European Urbanism and High-Speed Rail

Europe has a number of strong national high-speed rail networks, providing much inspiration internally as well as abroad, including in the United States. With Americans looking at an infrastructure bill including high-speed rail funding, there’s a lot of discussion about what can port, hence my proposal map. That said, caution is required when doing naive comparisons with Europe. European urbanism doesn’t work the same as American urbanism, in two ways. First, European cities are more compact and transit-oriented than most American cities, which is why I somewhat discount American lines unless at least one city connected has public transit. And second, Europe has more, smaller cities than the rest of the urbanized world. This post concerns the second issue.

French and American urbanism: an example

A few months ago I poked around European and East Asian metro area lists. The upshot is that whereas in the three East Asian democracies 70% of the population lives in metropolitan areas larger than 1 million, in France only 33% does, and the median resident sorted by metro area size lives in a metro region of 350,000.

We can apply the same analysis to the United States. At the CSA level, the median American lives in Sacramento, population 2.6 million, and 68% live in metro areas of at least 1 million; at the MSA level, the median is Milwaukee, population 1.6 million, and 56% live in metro areas of at least 1 million. American metropolitan areas are unusually weakly-centered, especially at the CSA level, but otherwise they’re pretty typical of the urbanized world; it’s Europe that’s unusual in having such small cities.

The upshot is that people who are not used to this peculiarity of Europe who look at a map of European cities focus on million-plus metro areas, which are not the whole story here, especially not in France. This makes Europe look emptier than it is, which can lead people to overrate how much ridership a high-speed rail network would have at a fixed population.

France and the Midwest

Scott Hand posted a map on Twitter superimposing France on the Midwest with Chicago taking the place of Paris, arguing that they are similar in population and area:

This is a good sanity check: your Midwestern network should be of comparable magnitude to the TGV network, rather than much larger. It’s easy to say, Lyon has 2.5 million people, Detroit has 5 million people, so clearly a line to Detroit is twice as good as one to Lyon, right? But no: French urbanism supplies many more small cities, which must be accounted for as well. At the end of the day, the populations are similar, even though, in addition to Chicago, the map has three cities (Detroit, St. Louis, Cleveland) with larger metro areas than Lyon and six more larger than Marseille (Milwaukee, Indianapolis, Nashville, Cincinnati, Columbus, Pittsburgh).

The LGV Sud-Est

It’s tempting to compare Paris-Lyon to Chicago-St. Louis. Yonah Freemark did this in 2009, and Jarrett Walker already pointed out in comments that the LGV Sud-Est was always about much more than this. On hindsight, I’ll add that even that sells the LGV Sud-Est short. High-speed rail between Paris and Lyon unlocked fast service from Paris to not just Lyon but also the following metro areas, all with 2016 populations:

  • Dijon (385,000), demoted from the PLM mainline to a branch but still served
  • Grenoble (688,000)
  • Saint-Etienne (520,000)
  • Chambéry (225,000)
  • Annecy (236,000)
  • Valence (187,000)
  • Vienne (115,000)
  • Bourg-en-Bresse (128,000), not on any direct train but still close enough by regional connection or car

What’s more, TGVs would branch from Part-Dieu along legacy lines to serve these smaller cities, albeit at low frequency. Now, with the LGV extending as far south as Marseille, Valence has a through-station on an LGV just outside the built-up area. There’s also Lyria service to thee major Swiss cities; Geneva, a metro area of 1 million, lies on a low-speed extension of the LGV Sud-Est, 3:11 from Paris.

Other than Geneva, which is invisible on the map because it is farther away, the other cities listed are all very small. In the United States, people don’t usually think of metropolitan areas of such size as urban, because they are extremely dispersed and socially identify as not-urban, and because metropolitan America operates at much larger size classes. But they have recognizable urban cores and their populations must be put into any ridership model trying to train data on TGV ridership. In fact, a gravity model with exponent 0.8 predicts that the combined TGV ridership from Paris to all the above cities, excluding Lyon, is nearly twice the ridership on Paris-Lyon.

And in this context, Chicago-St. Louis simply doesn’t compare. St. Louis is somewhat larger than Lyon, yes, but within 60 km, within which radius Lyon has independent Saint-Etienne, Vienne, Bourg, and Mâcon, St. Louis only has its own exurbs. To find a proper Midwestern comparison for the LGV Sud-Est and its extensions toward Marseille, one must go east of Chicago, toward Detroit and Cleveland. Within 60 km Detroit too only has its own CSA plus Windsor, but that CSA has 5 million people, and the same line also reaches Cleveland (CSA population 3.5 million), Toledo (900,000), and Pittsburgh (2.6 million) and points east.

