me on On-Rail Competition
I’m at InnoTrans this week, which means I get to both see a lot of new trainsets and talk to vendors for things I am interested in. Those are interesting conversations and much of the content will make it to our upcoming report on high-speed rail in the Northeast and to some ETA reports. But then, in broad stroke, the presentations about the trains here have deeply bothered me, because of how they interact with the issue of on-rail competition. The EU has an open access mandate, so that state-owned and private railways can compete by running trains on tracks throughout the Union, with separation of operations and infrastructure (awkwardly, state railways do both but there are EU regulations prohibiting favoritism, with uneven enforcement). As a result, I’m seeing a lot of pitches geared specifically for potential open access operators, all of which remind me why I’m so negative about the whole concept: it treats infrastructure as a fixed thing and denigrates the idea that it could ever be improved, while enthroning airline-style business analysis.
Proponents of the model cite higher ridership and lower fares due to the introduction of competition in Italy and Spain, but even then, it’s never really invented anything new, and only gotten some city pairs in those two countries to the service quality that integrated state-provided services have always had in France and Germany. In effect, the EU is mandating a dead-end system of managing trains and making a collective decision not to invest in what worked – namely, building and running high-speed lines.
For people unfamiliar with the argument, I wrote a year ago about TGV ridership and traffic modeling. The TGV overperforms a model trained on Shinkansen ridership, which can be explained based on lower fares, leisure travel to the Riviera, and underdeveloped air competition in small metro areas whose residents mostly drive to larger ones to fly. Relative to the same model, Italian and Spanish ridership underperformed the TGV before competition, and rose to roughly match the TGV or be slightly deficient after competition. So competition did lead to growth in ridership, as the competitors added service and lowered fares – but it only created what the French state did by itself. The German state seems to have French-like results: the trains here are much slower than in France, but relative to that, ridership seems to be in line with a TGV-trained model on the handful of city pairs for which I have any data.
This is causing quite a lot of the buzz in the intercity rail industry in Europe to center cross-border competition and new entrants. But this is, judging by the examples the proponents of competition look up to, not creating anything new. It’s not moving rail forward. It’s just filling in gaps that some state-owned railways – but not the largest two – have in their operations.
And worse, it’s making the long-term issues of intercity rail in Europe worse. There’s practically no cross-border high-speed rail construction in Europe, nor any serious push for making it happen. After a great deal of activism by Jon Worth (and others), the European Commission is announcing regulatory measures in its agenda, starting with passenger rights in case of delays. Physical construction is nowhere on the horizon, nor is there any serious advocacy for (say) a Paris-Frankfurt high-speed line or a Bordeaux-Basque Country one. This is a recent development: in the 2000s there was more optimism about high-speed rail, leading to plans like Perpignan-Figueres. But since then, the TEN-T corridor plans turned into low-speed lines and vaporware, and there’s no real interest in fixing that.
Instead, the interest is in letting the private sector lead. State-built high-speed railways – more or less the only high-speed railways – are not in fashion. The private sector is not going to step in and build its own (despite the sad Hyperloop capsule on display at InnoTrans), but instead look for underserved city pairs to come into, competing with state railways. It’s a story of business analysts using techniques brought in from the airline industry rather than one of infrastructure builders.
And it’s exactly those airline-imitating business analysts who are why RENFE, FS, and Eurostar underprovided service to begin with. The airline world lives off of segmenting the market; there are periodic attempts at all-business class airlines, and low-cost carriers entering and exiting the market frequently. It does not build its own infrastructure, not think in terms of things that could work if the infrastructure were a little bit better. A railway that thinks in the same terms might still build, but will not build in coordination with what it runs. It will do the exact opposite of what Switzerland has done with its tight integration of infrastructure and operations planning; therefore, it will get results inferior to those of Switzerland or even France and Germany.
The trains on display at InnoTrans announce proudly that they are homologated for cross-border travel, listing the countries they can operate in. The main high-speed rail vendors here – Siemens, Hitachi, Talgo, Alstom – all talk about this, explicitly; Alstom had a presentation about the Avelia Horizon, awkwardly given in an American accent while talking about how the double-decker cars with 905 mm seat pitch (Shinkansen: 1 meter) minimize track access charges per seat.
In contrast, I have not seen anything about building new lines. I have not seen booths by firms talking about their work building LGVs or NBSes. I have not seen anything by ADIF selling its expertise in low-cost construction; there are some private engineering consultants with booths, but I haven’t seen ADIF, and the French state section of the conference didn’t at all center French construction techniques. The states that have figured out how to build high-speed rail efficiently seem uninterested in doing more with it than just completing their capital-to-provinces networks; even Germany is barely building. Naturally, they’re also uninterested in pitching their construction, even though they do do some public-sector consulting (SNCF does it routinely for smaller French cities). It’s as if the construction market is so small they’re not even going to bother.
Every other booth at the conference talks about innovation with so many synonyms that they swamp what the firm actually does. But beneath the buzzwords, what I’m seeing, at least as far as physical infrastructure goes, is the exact opposite of innovation. I’m seeing filling in small gaps caused by last generation’s bad airline imitation with a different kind of airline imitation, and nothing that moves intercity rail forward.
Pedestrian Observations 
Hi Alon,Very good piece. One question: Do you oppose open access access or do you just think it should not be a focus area for growth/innovation/activisim?I could imagine a situati
The problem with open access is tracks are limited. Eventually two different organizations will need (not want!) to use the track at the same time and one must lose. There is no fair way to solve this problem, so who ever loses will start fighting politically since they are losing money by not being able to run the service they want to.
Access should be possible. It is useful to be able to close your tracks overnight and use someone else’s, then they close their track the next day and use yours – but the only works because overnight is not peak and you can accept running less service – plus you are getting some gain from that. There is also unexpected emergencies (terrorists destroyed my bridge) and so the other tracks suffer while doing repairs, but these should be called shared suffering.
The real problem is that it is easy and common to not run as many trains as you should and then when you realize you should run more trains you discover someone else is running a train in that time and now you can’t. This is particularly true if someone is running airline style seat management – when the next leaders realize the economics of trains are different and so empty seats are not a bad thing and so want to run better service they can’t.
Would be interested in your response to this – https://manhattan.institute/article/diminishing-returns-and-increasing-political-demands-on-infrastructure?utm_source=virtuous&utm_medium=email
Not sure what Alon thinks (I can guess but I’ll let them respond if they want).
I’ll remind you that the US is investing in transit not for purposes of good transit, but for community monuments, and political supporters (some would call this legalized corruption). If you only count the transit benefits you will get a bad picture of the real return on investment. I personally do not weigh community monuments or the political issues very high, but that is a subjective judgement.
Even though there are a lot of bad transit investments made, that doesn’t mean all are bad. Some of them do have a high return on investment. If the investment is in mass transit, the network counts and so the true return will be much lower if we decide the first was a bad investment and so don’t do the second project – but if we stay the course and finish the 5th project that total network becomes useful and so the return on investment for the 1st is much higher.
We have also only looked economics. Some transit investments make things safer – this won’t show up on economic reports but still makes thing better. Of course some of that benefit is lost because people just drive faster but overall highway deaths are way down on a per mile basis. (Most of the really bad ideas from the 1950s have been replaced with something better by now – there are still a lot of bad intersections but they are not nearly as bad as the worst of the 1950s)
Take that for what you will.
Thanks, Alon!
Was JR Central there? As I recall, they were promoting their integrated approacxh to both railcars and infrastructure.
I’m wondering if Lohr are showing their electric mini-train called Draisy that I read about this morning. Battery electric that recharges at each station. Like a bus it holds up to 80 people. SNCF are planning to deploy it to 3,600km of regional rail that has fallen into disuse and would otherwise require costly track rehabilitation. Specifically, lignes de desserte fine du territoire, short track connecting to towns or mainline stations etc. This thing is only 20 tonnes, about an eighth of standard locomotives. Can’t tell if it is driverless.
See https://www.groupe-sncf.com/en/innovation/local-mobility/draisy
I wonder what the cost might be?
JR Tokai is not exhibiting this year, in fact the only Japanese railway operator represented is Tokyo Metro. All other exhibitors are suppliers or research institutes, under the umbrella of the Japan Overseas Railway System Association.
https://www.jorsa.or.jp/resources/News/files/NEWS20240919204159.pdf
Were you ever not sour about “on-rail-competition”?
I mean its not wrong, on-rail competition is bad because timetabling is the most the basic element of output management for a train operators. Add in the trade-offs of adding services.
I also disagree with your dismissal of JR as “state-adjacent” I mean pretty much every large scale industry is by default, housing, commercial development, roads, construction etc etc. At a higher level I’d argue market-state distinctions are like matter-energy one with observer effects to boot. And Japanese rail operators are effortlessly more privatised than British style rail franchises where the government determines all major operational and capital decisions. The current fracas between Chiba city and JR East over the latter cutting expresses to Tokyo is case in point.
I mean JR success is because they replicated the existing model of the legacy privates onto the JNR. If Kintetsu could operate a big system of intercity, urban, rural services profitably for decades, it wasn’t much of a stretch to extend. Its pointedly not about on-rail competition, instead Japanese rail companies face competition on whole host of the levels. Most notably in urban Japan, the mess of operators have overlapping commuter zones sensitive to service quality/location/fare cost.
Furthermore the permissive planning system has allowed affordable low-segregation cities where people can and will move based on transit choices. The financial problems of the Hokusou, Chihara and Toyo Rapid lines not generating the expected population growth in their corridors stems in part from their high fare costs (Takagi Railway has recently done some English subtitled videos on this).
At the level of the Shinkansen, they face competition (which they mostly defeat on Honshu) from planes, cars and Highway buses. The struggles of the Shinkansen network in the late 1970’s and early 1980’s at JNR’s nadir when planes and the highway system really came online is case in point. And JNR’s financial troubles meant it had very expensive fares relative to the competition without the capacity to cut costs. The privatisation process allowed them to cut the fat while increasing service, the competition really did suffer, especially the Legacy privates in Kansai and Aichi.
I don’t think the Japanese model is necessarily worth copyable. My rule of thumb on railway privatisation is “have you privatised parking”
Except that Perpignan-Figueres was built, but the group that has failed to deliver Montpelier-Perpignan is SNCF, your supposed model owner-operator. I think you are confusing two different things: how much HSR should be built and how it should be run. Allowing open track access does not preclude building more HSR lines, just as building more lines does not require a state run operator.
Manufacturers touting cross border compatibility isn’t an open access issue; the lack of uniform platform gauge and traction gauge in Europe means this is a concern even with state operators. As we discussed before, SNCF had to develop multi-voltage train sets to provide service to Switzerland in conjunction with also-state-owned SBB.
Doesn’t Innotrans have segments dedicated to infrastructure and tunneling separate from rail vehicle technology?
I don’t think you are correct that airlines do not think about infrastructure. In some ways aviation has tighter infrastructure constraints than rail: things like gate counts for different ICAO groups, ramp strength to take the weight of group E/F aircraft (777, A380, 747), runway spacing, etc. all affect the type, number and frequency of aircraft operations and receive a lot of attention from airlines. The open access model of aviation did not prevent major international cooperation between private airlines / manufacturers and public airports to allow introduction of the A380, for instance.
I think part of the point Alon is implicitly making is that exclusive operators can capture monopoly rents and use those rents to pay for infrastructure, in a way that selling track access to competing operators cannot. Competition can mean better service, but it also means less money, and less money means less incentive for new services.
But does it mean less money ? IIrc open access only mandates to have no favoritism to access track, not that access costs should be low.
I don’t know the exact details, but I can’t help but see a correlation between the french having the highest access costs, and them planning several HSR lines in parallel -Perpignan, Toulouse, Normandy, Torino-
idk, I’m no fan of privatization, but I can see how it can be used to make infrastructure pay for itself in some conditions
Open access law doesn’t mean access charges must be low but this is what happened in Italy, HSR open access pioneer. The operators were losing money and the state bailed NTV out by lowering track access charges from 12.81 to 8.2 Euros/km in 2016, a value low enough to insure that RFI has enough to maintain the lines but no money left to invest. The three AVE operators lost money in 2023 and the new entrants are asking for lower tolls. They may not get them since the transport minister has accused Ouigo Spain of price dumping.
French high tolls are now largely congestion charges on the paid off core LGVs, and are needed to cover the loans on later ones. However, correlation between HSR construction and toll levels is lacking. Low toll Spain has the largest European HSR network and is expanding it. AV lines are also being built in Italy towards Bari, Genova and Verona, not in France where multiple planning is the rule but actual building only starts when all the stars are aligned. The transnational tunnel is the only active site on the Lyon-Torino project thanks to 50% financing by the EU and 29% by Italy. It will be shared with freight trains which typically cannot afford high tolls and SNCF is not financing it. Two projects need partial financing by the Occitanie region and it seems the Toulouse one will go first.
“I think part of the point Alon is implicitly making is that exclusive operators can capture monopoly rents”
No, because Alon is arguing that the owner and operator of the line (both track and trains) are the same. In that case monopoly rents don’t really exist because the entity paying the rent and the one receiving it are the same. If a homeowner rents their house to themselves for $1B/mo, that doesn’t actually give them any more money in the bank.
If anything a non-monopoly should generate more rent for the track owner, because the competition raises the possibility of different operators bidding up the price they pay to gain track access (increased demand for fixed supply raising prices). Alon is instead arguing the state should just give more money to build more track.
It is possible for monopoly operators to raise fares (to a point) because they don’t face competition, but Alon has explicitly criticized this in the context of Amtrak fares on the NEC/Acela, and instead argues for passenger fares per km being as low as possible.
Most places aren’t Tokyo or Manhattan. Or even Chicago. Most places aren’t the Tokaido or Paris-Lyon either. Most places are going to have more capacity than there are trains. Because they AREN’T TOKYO.
The monopoly is on the passenger side. Industry profits are almost always maximized with a monopolist.
Having multiple operators competing for track space will increase the profits of the infrastructure owner, but the total will still be less than if you have everything under one roof.
The only way this is not true is if you think the monopolist is not profit maximizing and allows costs to be inflated, if, for example they are lazy, or don’t actually care about profits (like non-profit hospitals).
@a eskpert
You are entirely correct on all counts. However, Alon is specifically advocating increased funding for new high speed lines. If multiple operators increase profits of the infrastructure owner that should be a good thing by his goal. As to a monopoly on the passenger side, Alon has consistently criticized Amtrak for having high prices because there is limited space on the Acela and so it wrings profit out of limited demand. He has expressly stated on many times that the maximum possible service should be run and flares should be flat at all times for a given distance to maximize ridership, not profit. Alon is not at all suggesting a monopoly to get greater revenue, he believes the additional revenue for more high speed lines should come from state funding.
Alon is suggesting a monopoly to get integrated service planning, both between services (the train From Berlin to Paris arrives in Frankfurt at 9:57 and the train from Stuttgart to Amsterdam arrives at 9:59 so if you get on the train in Stuttgart you can still go to Paris by making a connection) and between service and infrastructure (building a particular tunnel cuts trip time from 1:01 to 0:53 so the train that leaves A at 9:00 and arrives in B at 9:53 can be the same train that leaves B at 10:00, leading to more consistent service than if that train arrived at 10:01 and departed at 10:08).
@Adirondacker
The places Alon is talking about (Paris-Cologne, Frankfurt-Milan) and the kind of major routes where demand can fill the practical capacity of the tracks at certain times of the day, even if it does not fill the theoretical capacity. This is particularly true when you consider that HSR can have fewer trains per hour due to larger stopping distance required between them and that demand is not even at all times of the day and having plenty of room for trains to run at 11pm doesn’t help because that is when most people are sleeping.
>he
😦
If you don’t like Acela fares don’t take it. It there are alternatives. The Regional trains. The commuter operators. Buses. Either intercity or cobbling together a trip using the same agencies that run the commuter trains. Airplanes. Driving. Renting a boat and sailing. Because it’s not whatever forsaken desolate place you live where the bus comes three times a week it’s the Northeast Corridor.
Maybe in your imagination. One of the things Alon wrote is
Paris isn’t in Italy and it’s not in Spain. Neither are Cologne or Frankfurt.
or
Paris isn’t in Japan either. Neither is Cologne, Frankfurt, Milan, Buda or Pest…
Paris isn’t Barcelona and Madrid isn’t Tokyo. Or the center of the universe, Connecticut. And even at the peak of travel demand between Hartford and Stamford it’s not going to use up all of the capacity. Most places in the world are not the Tokaido or Lexington Ave. And just like red state Senators warp landing fees so they can get puddle jumper flights, whoever is regulating track access can regulate the prices. In the very few places in the world where there will be the kind of demand there is along the Tokaido.
Minimum sane interchange time between German high speed trains is probably 45 minutes if not longer.
Ummm, pretty exactly the opposite of a “model”.
SNCF management may choose to operate TGV service in mimicry of a flight level zero airline, and SNCF may choose to freeze out open access operators by every means, but those two separate organisational choices aren’t directly linked, other than by an generalised encompassing mentality of hostility to passengers.