What this means

Having fewer, larger cities doesn’t make it harder to build high-speed rail. On the contrary – it’s easier to serve such a geography. Asia lives off of such geography; Japan and Taiwan serve nearly their entire populations on just a single line, and Korea does on one mainline with a branch. An Asianized France would be able to serve nearly its entire population on the LGV network as-is without needing low-frequency branches to Chambéry- and Valence-scale cities, and an Asianized Germany would be able to just build an all-high-speed network and connect nearly everyone and not just half the population.

There are small cities that happen to lie on convenient corridors between larger cities, the way Valence is between Lyon and Marseille, or Augsburg and Ulm are between Stuttgart and Munich. Other small cities are close enough to large cities that they’re decently-served by a large city-focused rail network, like Saint-Etienne. Those cities are compact, so a large share of the population has access to the train – this is the explanation for the 0.8 exponent in the gravity model of ridership. But overall, most cities of that scale are strewn haphazardly around the country: examples include Limoges, Amiens, and Caen in France, and Osnabrück, Chemnitz, and Rostock here.

However, this doesn’t mean that, in analyzing the impact of population on ridership, we should just pretend the small cities don’t exist. They do, and they supply extra ridership that isn’t visible if one thinks city = metro area of 1 million or more. It’s an understandable way of thinking, but Europe has a lot of ridership generated from intermediate cities and from cities that have a regional rail connection to a big city or a less frequent direct intercity train, and the models have to account for it.

So yes, that the US has so many large-by-European-standards cities means high-speed rail would work well there. However, it equally means that a naive model that just says “this looks like the LGV Sud-Est” would underperform. A better model has to account for specific city pairs. American city pairs still look okay, even with extreme levels of sprawl at the outer ends, but ultimately this means the US can have a network of approximately the same scope of the LGV network, rather than one that is much denser.


  1. wiesmann

    Swiss cities tend to weight more than their population, because of the economical disparity with France. Among the 10 most important airports of France, two are closest to Swiss cities (Geneva and Basel). The Swiss government is also paying for rail infrastructure on French soil, one third of the Haut-Bugey line between Bourg-en-Bresse and Bellegarde was paid for by Switzerland.

    • Max Wyss

      Switzerland has a budget to help financing international rail projects bringing big advantages for the country. The Haut-Bugey line speeds up the TGVs to Genève, and (on the other end of the country) the contribution for the Lindau – Geltendorf electrification speeds up the ECs between Zürich and München.

      I think there was also some contribution to Italy for local lines.

  2. michaelrjames

    Just to point out that I’m pretty sure one could take a TGV from Paris to Geneva not long after the Paris-Lyon LGV opened, ie. in the early 1980s. About 4/5ths of the line is LGV, just like about 4/5ths of the Paris-Dijon line was LGV. Thus when this first TGV line opened it brought significant other destinations like Dijon and Geneva much closer. (I can’t read it on that juxtaposition map but Geneva is approx. a bit NW of Cincinnati; I believe the TGV route is direct to Geneva, not via Lyon ie. using the Bourg-en-Bresse line, single seat and fast)

    Another thought this piece provokes, is I wonder if American HSR has to use a different terminus model to that in Europe or Asia? That is, due to the low-density sprawl and the weaker centres, perhaps they need multiple termini with one in the middle-to-outer sprawl plus one in the centre, and maybe thru-running to mid-sprawl on the other side? So, three stops per major city and very awkwardly adds significant travel time to have these extra stops but for most travellers it is more than compensated by delivering them closer to their destination. In the absence of a developed, rapid regional rail system it could make the difference in attracting ridership. Plus, with a bit of joined-up thinking the ROW from the sprawl stations to the centre could provide new routes for future regional rail/S-Bahn. This last stretch–from the sprawl edge to inner-city–is typically the most expensive bit of ROW to build so it makes economic sense to make it dual function. If some of it is tunnel then that tunnel should be the biggest feasible so it can hold at least four tracks. The biggest single-bore tunnels today can hold two levels, each of two tracks.

    In fact we already see something approaching this in the few serious extant HSR plans. Namely CaHSR which stops at San Jose (and SFX, and maybe east bay) in SF’s sprawl zone (and inevitably will cross the bay from SF to Oakland), and Texas Central clearly wants/needs termini in both sprawl and CBD (but so far is not getting them, almost certainly due to the cost). Even though Paris is bigger, both in population and geographically, than any American city on this map, it didn’t have this problem because it already had well-developed regional rail, though it is strange to think that the RER was only about 4 years old when that first TGV began!

    • Alon Levy

      Three stops in a city isn’t an unusual thing, it’s already how things work in rich Asia. The Shinkansen does this, THSR does this, KTX has an awkward reverse-branch in Seoul. It’s rare in Europe, but Berlin has it too – trains don’t just stop at Hbf but also at Spandau and Ostbahnhof if they run east-west and at Gesundbrunnen and Südkreuz if they run north-south.

      • Nilo

        Doesn’t Paris technically have this too like don’t trains technically stop at both Marne-la-Vallée and Charles de Gaulle?

        • michaelrjames

          As soon as I submitted that post, I was prompted to follow-up, in that yes, Paris has actually slowly evolved to that same ‘model’ though it has yet to attract much custom (after 40 years of TGV! GPX may make the difference). Further it is more a reaction to the crowding of the inner-city stations than directly serving the suburbs. Thus the under-construction giant Pleyel-St Denis station is designed to relieve Gare du Nord and other Metro & RER lines there. Ditto Massy for lines into Montparnasse (and possible future La Defense-Nanterre). Likewise I’m not sure Interconnexion Est at CDG (or the abandoned Interconnexion Ouest), or Lyon’s St-Exupery fit the description because they are primarily to bypass the main city lines/stations. Lyon’s Perrache and Part-Dieu are both inner-city stations.
          So, I’m not sure these–or even the Tokyo examples which afaik are really in previously separate cities (Shinagawa & Yokohama)–negate my point which is that perhaps in the US they should explicitly set out a model to achieve this on any of their LGVs. That is, it looks like a sensible strategic rail transit thing for Buttigieg, esp. to integrate with regional rail. This is just a direct extrapolation of the issue that Alon is addressing in this piece, ie. the biggest underserved population centres aren’t those smaller cities but the exurban sprawl in the biggest ones.

          • Henry

            Shinagawa is in the 23 special wards and is 7.5 km away from Tokyo station, I wouldn’t really call that “a separate city” except if you mean by purely arbitrary administrative distinctions.

            You generally put these stations to intersect with perpendicular regional rail lines (or build perpendicular lines after the fact like in Taiwan), so the most logical place would be to put them wherever the TGV intersects with future Line 15. Are they planning to make Saint-Denis Pleyel a TGV stop? Didn’t see that anywhere googling, but that might just be the limitations of looking in English.

          • Alon Levy

            No TGV today stops at both a Paris terminus and another Paris-area station. This is an important distinction. There are TGVs that stop at Massy, but they don’t also stop at Montparnasse. The TGV in general is not designed for fast boarding or disembarking, so intermediate stops are long enough for people to go out and smoke, and at the Paris terminal passengers take 5-10 minutes to clear the train, the circulation is so obstructed. ICE intermediate stops are shorter, maybe 2 minutes, because they are single-deck and don’t try to unload an entire train in one city; Shinkansen stops go down to 1 minute when the train is not being overtaken, with two door pairs per car rather than just one as is common in the Virus Union.

          • Herbert

            Paris should build a main station to the same principles that Berlin has.

          • Alon Levy

            The time for that was 50 years ago when Les Halles was being dug. Today the best that can be done is an Hbf around where Gare du Nord is, with tunnels to Gare de Lyon and Gare Montparnasse. Subsidiary stations at Gare de Lyon and Montparnasse would be great too, but the TGVs would need to be redesigned for faster access and egress. Those 10 minute dwells mean that a line as busy as the LGV Sud-Est would need 3-4 platform tracks per approach track, so Nord would end up as a 12-16 track dig, so S21*2, and then Lyon and Montparnasse would be S21-scale each. And at this point Parisian construction costs are (slightly) higher than in most of Germany, because the enarcs are importing the privatization of the state from the Anglosphere.

          • michaelrjames

            [I reposted this to the correct position; you can delete the original if …. ]
            Gasp, shock horreur, it’s the end of the Enarques!
            This article claims it was provoked by the Gilets Jaune. Whatever next, Boris abolishing PPEs. Who knows, maybe because unlike Macron who is an Enarque, Boris was a classics major with no discipline to master anything, and probably hates the PPE-ers like David Cameron.

            Macron to close elite school that hothouses French leaders
            Institution has been pathway to power for country’s elite, including four French presidents
            Kim Willsher, 8 Apr 2021
            Emmanuel Macron is expected to officially announce on Thursday the closure of the École Nationale d’Administration, the elite French finishing school for the country’s leaders where he studied.
            Known as ENA, the grande école has been the hothouse for France’s top civil service and a pathway to power in the public and private sectors. Four French presidents, including Macron, have passed through its doors as have dozens of ministers and business leaders.
            Founded by Gen Charles de Gaulle in October 1945 with the idea of breaking the upper-class hold over France’s higher echelons, ending nepotism and making the civil service more democratic, it has instead become a byword for an establishment elite that critics have accused it of encouraging groupthink.

            FWIW, I’m not sure the problem is the ENA or the Enarques in quite the same clearer way that PPE and its illuminati graduates are guilty of creating and perpetuating their toxic neo-feudal economic caste system. But ENA does need to widen its intake. Macron has made significant changes to the French two-class tertiary system so he’s serious about this stuff. It’ll take a generation or longer.