It’s possible to be bad in many different ways at once; it’s also possible to choose only to be bad in some ways without also making the choice to fuck up in others. And of course, it’s theoretically possible to just, you know, Don’t Be Evil. Theoretically.
What innovation was there? Was there anything that was actually new, or is it just small updates on what we already have. I think what rail really needs is not innovation but just iteration. This is the same switch as last year, but we changed it to last longer; then next year, same switch as last year but we figured out how to manufacture it cheaper so the price didn’t go up by inflation.
It seems to be that full automation is a solved problems for metros. (but not trams or buses). What we need it to figure out how to make them for less $, make them faster (both building and running speed), and other such iteration.
“Real” innovation? I guess the sad hyperloop ~coach~ pod exhibited by TUM (Technischen Universität München) might count? (My understanding is that they don’t want to actually build revenue lines, instead this is a currently-cool project on which their students can learn how to be engineers.)
A few years ago, ÖBB’s capsule sleeper was a fairly non-iterative step; incidentally, this year Škoda brought an improved version (herringbone layout). There are several more, perhaps even less revolutionary, but as of yet unconventional ideas for interior layout geometry knocking about (1-seat or perhaps standing-only telephone booths? 2-seat compartments? Half-seats (I’m not sure what the correct term is)?).
A few hydrogen-combustion or fuel cell vehicles. Modular locomotives easily configurable between {genset-diesel, hybrid, battery}×{yes or no overhead electrification}.
I think, this is overly pesimistic article. There are cross-border projects in construction: Turin-Lyon, Fehrmanbelt tunel, Brenner tunel, rail baltica…
Most countries are not on the level of Switzerland, railway “culture” is very much “we always did it like this”. Some competition presure on state railway can make good.
I agree on what you said on airline immitation – very bad trend. Recently I traveled in Spain and although their high speed rail is amazing, overall traveling experience was not ideal. Lack of flexibility, confusing tarifs and route finding, time lost on security check, bad frequency, lacking connections and unreliable buses, bus timetables not matching trains.
So it is nice, that train goes 300 km/h but what if all places are booked or you miss a connection by bus ? Anyway, Spain infrastructure is amazing and we should learn from them!
There are megaproject tunnels, yeah, combining freight and passenger rail, like Brenner and Fréjus. But for the most part the approaches to them are slow (Fréjus is an exception, but as I understand it, the Austrian approaches to Brenner are at best medium-speed and don’t go all the way to the bigger cities), and there’s disinterest in workhorse lines that are not generational projects, like anything within the North European Plain.
The biggest issue with cross-border trains is the frequency. The only direct ones between the bigger European cities in different countries that aren’t at best the same as Fort-William Glasgow (i.e. 4tpd) are London-Paris, London-Brussels, Paris-Brussels-Amsterdam, Brussels-Cologne/Frankfurt, Paris-Stuttgart (but that is only 5tpd), Munich-Vienna and services from neighbouring countries into Switzerland.
I mean if we are really fair, and much as it pains me to say it Eurostar/Thalys are really pulling above their weight.
In the US, privately owned and funded airlines use public airports. Private vehicles use public expressways. Why can’t rail work the same way with public capital spending going into infrastructure and private rail companies owning trains, employing staff, and managing passenger facilities? The devil is in the details, not in grand declarations about the morality of public or private ownership in isolation from the rest of society.
I presume Alon’s argument would be that scheduling track space/time for trains, especially if you have a single track line, has much more intricacies than the other two modes. Planes and cars can pass each other without schedules.
Some airports and many expressways are chronically congested. That helps make the case for new public spending on new capacity. The argument for new rail capacity is much easier to make when there is demonstrated demand.
The appropriate comparison here would perhaps be an airport that is not congested by number of aircraft, but where e.g. the length of the runway limits the size (MTOW, whatever) — and thus, the range — of the planes that can fly there. There would be a respectable (though not breathtaking) level of demand for non-stop flights of a longer range. There are not, currently, such flights (obviously).
The problem — if it still needs description — is that service planning institutionally lives at the airlines, plural, and there don’t necessarily exist strong (or for that matter, any) communication channels from there to the organ that decides whether to build a runway extension. Thus even if everyone, from passengers through the airlines to the airport, would benefit from the extension, what is effectively a communication snafu (or even summarization snafu; each airline individually relays that it would be nice if the runway were longer, but this does not rise to the level where they would individually pay for the project out of pocket, and there is nobody anywhere who would bother to add together the per-airline Benefit:Cost ratios) prevents the project from happening.
A few, more granular details relevant to airline service planning are:
– jet fuel is a major part of their cost structure, and its price is fairly volatile;
– even if we pretend for calculation purposes that they own the appropriate fraction of the airports and ATC infrastructure, they are still opex-heavy.
Thus it is entirely appropriate for service planning to have a relatively short lookahead. It is highly probable that a few years from now, the price of jet fuel will have changed, thus writing a new service plan will be necessary. Thus it makes very little sense to “fit” infrastructure projects with multi-decade lifecycles around service plans with <5-year expected lifecycles.
Obviously, with rail the situation is exactly reversed. The capex of the infrastructure utterly dominates the cost structure. Thus to a first approximation, new construction sets some particular schedule in concrete for decades, because revising it would require further construction. (This is most dramatic with passing loops and overtaking tracks, where changes to the schedule change where these facilities need to be located along the line, versus where their presence is superfluous.) Therefore construction needs to be coordinated with service planning, which pragmatically speaking requires that these two functions (activities) be located in one organization.
Even in the countries with the best railway systems, there is limited public appetite for spending on new construction. Whether this budget reacts to demonstrated success, and if so, how, is irrelevant to the question of how best to use whatever funds are available for the purpose.
One other point – if you are not using the tracks a lot you can’t afford to build them. It is just barely possible to get an acceptable ROI if you are running 2 TPH (trains per hour) all day, but only if those trains are unusually full (at peak time you turn riders away), which suggests if the cities or the network grows at all your will need to upgrde to 3tph soon – but that means passing loops in different places. So realistically you should have at least 3 tph just to make a good ROI. You can’t take the above logic and say 4tph because once you reach 4 tpr you are building passing loops for most of the distance so you may as well just build two tracks and get the advantage that your station go were the population density demands it not where the need for trains to pass forces you to put one (you don’t have to put a station on each passing loop, but you must slow down and so it only costs you an extra minute to stop and open those doors so you may as well)
Note that in others thread many are talking about tracks running 4-5 trains per DAY not hour. If you are running that few trains on your track there is plenty of extra space – but you are not getting a good ROI for your track investments. They likely are making this work because the track was built in the 1800s and so the land is paid for (the expensive part), and so getting a good ROI is easy. However if you apply various formulas you often discover that despite only running 5 TPD, there is enough population in the various cities along the way to expect that you can fill 4 TPH, which means the current situation shouldn’t last.
“The problem” does need description. It’s not obvious. Airlines most definitely have influence with the public bodies that own airports in the US. Airlines are paying billions toward the new terminals at JFK in New York ,for example. Airports do assess longer term forces in planning runways and terminals.
Rail services are not set in stone for decades. Service schedules are changed all the time on the Northeast Corridor in the US. As many as three rail operators operate on some sections of the Northeast Corridor in the US now. There is no reason that a for-profit rail operator such as Brightline couldn’t be incorporated into the operation of the Northeast corridor.
Brighline isn’t going to be particularly interested in running empty trains from Perryville Maryland to Newark Delaware so I’m going to assume you mean running something to Manhattan. Whose standing room only train are you going to cancel to do this?
Brightline will put trains where it gets riders. The reason trains from Maryland to Delaware have few passengers is because the service and price are unappealing. There are many who would never enter an Amtrak train who would take Brightline; call them ‘The Divas.’ They actually care about customer service and comfort instead of operation agreements and efficiency. The Northeast Corridor isn’t at maximum capacity, even if some sections of it seem to be. This isn’t about dividing up the existing pie. It’s about baking new pies. Trains aren’t fundamentally different from cars or planes.
@henry, I think if you want a train service people will use you need two tracks the whole way. Otherwise you get lots of delays if one direction gets a bit late – or you have lots of padding which makes it slow.
Increasing ridership will make the case for rail bypasses or even a third track. Then the work of organizing the funding and construction begins.
There is almost no one in metropolitan Harve De Grace/Perryville and equally few in Newark. There cannot be “many”. I doubt there are many people driving between Aberdeen Maryland and Bear Delaware because there aren’t many. And since none of them have personal helicopters, none of them fly.
One train can serve a number of stations. Surely, you are being facetious.