          • Alon Levy

            The difference is that ENA is being replaced with an institution that is like ENA but with a different name. All because Macron didn’t have the guts to mass-arrest rock throwers in December of 2018, corona denialists last summer, etc.

          • michaelrjames

            Well, as I said, it’s really the narrow nature of the intake students that likely limits the outcomes. So, in that sense I agree that why bother ‘closing’ the ENA to recreate something similar, when the real action is to make sure more students from more diverse backgrounds get into it, as Macron has stated as his goal. There may be valid reasons, such as having a complete turnover of senior staff and direction etc.

            The entire world has this problem to some degree. Grand Ecoles esp. ENA, PPE@Oxford, Harvard & other Ivys etc. Australia’s system is still reasonable in that the threshold is not so steep to get into ‘good’ universities, plus not having such a disparity between universities; I think the really sharp dropoff in getting into too few to schools, is too extreme in UK, USA and France, and probably some others (certainly China is brutal; Japan, Korea?). Selecting the top few percent for the top schools is a pretty simple trick to produce ‘top’ achieving institutes and is self-reinforcing but I’ve never thought it necessarily was the best outcome. And its very nature is bound to suppress innovative thinking and produce self-perpetuating social exclusiveness. At some point those chickens come home to roost.

            Also as I said, there is no judging this action of Macron in the near-term because any effect will take an age to filter thru. That’s why I award him some provisional brownie points for both recognising the problem and trying something to fix it. Especially when a product of that system. How many politicians do anything that will only have an effect long after they are gone and forgotten, and will do zip to win their next election?.

          • Henry

            Re: dwells: how does Europe handle larger than airline carry-on size luggage?

            I’m wondering how short dwells would work with baggage; Shinkansen have an oversize baggage area but you have to reserve the space before hand and space is limited.

          • fjod

            Henry – usually a couple of small baggage racks per carriage for intercity or airport trains, usually near the doors; here are some plans. In my experience this is not really a constraint on dwell times; I don’t know of any luggage that can’t be unloaded in 1-2 minutes. Sometimes there are restrictions on bikes too.

          • Herbert

            “Sometimes there are restrictions on bikes too” in Germany means in practice that only the ICE 4 allows non-folding bikes to be carried at all (old IC stock also do, but they are not “true” hsr)

            In general there is always a calculus with cyclists as they could in theory be valuable customers (who already kinda value the environment) but the marginal revenue of a bike rack versus extra seats is not a particularly attractive proposition…

          • Sascha Claus

            Note that the number of bikes travelling to and from holidays fluctuates widely with season, and that the high season for bikes is also the high season for tourists in general. Fluctuating demand is the bane of and accountant, as additional trains/cars need to earn enough money to compensate for sitting idle the rest of the year.
            Add in the fact nobody reimburses the operators if children travel at reduced fare, which makes it unattractive to offer such perks, and families with children have to fork over quite some money to make it worthwile for the operators to target seasonal touristic travel. This makes the automobile quite attractive for holiday travel; even more so, if accomodation near the stations is more expensive.
            Now add additional space needed if you want to carry bikes, and seasonal holidays by train look like something for the wealthy.
            If you’re lucky, you can spread your bicycle season and use the same rolling stock to serve winter destinations, using the space for ski equipment. That would allow some more income from tourist travel.
            PS: I wouldn’t consider the ICE 4 high-speed, as it is the replacement for the old IC stock and has a maximum speed of only 250 km/h.

        • Max Wyss

          These two stations are on the bypass route, and served by trains from the south(-east) to the north (and beyond to Belgium).

      • Sascha Claus

        Suburban stations come with the advantage of not only offering better car access, but also lower real estate prices for parking garages and car rental lots, should the need for them arise. Just in case somebody is wondering how people will get around without local transit.

        • Herbert

          But there is also a big incentive to do something more high-value with the near-station land than dumping motorcars on it… Like retail or offices…

          • Henry Miller

            When we build the edge station there isn’t much demand in the area so you get a “strip mall” across the street from the station and then parking lots or green grass as far as the eye can see. Local transit will also make the station a focus of their network, because it is an important destination. Over time businesses figure out this is where people are so they try to relocate there. This process over time builds the area and the parking lots get pushed farther out, and then someone figures out they can put in a paid parking ramp and get paying customers close to the station. Depending on lots of factors (some I don’t even know), this process can take between 10 and 500 years.

            The above doesn’t apply to non-edge stations. If there is something there, then it skips the whole parking lot.

      • michaelrjames

        So the same year service Paris-Lyon began. That was nifty of them. Obviously it must have been in planning for years but I don’t remember any media about it, like Paris-Lyon got. Must make it the first transnational Euro HSR, even if it wasn’t that speedy at the border.