If the train stops someplace where almost one lives, then a place where there isn’t many of them at all and then someplace with even fewer for a dozen or so stations it’s still almost no one. They all own cars, there is no congestion because there aren’t many of them. And because there aren’t many of them there is plenty of parking. Except for people who can’t drive for some reason they aren’t piss away a lot of time getting to and from a train and they will drive.
Train operators KNOW if the train stops someplace where almost no one lives, then a place where there isn’t many of them at all and then someplace with even fewer for a dozen or so stations it’s still almost no one. Brightline wouldn’t have a statutory obligation to stop at all of those stations. It would stop at stations where there was demand. If people preferred Brightline services they’d take it instead of Amtrak. Amtrak would have to withdraw service for lack of use and Brightline would be able to expand that service. It’s not a magical panacea, but it is a believable path to more passengers. More passengers means more support for maintaining track and facilities. Better track and facilities means more reliable trains, etc. The circular debate here seems to say that the only way to improve passenger rail in the US is top down; that changes must be imposed from above through elaborate command and control. That may work in France. It doesn’t work in the US.
You have a surprisingly cynical attitude about trains for a forum focused on trains.
The station where their is demand is at capacity and the only way for you and Dagny Taggart to send the valiant Brightline train there is to cancel some other standing room only train. Which one do you want to cancel?
Exactly. Amtrak trains will be cancelled and Brightline trains will take their place. In the longer term, the greater demonstrated demand will make it easier to get increased public spending for new capacity with additional track. There is still a large role for public spending, even if it doesn’t go through Amtrak directly.
I want to know which one. In other words who are you going to piss off by canceling their train?
It all works out in Freemarketland. In the real world you are going to piss off people, who know who their Senators and Governor are, when you cancel their train.
Who decides which flights are cancelled at airports? Who decides how many lanes a highway has? Who decides to stop flying to a particular airport? Those decisions piss off a lot of people.
Henry: I was particularly thinking of branch line reactivations (for passing loops — usually the track needs complete rebuilding even if some freight traffic survived to this day, but the right-of-way and the embankments/cuttings generally exist) and mainline commuter R-bahn service (for timed overtakes). 30 minute takt is very common for both. Incidentally, stops on single-track segments are very elegant: serving both directions with a single platform edge, they can also provide cross-platform transfers to “local” buses (such as they are in the formerly-rural “commuter villages” landscape, as it occurs in Europe).
Matthew Hutton: that’s exactly the reason to pay attention to the delay-minimization practices written about by Alon, Clem Tillier (caltrain-hsr.blogspot.com), Hans-Joachim Zierke (zierke.com/shasta_route) and others.
Matt: “Amtrak would have to withdraw service for lack of use […] More passengers means more support for maintaining [and improving] track and facilities. [The consequently] Better track and facilities“
That is indeed how things ought to work in principle. Unfortunately, in practice many organizations that touch passenger railroads in the US flat-out don’t work that way, not just in the sense of ordinary imperfection or incompetence, but perversely so.
– Amtrak continues to provide plenty of barely-used services. (NEC fares are so high partly to subsidize these exercises in money-burning.)
– There is indeed pretty good political support for maintaining and improving the NEC, with corresponding funding. It has not resulted in improved service.
– Adirondacker was making the point that by far the most lucrative option(s) for additional service (…since, by assumption, the timeline is that Brightline trains are added first, Amtrak trains are emptied second, Amtrak trains withdrawn third, right?) for e.g. Brightline would be direct Acela/Regional competitors: intercity trains into NY Penn. Which is, currently (due to poor choices in what infrastructure to build, no less!), more or less at capacity with Amtrak and NJT commuter trains. Thus, unless/until new capacity is built, a hypothetical competing Brightline service into NY would require the withdrawal of some existing train service.
Alternately: in principle, trains aren’t that different to road vehicles or planes. If a (suburban) road is congested with cars, it is relatively easy, politically speaking, to get additional car lanes built. It is likewise easy to start a (mixed-traffic) bus service to compete with cars, in the anticipation that people who currently drive will instead opt to use the bus, and this will clear up the congestion. It is rather more difficult to add a dedicated, bus-only lane to the road, to say nothing of kicking cars out of an existing lane to convert it to a bus-only lane. Likewise, managing traffic by surrounding the destination with a congestion charge zone is politically difficult. You might notice a pattern here… and why people would be cynical about the road-related institutional structure (bus agencies, DOTs, politicians, consultants, etc.) as a whole. It is also surprisingly difficult to fix bad practices in bus networks — overly dense stop spacing, winding routes, etc. (I trust you are familiar; I’m happy to explain if not) — even if they would result in more ridership, thus simultaneously providing more public benefit as well as a better ROI to the operator. Precisely none of this means that the cynics think little of buses, or that standard economic considerations don’t apply to them. (If I may turn the comparison on its head: in the United States, planes are — not fundamentally, but for the time being — different from other modes of passenger transportation, inasmuch as the airlines, the FAA, etc. are competent.)
The airline. Usually because it’s going bankrupt, Again. When someone complains to an elected official he or she throws up their hands and exclaims they have no control of the free market.
There are no ‘free markets’ without regulation. That’s how all other transportation in the US works.
If you are so well versed in how things work, why did you need information about who cancels flights? You still haven’t answered which standing room only train you are going to cancel so the valiant free marketeers at Brightline can stun everybody with their offering?
I don’t need that information, you do. You are the one who asked. You just don’t like the answer.
You are the one asserting it will be rainbows and lollipops is a standing room only train is canceled so Brightline can run a train, not me. You are the one asking why and who cancels flights.
No I’m not. Are you an Amtrak employee? Why are you so threatened by new passenger rail arrangements in the US?
Yes you did
And you lied about the Northeast Corridor being at capacity.
It doesn’t seem to be, it is. It’s why there are standing room only trains. I still want to know which one you want to cancel so a different train can use the tracks.
You seem blinded with rage that anyone would question you in any way.
I’m not an Amtrak employee and I’m not blinded by rage. I’m not even mildly perturbed. I’m just pointing out you aren’t living on Planet Earth and then lie.
Such rage….
The integrated rail companies located in Japan, Switzerland and the freight specialists from North America and Australia are the most efficient operators. The EU’s plan to shake up state owned railways by separating operations and infrastructure was a bad idea but is not the end of the world.
The LGV project between Bordeaux and the Pays Basque is slowly moving forward with an opening expected in 2035. France is counting on a very Spanish 20% financing by the EU. Alta Velocidad Espanola should reach San Sebastian in 2029 with a French style partial devolution to the Basque government.
You were disappointed because infrastructure managers like SNCF and Adif did not pitch their expertise, but it doesn’t mean that they are not interested. InnoTrans may not be the right place. High speed rail managers won’t be selected because they are technically the best but because it is diplomatically expedient or their state of origin comes with an attractive financial package.
Competition can be harsh on the operators. In Spain, Iryo lost 79.9 million Euros in 2023 on a modest 65% seat occupancy ratio, Ouigo lost 42,7 million despite selling 80% of its seats, and RENFE 65 million after selling 70% of its seats. That probably means more money for Adif and many happy new AVE costumers, an important constituency when a government considers investing in infrastructure.
Japan and Switzerland are vastly different from the US. Changes in the US can only be understood in relationship to the current situation in the US, not to any other countries. The variable are simply too many to account for. Does “integrated” mean everything owned by the same organization/company? How do you define “efficient?”
Integrated means that the railroad owns and maintain the tracks on which its trains are running. This was how railroading was done until British Rail privatization or EU mandates separated track ownership and rail operations.
Efficiency can be evaluated by looking at productivity, i.e. how many passenger and ton-km/employee a railroad manages over a calendar year. Last time I looked was over ten years ago. Amtrak was at 0.5, SNCF 0.59; CFF 1.09, BLS 2.08, Japanese average was about 1.94, China 2.46, US Class 1 16. China benefits from carrying so much coal. The Class 1 do so well because they specialize in long distance traffic but the US short lines are efficient in their own way. They can survive and pay their bills on marginal lines where larger companies would lose money. That is also why I said the Swiss were efficient. When all other railroads abandoned the small rural branches to concentrate on the trunk lines, the Swiss never gave up.
The numbers are for millions to ton or passenger-km per employee.
Is “efficiency” the CAUSE of high ridership or the effect? What transportation alternatives are there in Japan and Switzerland? Are expressways toll? How much are cars and fuel taxed?
Cause or effect is a good question. As you know, high tolls and various regulations raise the cost of car ownership and operation in Japan. In the metropoles, a lot of the salarymen and women are a captive clientele for railways which nevertheless very efficiently use their limited footprint. As John D explained, the answer is both.