        • wiesmann

          Technically, the two last tracks in Geneva main station are under French juridiction, you need pass custom controls to reach them, so technically, this train was only running between French stations. The line between Geneva main station and Bellegarde also has a strange status and is under French voltages – at that time 1.5KV, I think they switched to 25KV.
          The lack of media coverage is hardly surprising, Geneva is a french speaking city, known internationally, the center of gravity for a good chunk of French territory, but is not in France. This does not jive well with the French narrative.

          • michaelrjames

            Right, that makes sense as the French didn’t have to convince anyone else to run that service. Though I see that today the SBB has a quarter ownership of Lyria. The 3 to 4 hour travel times in the comfort of a TGV from the major Swiss cities to Paris won them over …
            I remember that arrangement exist(ed) in Basel where the line from France was the first trainline built in Switzerland, and that the station had French extraterritorial status. I think it became three stations under one roof, the French, plus German and Swiss?
            But do these stations still do passport control, ie. since Switzerland joined Schengen? A brief browse of Seat61 says that the physical separations still exist but no passport control despite various officials “lurking in the corridors”.

          • Yom Sen

            In 1981 the line opened from Saint-Florentin (near the junction to Dijon) to Lyon: All routes from Paris to Lyon/Macon and further, entirely on electrified lines were switched to TGV: Paris to Nice, Marseille, Saint-Etienne, Montpellier, Chambéry, possibly even Port-Bou and Modane. Only Grenoble had to wait until 1985 and electrification of Lyon-Grenoble to have its TGVs.
            I thnk unlike Paris-Lausanne and Paris-Zurich run in cooperation between SNCF and SBB, Paris-Geneva was run only by SNCF until the creation of Lyria.
            Switzerland is part of Schengen but not part of the Customs union, so in theory there are no more passport controls but still controls on goods. Last time I was in Geneva I saw customs officers but I was not asked anything. It’s the same by car, only random controls are done.

          • wiesmann

            Basel has a similar arrangement for the French part, DB runs out of Basel Badischer Bahnhof, a different station on the other shore of the Rhine. Passport control in Geneva and Basel exist, but they were pretty relaxed before Schengen (many people commute over these borders) and even more so afterwards. No sure what the situation is now with Corona…
            Note that in Basel the TGV don’t alight on the french tracks, as the train comes from Zürich, so passport control happens on board…

          • Herbert

            There are even “international trams” in the Basel area and of course the airport with three different IATA codes…

          • Sascha Claus

            Basel Bad. Bf. is German customs territory and was German “immigration territory” even before Switzerland joined Schengen. This was arranged so that travellers could connect between the various branch lines to Germany and long distance trains; Basel Bad. Bf. was the main station for that corner of Baden. Just look where the rivers and their valleys are to see why.
            Passport control was at the exit/entry from the station to the street and … at the beginning, passengers between Germany/Baden and Switzerland had to alight and pass a passport control booth on the platform, while the empty train was slowly moving in parallel.
            I don’t know if and when that was changed to controls on moving trains or replaced by occasional, random checks. I only travelled there in the 2000s, when controls were practically nonexistent.

          • Max Wyss

            @wiesman: The line from Bellegarde to Genève was built by the PLM, and is still following French standards. It does, however have local stops within Switzerland. When it was 1500 VDC, the SBB had two train sets for 1500 VDC which shuttled between Genève Cornavin and La Plaine (the border station). They got replaced by 5 light rail vehicles similar to the Lausanne T1 ones.

            With Schengen and the Genève RER, local trains are now running beyond La Plaine.

            There are no official border checks, but (as stated) customs checks are still possible (in fact, Swiss custom officers may check anyone suspicious within 50 km from the border).

          • Max Wyss

            @wiesman: Before TGVs, the direct trains into France all stopped at Basel SBB, and Customs and Immigration was on the moving train between Basel and Mulhouse. These trains used Track 4 in Basel SBB, which is switchable between 15 kV and 25 kV. This was also the case when they had through-running of local trains between Rheinfelden and Mulhouse (unfortunately, that stopped when France changed the crashworthiness rules, and it did not resume when the FLIRT France became available.

            Even in the last years before Switzerland joined Schengen, the border checks in local trains were not that strict. My experience is more between Schaffhausen and Singen, and more than often, at Thayngen, the official just looked into the train from outside; he didn’t even bother to come on board (but these local trains were mainly used by regulars (workers and students). The REs not stopping between Schaffhausen and Singen had a passport/customs check on the access to the platform in Schaffhausen, however.

            About Basel SNCF, they are rebuilding that part of the station, and there is a well hidden customs office around (but I am not sure when you actually can see an officer; if you have anything to declare, you will do self-declaration).

          • Max Wyss

            @Herbert: Yep, that’s correct. One tram line goes to Weil am Rhein in Germany, and another one to St.Louis in France.

            The access to the airport is kind of interesting. There is an exterritorial road (for France) connecting Switzerland with the airport (Swiss section). This road has been in existence for decades already (and is served by bus line 50).

            The airport has 3 IATA codes: BSL (Basel, for Switzerland), MLH (Mulhouse for France) and EAP (generic “Euro Air Port”, for Germany). Some decades ago, my boss (who was somewhere near Olten) had regular trips to Paris, and pre-TGV, he flew. However, he never flew from BSL, but always from MLH… BSL – CDG or ORY was an international flight, but MLH – CDG or ORY was a domestic flight, and considerably cheaper. All he had to do, was crossing the border between the Swiss and the French section of the airport…

          • Sascha Claus

            In Basel, don’t forget the BLT overland route 10 to Rodersdorf, which passes through the French village of Leymen on the way. And the foot- and cyclepath along the exterritorial road to BSL/EAP/MLH. Both the bus route as well as the foot- and cyclepath seem to be missing from the exterritorial road to Geneve Airport (GVA). (The Swiss seem to have a knack for extraterritorial customs roads to airports. 😉 )

  3. Onux

    “This is a good sanity check: your Midwestern network should be of comparable magnitude to the TGV network, rather than much larger.”

    Why shouldn’t it be larger? France thinks that its LGV network should be larger, it is building or planning extensions to Turin, Nice, Perpignan/Spain, Toulouse, Le Havre/Caen, and another line to Calais, in addition to things that may not happen, like the Orleans-Lyon line, all (supposedly) within the next ten years. If that is the network that should exist for this population size in ten years, why shouldn’t the midwest plan for a similar sized network for itself in ten years? Especially, since as you said, the larger city sizes in the midwest should make HSR easier, not harder.

    • Alon Levy

      Yeah, there’s Toulouse and Nice, but that’s about it. The Normandy lines are really weak, and with the tunnels that got foisted on the project by NIMBYs, so is Paris-Nice to the point that it’s being downgraded to a mixed high- and low-speed line. And more conceptually, SNCF is bad at capturing markets that are not provincial city-to-Paris, so for example Marseille-Lyon has insufficient frequency.

      • Onux

        I also forgot to mention that Scott’s map does not include LGV Nord to Lille/Calais (roughly corresponds with Chicago-Milwaukee-Madison) and LGV Rhine-Rhone (roughly Indianapolis to Columbus). In that case the map would then cover almost everyone’s practical midwest HSR network, plus has the “extra” route toward Des Moines. The fact that France is also building the Toulouse line, the Nice line (if watered down), the eventual route to Turin if they finish the base tunnel, and the short extensions to LGV Rhine-Rhone, plus plans for Montpellier-Spain, the Normandy line, and concepts like the Orleans-Lyon route, Interconnection Sud, further Rhine-Rhone extensions, and Toulouse-Montpellier as part of a “Southern Line” (Bordeaux-Nice), I still think France has/is planning a lot more HSR than most people assume for the midwest.

        • Richard Gadsden

          If the Spanish ever build the Basque Y, then I think a Hendaye-Bordeaux line is inevitable – that would be a much faster Paris Madrid route than going via Barcelona, and the French are much bigger fans of an occasional super-long-distance train to Paris than they should be.

          • Onux

            My understanding is that the Hendaye-Bordeaux line would be mixed passenger/freight, so that while it could/should be a faster route, travel time Paris-Madrid would be the same as through Barcelona. I could be wrong about this.

          • Car(e)-Free LA

            The other issue there is that getting from Paris-Madrid in under 5 hours requires cutting from San Sebastian down to Zaragoza instead of looping through Segovia.

          • Mikel

            @Car(e)-Free LA: according to Alon’s crayon, the easy part in that option is carving a new ROW in the Bayonne-Hendaye sprawl, which is scheduled for not earlier than 2040. Then you’d have to punch 50 km of tunnels between Hendaye and Pamplona…

            To get Paris-Madrid in under 5 hours you probably need to bypass Bayonne and San Sebastian altogether, at which point a Pyrinees base tunnel becomes a good option to make the money printer go brrr.

          • Herbert

            There is occasional talks of a base tunnel roughly at the site of the old Canfranc crossing…

            As a “first step” they seem to be in the process of reopening the line to France (you can already catch a domestic train to Canfranc and iirc the French line is also in service to within a few kilometers of the border, but the link is missing and the most difficult part)

          • Car(e)-Free LA


            I’m not completely sure you’re right about the times. We start out at 2:05 from Paris-Bordeaux, nonstop. Madrid-Pamplona should be doable in 1:20, nonstop. That requires you to do Bordeaux-Pamplona in 1:35 to keep the trip down to 5:00.