Switzerland doesn’t have the same obstacles to car ownership or operation. The annual fee allowing unlimited use the highways is a modest $56. A few years ago, Swiss rail share of distance travelled was 22% for those earning less than 4000 Francs a month, 24% for those earning more than 12000 Fr. People don’t ride Swiss trains because they can’t afford the alternative but because the service on offer is excellent, including in rural areas long abandoned in other countries. When it comes to passenger rail efficiency, the answer is “cause”. For the freight, the answer is both. Domestic rail service is extensive and still includes single carloads largely abandoned by neighboring networks. Transalpine service has a high market share (60%) thanks to policies discouraging lorry transit.
Both. High ridership from the contextual factors you mentioned spurs operators to develop technologies and practices to deal with the traffic, which results in a high quality of service that retains and even encourages ridership. Granted, not all countries will have this virtuous cycle, but the absence of part one does not preclude learning from part two.
Pretty much everyone in the Chilterns who is using Chiltern railways can afford a car and uses the train because the service is good.
Yes, the Chiltern Railways have done a good job but note the contrast with Switzerland. The 1960’s electrification of the West Coast Main Line was seen as an opportunity to downgrade the parallel Chiltern Mainline and close the Great Central Railway with the exception of its termini, Manchester London Road now Piccadilly and the London segment towards Marylebone whose closure, decided in 1984, was never enacted. The Chiltern Railways have been able to restore platforms and the second track on the downgraded areas. HS2 is trying to provide, at a huge cost, the capacity lost when the Great Central Mainline was closed.
This steadiness distinguishes the Swiss railways from the rest. This is why they lead the world with rail electrification to the smallest rural village, when Marylebone is still 100% diesel, and a 2023 ridership of 2505 km/inhabitant, almost three times as much as the UK’s 883 km.
Britain has been 100% wrong on electrification. We are trying out battery trains in Liverpool!
Not 100% wrong but all over the place. While Switzerland, devoid of fossil fuels, had no place to look for energy besides hydroelectricity, the UK has been an energy superpower. It started with coal, still the source of 67% of electricity production in 1991. The UK was a pioneer of nuclear electricity generation before finding oil and gas in the North Sea, becoming forty years ago the third largest oil exporter and leading to a Dash for Gas in the 1990’s. It is now exploiting its abundant wind, at a high cost in transmission lines, since the windy places are distant from where the demand lies.
The railroads were affected by this. Electrification was the obvious option in Switzerland but was stop and go in the UK: no electrification for decades, then an intensive program the supply chain cannot fulfill at a reasonable cost.
I’m surprised to see arguments that individual transportation costs don’t matter in public transit use, funding, and ‘efficiency.’ If US states dramatically increased the cost to register a car, taxes on gasoline, and introduced tolls on all expressways, millions of americans who have never even thought of using trains would suddenly become interested in doing so. It would create a whole new constituency for passenger rail in the US.
I don’t see anyone making the argument that individual transit costs don’t matter. Alon stated on this page that one of the reasons why the TGV ridership outperformed the Shinkansen was lower fares. I wrote that high costs made car ownership impractical for many Japanese. Americans and Canadians would respond to the same incentives but only to a degree because suburban sprawl favors individual transport, wide urban highways are the norm, and transit construction costs, the basis of Alon’s work, are out of control.
This is unthethered insane blathering. Typing because you like typing.
We don’t need to speculate what might happen if “US states dramatically increased the cost to …”.
We know what actually happens: Vilification, unlimited lobbying and vote-buying expenditures, voted out, all meagre traces of “baby steps” policies erased.
So, you’re agreeing with me, Richard.
I’m not talking about an alternative future in the US. I’m talking about an alternative present. None of these things are going to happen in the US. I’m just trying to understand what could have been instead of what will never be. Moderately improved commuter rail, more transit oriented development, and better connections between different forms of transportation are the most we can possibly expect in the US given it’s governmental structures, cultures, and economy.
The comments about HSR and Amtrak are timely.
Amtrak just announced it is going to restore service between Chicago and Miami (The Floridian). 47 hours. The route is not like you would choose when driving (Toledo, Cleveland, Pittsburg, Washington DC, Columbia SC, Jacksonville, Orlando). Google Maps says it takes 20 hours driving time on its route (Indianapolis, Nashville, Atlanta, 1400 miles) – and says that Flixbus can do it in 27 hours.
Apparently temporarily while the Silver Star cannot originate in New York. Calling it the Floridian will get a few especially frothy railfans to make the trip.
https://media.amtrak.com/2024/09/amtrak-launching-the-floridian-with-daily-service-between-chicago-and-miami/
Pre-open access Hull-London had 1 return train per day, now it has 8. That is a huge improvement.
And frankly if an open access operator put on 8tpd between Barcelona and Montpellier that would be a huge improvement as well.
Fundamentally only really Britain, the Netherlands, Belgium, Germany, Switzerland and Austria and Japan, Taiwan and South Korea are running an acceptable train frequency in general, and even there there gaps like London-Hull.
No one else really manages it.
Hopefully, they haven’t made the same mistakes as Britain made. Our rail privatisation was very badly thought out with train operations, train ownership and track all split apart.
Fans of privatisation like to say that passenger numbers increased hugely under privatisation, but the fares also increased way above inflation, tickets got less flexible, the trains got less comfortable and only very slightly faster in a few places. e.g. 20 mins off a 5 hour journey. The sandwiches got very slightly better but this seems unlikely to have attracted people to using the trains.
My hypothesis for what has driven passenger numbers on long haul routes is: increased private car use causing congestion for people driving on longer journeys making the drive more unpredictable and stressful; parking restrictions making parking in cities expensive particularly in London meaning that it is inconvenient to visit your friends and family there by car because you have to pay to park in the street outside their house using a complicated permit system; more thorough driving tests increasing the cost of learning to drive out of reach for young people. I looked at taking my wife and kids to London to visit family by train and it was going to be £520 return booked two weeks in advance (that was with a discount card £800ish without) compared to about £180 for the three tanks of fuel for a journey of about 480 miles each way and only an hour or two slower door to door. Clearly something is forcing people to take the train despite privatisation not because of it.
On top of all this, for short haul routes house prices have been increasing way faster than inflation and wages for years and therefore pricing more people out of cities and onto commuter lines.
Privatisation did nothing to drive down the cost of rail travel to passengers. It also made the whole system significantly less flexible and more bureaucratic because any change requires renegotiating a three way nest of contracts between the operator, the owner of the rolling stock and the owner of the track. So if you wanted to upgrade signalling on the main West coast trunk route between London and Glasgow to digital control with something like ETRMS you have to negotiate with the ~20 train operators that use that route, the various owners of the trains that need to be upgraded with new equipment and the owner of the track to fit the equipment. For every electrician out there on the track fitting equipment there are probably 3-4 managers talking about who is paying for what and whose responsibility it is if things go wrong.
(Rescued from spamfilter, sorry it took this long.)
The service frequency in Britain is much stronger than almost anywhere else and that improved hugely under privatisation.
In France there are very few places the size of e.g. Bristol that have half hourly service to London/Paris, and the 20 minute London to Manchester frequency is only bettered really by the Tokaido Shinkansen – but that involves far larger cities.
The off peak tickets are also pretty good value given their flexibility, and the advance fares to London can be pretty keen compared to other European countries. An ICE any train ticket certainly isn’t cheap.
And yeah the French trains are more reliable, but the German trains certainly are not.
Yeah but service frequency on a congested railway increases because centralised planning and investment makes more paths and more trains available. We don’t know if this would have happened anyway under British Rail but it would be safe to assume so because they did it for the East Coast line with the intercity 225 programme where they ordered enough units to upgrade service frequency on the East Coast. In most cases the number of trains to be purchased is ultimately decided and funded by the treasury department at the government and somehow they end up in private ownership and are then leased to the train operators for vast profits. It’s not a real free market, just a very bureaucratic and inflexible way of buying public transport.
Advance purchase tickets can be good value, but they are like gold dust. Good luck trying to find any if you need to travel with less than a few weeks notice, or if you’re trying to travel in peak hours. Generally you can only find them if they are at an inconvenient time. E.g so early you need to stay in a hotel. I don’t think it’s fair to include them in comparisons of average prices unless you have data on how many are available compared to normal tickets. I suspect token amounts.
UK commuter trains are also expensive. Compare the cost of a return from a town in the Home Counties with Paris and the uk price is extortionate. I searched for a couple of minutes looking for travel in the next few days travelling one day and returning the following day. I saw two singles at maximum price give maximum Chartres-Paris return £32 (€19.20 each way) but I saw as low as €13 single. Return Maidstone to London; Anytime return seems to be £52. Off peak seems to be £46. But could possibly be as low as £33 if you travel in the middle of the day. Colchester-London prices are pretty similar. I didn’t check other routes but I’m sure they’re similar.