            Bordeaux-Bayonne in 0:45 is easy. Taking the bypass around San Sebastian on the Basque Y and doing Bayonne-Pamplona in 0:35 is definitely possible. That means Paris-Bordeaux-Bayonne-Pamplona-Madrid in 5:00 can be managed. I’m very skeptical of the utility of a base tunnel considering how few city pairs it actually helps–even with Spain’s ridiculously overbuilt HSR network–and actually doesn’t cut Paris-Madrid track mileage by much. A short tunnel just from Bayonne to Pamplona could drop Paris-Madrid to 4:40 which *might* be worth it.

            It’s also important to take into account how a takt system involving Valladolid, Burgos, Logrono, Vitoria-Gasteiz, Bilbao, San Sebastian, Pamplona, Hendaye, and Bayonne might operate when considering what additional infrastructure might be worthwhile. After all, seeing city pairs like Bilbao-San Sebastian-Bayonne as proper high-speed intercity rail is pretty meaningless, considering metro size and speed.

          • Mikel

            @Car(e)-Free LA: I think you’re being too optimistic. Madrid-Pamplona will be around 430 km on the proposed alignment, so a 1:20 time would require an average speed of ~320 km/h. Given that nonstop Madrid-Barcelona trains average 250 km/h, 1:35 would be more realistic. Bayonne-Pamplona in 0:35 is also difficult considering that the San Sebastian bypass will have a 170 km/h top speed and a 140 km/h junction where it joins the Basque Y and the Madrid-Hendaye mainline. And that’s assuming that a huge tunnel can be built roughly parallel to A-15, an option that was ruled out in previous studies due to the high cost and environmental impact and also would increase the Bilbao-Pamplona travel time. Possibly cheaper speedups would include bypasses of Valladolid and Vitoria and a redesign of Bordeaux-Dax to make it straighter.

            All things considered, I’d say 5:30 is a more realistic target for Paris-Madrid, and it’s the proposed travel time in Alon’s Eurail crayon. I’m not sure if their calculations include a stop somewhere in the area between San Sebastian and Bayonne, which despite a polycentric, not-huge population (maybe three-quarters of a million, depending on how you count) is an important tourist destination with potentially strong local and regional rail connections. Anf if you can get from the border to both Paris and Madrid in under 3 hours, it’s going to obliterate flights.

          • Car(e)-Free LA

            I’m aware of present day speeds, but in my view, all that means is that Renfe is running trains unacceptably slowly and might need to straighten out some curves. There is literally *nothing* between Madrid and Pamplona and therefore, there is *no* excuse for running trains slower than 350 km/h. Which, incidentally, means that AVE trains from Madrid-Barcelona (with an actual stop in Zaragoza) should be 2:15 or better. If there’s one thing CAHSR and HS2 are doing right, it’s running trains at 350 km/h.

          • Eric2

            According to Wikipedia, Madrid-Barcelona takes 2:30 with a top speed of 310km/h (and the line was designed for 350km/h). I guess you could get down to 2:15 by running at 350km/h, at a higher energy expenditure, but it doesn’t seem like a dramatic difference.

          • Car(e)-Free LA

            Well, even that is assuming a stop in Zaragoza. You could cut over 20 min nonstop at 350 km/h.

          • Mikel

            No. Nonstop trains take 2:30 at 300 km/h top speed, and the Zaragoza stop adds another 15 minutes. Running at a top speed of 350 km/h (a 16% increase), you could maaaybe do Mad-Bcn in 2:15 (an 11% increase in average speed) and Mad-Zgz-Bcn in 2:35. And that increase in top speed would increase track, catenary and rolling stock wear from ballast getting sucked into the train nunderside. It would also reduce capacity unless you kick out the 250 km/h trains. It’s better to invest in getting the Sagrera station and the Atocha-Chamartín tunnel finished to reduce door-to-door trip times.

            Here’s the math. This is the top speed diagram of the line. Let us assume that trains currently run at 300 km/h where possible, and that acceleration and braking is instantaneous. The 300 km/h begins in Bif. Vallecas AV and ends near Gelida, so 570 km = 114′ @300, 98 @350. That’a a 16-minute speedup in a very optimistc approximation, so the actual reduction would be more like 10-12 minutes.

          • Car(e)-Free LA

            I stand corrected. Either way, speeding up *every* train from Madrid to the Zaragoza wye to 350 km/h would be extremely useful for France-Madrid services. Assuming Pamplona-Zaragoza is built to 350 km/h, that does mean Pamplona-Madrid in 1:20 is possible. Which matters a lot to the Paris-Madrid bottom line.

          • Richard Mlynarik

            there is *no* excuse for running trains slower than 350 km/h

            There are many better things to do with energy than pushing air out of the way of fantasy-speed trains, and this is only going to be increasingly the case as (if! or doooom!) more of the economy electrifies.