A thirteen hundred seat Shinkansen every few minutes is the most efficient way of carrying millions every month.
While railway operators look at the Tokaido Shinkansen operations as an achievement beyond their reach, they look at high frequency on the congested UK network as a feature they would rather avoid. In 2022, UK trains carried an average of 115.6 passengers, below numbers in Germany (129), Italy (136) and France (258). This was not a deliberate decision of UKs operators. The twin deck commuter trains common on the continent don’t fit in the UK’s loading gauge. Few UK platforms accept 400 meters long trains. No one can be blamed for that historic legacy. The UK network was built at a time when engines were less powerful than when railroads were built on the continent or in America. On top of these inherent inefficiencies, privatization may have added some as you explained earlier and fixing the issue will be very costly given UK’s construction costs. High frequency is not a bad thing in itself but I believe most would rather have Long Distance French prices even if it meant less frequent trains.
I would argue each train carrying fewer passengers is actually good as it means you are offering a high frequency service.
If you have the track capacity for it then a half hourly 2 car train service is better than an hourly 4 car one.
Yes it is better but it is also more costly than longer, less frequent trains. The question is whether customers are more interested in lower prices than in low waiting time. The answer will depend on the thickness of their wallets and on the distance travelled. Low frequency long trains as seen on the Trans Siberian or on 2 to 3 miles long US Class I freight trains will be preferred on long distance but would be inacceptable on a commuter service.
On a new subway line, automatisation will allow higher frequency at a lower cost than with longer stations and intervals as shown by Lausanne but that is a different situation.
We know that frequency is super important because the faster, cheaper, more reliable but less frequent TGV has less passengers than the slower, more expensive, less reliable but more frequent British long distance train services.
Should I believe Matthew Hutton who believes frequency is super important or M O’Leary, Ryanair’s CEO who thinks not? Ryanair serves thin routes with frequencies as low as once a week when legacy carriers used smaller, more frequent flights. Ryanair has become the world’s number 1 airline in terms of international flights because low costs seem to matter more than anything else to a majority of Europeans.
Richard Voight, an economist at the Federal Reserve Bank of Philadelphia, evaluated the elasticity of demand in response to changes on the SEPTA network. Although these commuters are more sensitive to train frequencies than long distance ones, they were still more sensitive to prices (long term elasticity of |1.84| meaning the agency was losing 18.4% of its revenues if it raised prices by 10%) than to lower speed or frequencies. This may explain why the low cost/low frequency TGV has a higher ridership in terms of passenger-km than the sum of long distance, regional, London and South-East British trains.
(Bad spamfilter.)
There’s a lot of literature out there about transit ridership elasticities, and the fare elasticity is not -1.84; the long-term elasticity is a bit weaker than -1 usually, and also a bit weaker than the long-term elasticity of ridership with respect to service.
I’m sure there are many.
The Camden and Amboy – for service between Philadelphia and New York – opened a few years after the Stockton and Darlington. Chartered at the same time as the opening of the Liverpool and Manchester. There are 400 meter platforms scattered all over. Clearances are a bit tight in the North River tunnels between New Jersey and New York so they had to make some compromises with Bombardier’s bi-levels and call them multi-levels. So there are 300 meter double decked commuter trains and up until recently 400 meter long distance trains on railroads that have been around almost as long as the ones in the U.K. I’m sure there are lots of people to blame. And lots of excuses.
I was speaking to Mark Smith from Seat 61 and he failed to ask Adrian Shooter of Chiltern railways this question when he shared a taxi with him. That said he did say that he didn’t ask because the answer was obvious and that the privatisation did increase frequencies.
That said the benefits of privatisation were definitely front loaded and there have been significant costs too in that the railway has become more expensive to run for a given number of passengers.
Privatisation has also typically been bad at arguing to build new lines and capacity rather than simply using up existing capacity. Chiltern was probably the best at adding extra capacity but it didn’t four track between Gerrards Cross and the tunnel into High Wycombe and it didn’t do electrification.
I think you summarized the problem with UK rail privatization. By 2017, UK train fares were 396% of 1987 prices while overall car ownership costs, bus fares and retail price inflation were 246%, 464% and 272.5% of 1987 numbers. Rail passenger traffic increased because the factors you described created a captive clientele. A study looked at the costs distribution and concluded that operating costs excluding track access charges were marginally below the average found on the continent (SNCF, SNCB, DB, NS, DSB) despite the added costs of the ROSCOs but that Railtrack was inefficient. Overall costs per passenger-km were 20.2 pennies against about 12p on the continent. JNR privatisation did not separate track and trains ownership and lowered costs by increasing productivity. Fares which had been going up rapidly during the last JNR decade were blocked for 8 years with labor costs going from 73.6% of fare income in 1985 to 35.1% in 1994. Over 10 years, the newly privatized JRs almost reached the productivity levels of their private competitors.
But then Shinkansen fares are much higher than TGV and ICE fares per p-km.
Yes, Shinkansen fares are higher than on the TGV or ICE but everything is relative. The Shinkansen benefits from Tokyo’s restrictions to car ownership, and doesn’t have to compete with free Autobahns or ultra low cost airlines. The JRs can charge more to subsidize loss making rural lines. Furthermore, the JRs other than JR Central may not have paid the full cost on the new trunk lines but they do not receive the various subsidies available to DB and SNCF. Acela’s high fares do result from Amtrak’s inefficiencies. I am no Japan expert but I don’t believe the same is true of the Shinkansen.
Sure, but note that if you train a ridership model on the Shinkansen and ignore all of these factors (as well as the fare factor), the TGV greatly overperforms it.
From what I understand, Shinkansen fares are high because they can afford to be – the service level (despite 10tph Nozomi on Tokaido) is low compared to the corridor population. Unreserved carriages can resemble London’s Victoria line in terms of crowding in peak season, and getting a seat is so uncertain that most people pay extra to have reserved seats even on a bus-frequency service. Posters on this site have cited very high share of business travellers – suggesting the leisure travellers are priced/crowded off Shinkansen services. It all points to a very suppressed market so you would expect a good European system to outperform a Japan-calibrated model, and crucially a european system needs to cater to a bigger leisure market (esp old grannies with heavy suitcases) in terms of dwells and turnarounds.
Almost all your points are wrong or miss the point.
At the end of the day what is important is that more passengers are being carried. That these passengers are paying more just reflects a willingness to pay and there’s nothing inherently wrong about capturing it. I’d rather 1000 passengers paying £50 each than 500 passengers paying £20 each.
The important thing was railway operation was taking out of the Treasury. Shareholders are bolder at providing finance to marginal revenue > marginal cost initiatives. There may well be truth that there were underlying demographic factors contributing to a general demand increase, but the important point is that shareholders were better at responding to that and enabling that higher rail trip rate far better than Treasury on its own would have. To the extent that a 50% utilised railway soon became a 100% utilised railway. Now that the 100% utilised railway has been baked into the new base (even accounting for Covid wobble) you could argue that shareholders have outgrown their usefulness.
For all the ills of rail in Britain, it is a poor imitation of Dutch and Swiss railways, but an imitation nonetheless. The old BR under the dead hand of the Treasury would have remained an imitation of SNCF (though probably a slightly better one).
Rail in Britain has not become less flexible. Any Permitted tickets remain the default (apart from the naughty experiment from LNER which incidentally is fully state-owned) – there are other choices (Advance fares) that you are free to ignore. All French/Spanish/Italian long-distance fares are essentially Advance fares. Anytime Fares are a bit ridiculous but they produce TOC premiums which DfT can reinvest into the railway. Off Peak and Super Off Peak fares remain excellent value.
In Spain you can travel fairly cheaply in a guaranteed seat, but only in 10 days’ time. In Britain I have the choice to pay £300 to stand in the vestibule if I have to travel in a moment’s notice. The British model is the far smarter model. It’s better to have 1000 grumpy passengers than 500 happy passengers and 500 people you simply refused to serve.
I somehow feel the core problem is national borders. SNCF could do cross border lines in cooperation with weaker states which would basically allow french trains on their tracks (Switzerland, Benelux), once you have a serious state on the other side of the line (Italy, Germany), things got complicated. How long did we have to wait for Paris – Berlin high speed train? There was a Paris-Geneva TGV in 1984.
The alternative is having a supra-national system, like the TEE of old. Initially this was supposed to be a separate entity with its own rolling stock, the only cross country train was the RAm TEE, Every country instead had its own TEE trains. France did not even adopt the common livery and claimed the EMU model was not flexible enough so the first 4 voltage EMU, the RAe TEE II was replaced by a locomotive pulled train, with locomotive changes at the border.
Half a century later, everyone is selling their modern RAe TEE II…
The solution to the various Länderbahnen acting uncoordinated and having occasional pie-fights was to merge them into one single DR. Analogous things happened in the other European countries, mostly. Pretty obvious that today the corresponding solution is to merge them into one single European Rail.
The French situation was pretty different, as the network had been divided in largely disjoint regions by king Louis Phillipe, SNCF was founded late (1937) and the goal at that point was for the state to take over money loosing lines. There was enthusiastic competition in Switzerland, but in typical Swiss tradition, the solution was a coordinated approach while keeping many local semi-independent operators, while keeping some internal competition (BLS on the Lötschberg, SBB on the Gotthard).
This is why I find the Léman Express, the S-Bahn network around Geneva, very interesting: you get a miniature of the problem. The French part of the network was very neglected, the only reason the line to Évian survived was because of the trains carrying water out of the bottle plant. Basically, the Swiss built a new underground connection between the main station and the Annemasse station, the local hub in France. The line existed up to an isolated station in Geneva (Eaux-Vives) but was not connected, and was operated by SNCF, so service was minimal. Most of the work was done on Swiss territory, the center of gravity of the network is definitely Geneva.
The sane solution would have been to let SBB handle the lines to Annemasse and Évian, they already handled the ones to Bellegarde, which were quite appreciated, if only because they were not affected by strikes and the trains on the french side up until then were a joke. But no, this is France, so they created a two-headed monster called Lemanis. Lemanis does not own trains.
The trains are either SBB stock (Stadler Flirt) or SNCF (Alstom Regiolis), both train-sets have to handle different voltages, different signalling, different information systems, different drivers, with different procedures, incompatible waste-water evacuation process. The trains have sticker with different emergeny numbers depending on which country the train is. SNCF of course blocked the initial train sets from Stadler because they only complied with EU regulations, and during the transition period, they could assert primacy of French procedures. The Swiss managed to get one track in Annemasse which is 15 KV, so regular KISS trains can go there.
Now they are talking about rehabilitating the track between Évian and Switzerland at the other end of the lake and it is the same circus again. Of course, Switzerland is going to pay (region is spending money on a highway), but one point of discussion is of course if the track should be a Swiss or a French track.
I think a lot of it is whether countries are prepared to just do things the French way.
Eurostar is a better service than the France-Spain ones even though Britain is more powerful than France – probably because we were prepared to cave to the French on some of the operations stuff and/or that London is so important that you could do the discussions at a level where SNCF had to cave.
The challenge with a pan-European train agency would be that it would prioritise the biggest cities over everyone else.
Actually if the French were forced to allow open access properly you could probably get Hull Trains/Lumo or Grand Central to run a Lille-x service which while Britain centric would improve things.
I think the passage of time also plays a role, to some extent, the French political mind-scape never realised that something happened in 89. In the 80s the TGV was something incredible in Europe, and getting Swiss cities connected to Paris, or even Marseilles was something very positive. But the 80s are over, most countries have one form of high speed train or another, and France, which used to be very much in the center of Europe is now very much on the western edge (in particular since Brexit).
The problem of a pan-European agency would be the same than at the time of the TEE, national companies not wanting to let go of their home-turf and their national industries…
I didn’t mean an agency that exists alongside (or above/below) the national companies. I mean that the national companies should cease to exist, being merged into the European agency/company. This can happen either suddenly (Germany, Britain twice, etc.) or piecemeal over perhaps decades (France, Hungary, etc.).
The Paris-Berlin train is still a symbolic gesture instead of a useful service. It’s one train a day, with an 8 hour trip time, that saves almost no time over changing trains in Frankfurt. And waiting for the single daily train can effectively be much slower.
The Paris-Frankfurt high speed train has been around almost as long as LGV Est has.
The reason why a Paris-Berlin high speed train took so long to be created, and isn’t terribly useful, is mostly internal to Germany. Even with the political issues around cross border high speed rail, Paris-Frankfurt has a slightly faster average speed than the purely domestic Frankfurt-Berlin.
If Paris-Berlin could be a 3 hour trip like Paris-Geneva, or even a 5 hour, the high speed train would have existed a lot earlier, but Frankfurt-Berlin is alone is almost 4 hours.
Google says the road distance between Paris and Berlin, via Frankfort, is 1,125 km. If that take 8 hours it’s not high speed. It’s the speed of Amtrak between New York and Washington D.C., give or take a few. It’s not going to be 3 hours because it’s 1,125 km. …. why can’t railfans understand that if there is a wider space between the dots on the map that means they are farther away…. 1,125 km at an average speed of 250 is four and half hours.
Paris-Frankfurt is still offers a sub-Fort William service level though.
The first train from Paris to Frankfurt leaves after midday, and the other way there are no trains between 8:30am and 6pm. It’s a joke.
Your story about stronger and weaker states is just nonsense. TGV operations are by law a commercial (non-subsidized) entreprise. TGVs have been visiting small and large countries alike for years, and that includes Germany because 3h05 to 3h25 trips between Paris and Freiburg im Breisgau, Stuttgart or Cologne are a competitive offer.
The TEEs were a symbol of European rail cooperation but that doesn’t pay the bills. In 1974, the Zurich-Brussels Edelweiss switched from diesel to 4 voltage Swiss EMU. This TEE was not discontinued in 1979 because the SNCF objected to the EMU but because it was almost empty. The business clientele was no longer putting up with a seven hour trip. It was transformed into a pulled electric train with 1st and 2nd class carriages catering to a different clientele. Nowadays the business and budget clients fly direct between Paris and Berlin on Air France or budget airlines. If Lufthansa can’t even bother to offer direct flights, who is going to fill up 500 seat TGVs on a 8 hour trip ?
According to the French Wikipedia, service on the Cisalpino was switched from using the RAe TEE II to pulled Mistral 1969 carriages in 1976. I see no mention of second class service on this TEE train. The reason that is given for the switch is too many passengers. The service stoped in 1984 (8 years later) with the arrival of the TGV which took over the Cisalpin name for the route up to Lausanne.
Lack of capacity on a route with seasonal variations in demand was one of the reasons behind the Cisalpino switch to new pulled “Mistral 69” first class coaches on the same overall schedule on the 25 and 26 May 1974. The second one was not SNCF’s wickedness. It was that the Werkspoor/BBC/SIG diesel trains used since 1957 by the Edelweiss TEE were fatigued. One had 4 030 200 km on the odometer. The route was fully electrified since 1960 and switching to the ex-Cisalpino RAe TEE II shortened the Zurich-Brussels trip by 12 minutes despite one additional stop.
«Au début des années 1970, la SNCF reprocha aux CFF le manque de souplesse des rames RAe TEE II, indéformables et incompatibles avec tout autre matériel ferroviaire. S’il fallait renforcer le train lors d’un pic de trafic, il n’y avait pas d’autre solution que d’ajouter une seconde rame, la capacité du train passant ainsi du simple au double. Il fut donc décidé d’assurer le Cisalpin dès le 26 mai 1974 avec une rame tractée composée des nouvelles voitures Mistral 1969 pour mieux répondre à la demande croissante¹.»
The wikipedia quotes a paywalled article from Ferrovisime.
Hull trains is a commercial unsubsidised profit making enterprise that has 7 round trips a day between Hull and London with a further round trip operated by LNER making 8 in total. Why can’t at least the same service level be managed between Paris and Frankfurt or Paris and Stuttgart?
Yes it goes across an international border, but Frankfurt and Stuttgart are significantly bigger and richer than Hull.
Exactly..
The question is who funds and manages the supra-national operation or PSO authority. EU institutions are not funded or set up to manage operational endeavours. EU institutions are only funded and given mandates to pass and uphold EU-wide laws. They have no means to take revenue or fiscal risk for a commercial/operational undertaking.
If you were to do cross-border services properly you’d have neighbouring countries set up a joint PSO authority funded and governed by treaties – a bit like how German states set up multi-state authorities but at a proper international level – but we all know how that’s likely to go with some countries. Failing that non-discriminatory access is the only way.
In your specific examples – I’d say it’s the other way round. SNCF leans on the competence of the Swiss, Dutch and Belgian railways – notice how those trains are all clockface unlike SNCF’s domestic operations – and I think those authorities are also proactive in enabling those trains by helping SNCF – they are not ‘weaker’. Germany are not quite as good at timetabling and infrastructure capacity as the other examples and SNCF just puts it on the ‘too difficult’ pile.
I have spent a bit of time looking at the UK overseas travel data. If for the sake of simplicity we say you need 200k journeys a year or 100k return trips to justify a daily return service then the obvious place that deserves a daily connection to London is Barcelona/Madrid. Yes it’s further but the market is 18m trips a year.
The second one probably is also justified as there are 6m trips a year is Italy – especially as a single service can serve Turin, Milan, Florence, Rome and Naples without too much trouble, a 2% market share from the UK would be enough to justify a daily round trip from London.
Difficult to see anything else being justified. Switzerland only has 1m trips a year so you would need a 20% market share just to have one daily return service to Zurich and one daily to Geneva. Perhaps with the Low Countries as well you could double that – but it is still all far worse than a 2 hourly service from Paris – and 20% seems very high.
Germany I believe has only 3m UK visitors a year – but also lots of potential destinations that aren’t in a line. Undoubtedly better to simply offer better connecting services from Brussels/Paris Est.
Unfortunately the UK has made itself outside of the scope of any discussions on supra-national EU-level discussions.
Running one train per day is unlikely to be competitive against air to fill a train, unless the total market is so big that a small market share still lets you fill the odd train (I think Paris – Barcelona does fit into that category). UK’s border requirements render anything other than strong point-to-point markets uneconomical.
In a more enlightened parallel universe rather than being a tri-partite company, Eurostar would have been a tripartite PSO authority letting an operating contract with the dual expectations of offering an all-day clockface timetable on all routes and receiving a premium payment.
Longer-distance weaker markets would then be catered for through connections (and ticketing which is a can of worms I won’t open). For example, London – Brussels and Paris – Brussels would both be half hourly, arriving Brussels headway apart on an island platform. One train goes to Amsterdam and one train goes to Cologne and they alternate between origins every half hour. Each origin-destination pair gets a direct service in one half hour and a same-journey-time cross-platform interchange opportunity in the opposite half hour.
Brussels already has the DB train to Frankfurt running 20 minutes after the Eurostar from London. I believe the staff feel that is a bit tight, but it does mostly work.
As per government data UK-Belgium is 37%, UK-France is 20%, UK-Netherlands is 16%, UK-Germany and UK-Switzerland are currently 2.5%, UK-Italy is 0.6%, UK-Spain is 0.2%.
Gives you an idea even if we are outside the EU.
The point is if something is managed as a proper integrated thing, and at a reasonably high frequency, you can afford to go for Swiss-level interchange times where the infrastructure allows. (This kind of operating model really does require the UK to be in Schengen so is entirely academic …)
To be honest, you don’t necessarily need betterment on existing timings. Current typical timings at Brussels south eastbound:
Repeat the same thing in the opposite half hour but with London – Cologne and Paris – Amsterdam as direct services
It might be Germany is brought into this franchising authority so the existing Brussels – Frankfurt falls within the scope of the PSO.
@Weifeng Jiang, how are you going to make the Eurostar to London that reliable given it runs on the same track as London commuter services and then the same track as the channel tunnel trains.
The Japanese in part pull off their insane reliability because there is basically no overlap between the Kansai network, the Tokyo network, the Nagoya network and the long distance Shinkansen network. If there are delays in one part of the system they cannot cascade anywhere else. And even within those networks there is a lot of separation.
@Matthew Hutton – Eurostar operates in just about as benign a railway environment as you can ever imagine. The Swiss manage tighter connections on much more mixed operating environments. The Germans plan such connections at Mannheim, the Czechs do so at Breclav (3-minute recognised connection).
If you operate a 30-minute interval service the occasional late train doesn’t really matter – you just get the next train – and that’s still a better connection than half of timetabled connections in places like France and Spain.
If we cannot timetable and operate a reasonably efficient and reliable railway on a medium-intensity and low speed differential railway like HS1 then the entire European rail industry might as well pack their bags and go home.
@Weifeng Jiang, if the Germans plan a tight connection with hourly service then that leaves a very sensible for them of 60 minutes between trains 😂.
I think one of the Eurostar challenges is that the British and French systems are completely different so the British only know where the Eurostar is once it gets through the tunnel and vice versa.
The British and French can also fail to prioritise a late running Eurostar over other on time services. I have been on a delayed Eurostar stuck behind a slower moving empty coaching stock in the UK and I believe a slower moving Lille-Lyon service that could have gone faster and waited at the next stop.
Clearly the Germans consider their planned connection Mannheim an acceptable balance between performance risk and timetabling ambition. What I’m proposing here is less risky – a less severe outcome in case of delay on a better performing railway.
We are in such an academic debating context anyway (the UK being in Schengen) then the rest of the debate can be similarly academic. What you describe are basic railway regulation decisions – regulating trains taking into account different classes of trains and minimising overall network delay. This is the kind of thing where competence can be improved with fairly basic training. Or at the very least those are just day-to-day railway operating risks that any sensible person would just accept, under a mindset of continuous improvement of course.
@Weifeng Jiang, if you look at the station top destination data it looks very much like there is a pretty extreme reluctance to change trains at all. Mostly the top destinations are reachable without changing.
@Matthew Hutton – you need culture change. The Swiss and Dutch realised you could not possibly achieve a high rail market share with direct services alone. People will change if you make interchange front and centre of your timetabling ethos.
Are you sure you didn’t just attend the wrong event? Innotrans is an industry fair, so it’s mostly about exhibiting products. For something more policy-oriented, I’d suggest e.g. the Transport Research Arena.
I don’t mind the products, I mind the extremely private open access provider-oriented pitches.
Isn’t the point of open access that Dagny Taggert’s management skills and Howard Roark’s design skills will combine to get astoundingly cheap, comfortable and stylish rail experiences?
I assume that’s more a reflection of private open access operators being smaller companies and more willing to shop around for rolling stock?
The big state railways are beholden to EU procurement bidding rules, and what discretion they can carve out without getting caught, they burn on placating political pressure to unfairly (but remember to not look unfair) support domestic industry.
The big state railways are also big, i.e., companies trying to sell to them can give them tailored pitches directly. They obviously send representatives to look around at trade shows, but the generic presentation given to everyone isn’t really for them, it’s for the smaller fish.
here’s a video from a German rail nerd on Innotrans….
The EU doesn’t have an Open Access ‘mandate’. The EU’s goal is very simple – promoting cross-border travel. For international services to make sense everyone needs to be on a level playing field. An international train needs domestic passengers in whatever countries it passes to make its finances stack up. International rail service shouldn’t be hostage to national operators’ tit-for-tat horse-trading or insistence that the ‘foreign’ train is set-down only after it arrives into ‘my’ country. Non-discriminatory access is essential and the only way to ensure that is separation of track and trains. The old vertically integrated SNCF with entrenched mindset simply could not be trusted to treat its own trains and other trains fairly.
During the transition period (into the Fourth Rail Package) there was nothing stopping national governments from including their entire national networks into PSOs, as the Dutch and Austrians have done.
If countries like France were really interested the French government could have put all TGVs under PSOs and directly awarded them to SNCF Voyageur on an interim basis. They are simply not interested. Pre liberalisation SNCF ran a measly hourly service between Paris and Lyon. It’s only due to the threat of liberalisation and Trenitalia actually entering the market has supply on this O-D pair gone up to (almost as good as) an all-day half-hourly timetable. In the vertically integrated and politically integrated days SNCF or the French government showed neither the desire nor the aptitude in delivering this.
Different national operators had different levels of competence. The EU cannot mandate competence – it can only legislate to bypass incompetence. The competent authorities and carriers have no trouble managing PSOs. If someone can’t be bothered managing PSOs then let Open Access flood the market is the next best thing.
We see from Italy timetable coordination in an Open Access environment can still be done. On the Bologna to Rome corridors both Trenitalia and Italo have settled into their respective clock-face patterns.
Just to highlight a very specific point:
Decisions are whether to build high speed lines have nothing to do with the operating models of trains. If states are worried about market not responding to new infrastructure they can always specify a PSO to operate the service level the infrastructure business case is predicated upon.
@Alon To sour you even more on on-rail competition, the BEA-TT (French version of a Transportation Safety Board) just put out a bombshell report of compounding failures on one such service, run by EF OSLO, that nearly had emergency workers struck down by passing trains in the middle of the 2023 heat wave.
https://www.bea-tt.developpement-durable.gouv.fr/rapport-d-enquete-technique-sur-la-mise-en-danger-a1379.html
The causal effects tree on pages 62 to 64 (yes, there’s 3 pages worth of summarized failings) comes down to a lack of capacity, a lack of training, a lack of planning, and crucially, a lack of clear responsibility when it comes to managing emergencies involving third-party operators.