            Just because you can do something doesn’t mean you should do it.

            I know Alon disagrees, but I don’t see sustained 320+kmh as an economically/environmentally justifiable or “sustainable” (whatever that word means) in the medium-long term.

          • Nilo

            I mean the Chinese already run trains at 350 km/h don’t they? There may be reason to object 350 km/h trains beyond that, but that seems like it’s a real choice that can be made today.

          • Eric2

            @Richard Mlynarik
            What about during the day when there is excess solar energy? That seems likely to be a common situation pretty soon.

          • Sascha Claus

            250 km/h trains … are these the gauge changing twice-a-day Barcelona – Bilbao / Irun trains? With change of gauge in Zaragoza and splitting the train in Castejón de Ebro (for Bilbao and Irun). These are wonderful examples of the technological and timetabling possibilities, an absolute must-have for every model railroad, be it real or virtual!
            Just move these trains off the proper railway and replace them with a takt Zaragoza – Bilbao / Irun and timed connections to the Barna – Madrid trains. Same for the twice daily Madrid – Bilbao trains; I assume they go via Valladolid.
            These four trains and a fifth one to Vigo (far away on the western coast) were the sole amount of intercity traffic to be found in Bilbao three years ago. Typical for Spain: it’s crying for a network of frequent intercity trains (at least until HSR arrives), but they run two trains per day (sometimes four). But all the way 8 or 9 hours to Madrid (or Barna); with dual-mode trainsets if needed. Well worth working your HSR timetable around these trains.

          • Henry

            @Nilo: to add onto this, China used to run those lines at 380 km/h, but reduced them to 300 km/h to save running costs and after the Wenzhou train accident. It’s back up to 350 but that seems to be the upper bound they are willing to entertain.

          • Mikel

            @Sascha Claus: The problem is not the sporadic 250 km/h trains to northern cities, but the frequent 250 km/h trains to Valencia and beyond. Most criticism of Renfe’s weird operations and allergy to timed transfers is fair, and I usually write a lot of it in this blog, but gauge-changing trains account for ~30% of HSR ridership so they are, in fact, a great technology. There will be gauge-changing trains capable of 350 km/h next year, so don’t despair!

            I don’t see why you’d want to kick Madrid-Bilbao trains from the HSL — the entire point of the Guadarrama base tunnel is cutting 1 hour for all the trains that leave Madrid and then branch to multiple destinations. Galicia, Asturias and Bilbao are more or less equivalent in terms of population, so they all deserve direct trains. (Bilbao is economically much stronger but the business travelers fly).

            I agree that Bilbao currently has shit frequency to everywhere; the main problem is that the current line is so slow it’s slower than buses, and it doesn’t even serve the Bilbao-Vitoria commuter market. Ditto for the narrow-gauge line to San Sebastian, which nobody rides end-to-end except for a few tourists. All of these problems will be solved by the Basque Y, if they ever manage to finish building it…

          • Herbert

            An unmodified Siemens Velaro reached above 400 km/h at a particularly straight section of track. The question is not whether you can reach such speeds, the question is: Why should you? As outlined above, the speed gains are pretty low and the energy costs are pretty high… At any rate there is no reason to do it before lower hanging fruit are not plucked…

          • Sascha Claus

            @Mikel: How much of the ~30% of HSR ridership would still be on board with a timed transfer? 3%? 13%? 23%? 28%?
            If there is decent frequency on the broad-gauge section, through-running makes sense (and might be necessary for capacity reasons). But for Bilbao, the money might be spent wiser for more Zaragoza – Bilbao trains than for a direct train from Barcelona. (Seems like the second daily train has vanished during the last years.)
            Though 4,5 h for Zaragoza – Bilbao seems a bit on the longish side, but almost 300.000 people in the towns between them (only counting where the train stops) should be able to fill more than one train per day.

      • adirondacker12800

        Define “Midwest”. Just because Lille-Paris-Lyon is three dots it doesn’t it make the same as Detroit-Chicago-Saint Louis. The area of Michigan, Ohio, Indiana and Illinois is bit less than the area of France. Throw in Southeastern Wisconsin and metro Saint Louis they would almost exactly the same size. With very roughly two thirds of the population.

      • df1982

        Should France build an orbital TGV line after its Paris-centred radials are built out? Something like Nantes-Limoges-Clermont Ferrand-Lyon-Geneva-Strasbourg-Luxembourg-Brussels? And as a more general principle, does metro network design (with its radials + orbital structure) usefully scale to HSR?

        • Alon Levy

          Ooh, I like this question.

          I don’t know the answer. In the specific case of France I lean toward no, because the frequency would suck and the cities don’t really form a neat line. But Marseille-Toulouse is probably not a bad idea, esp. if paired with a connection to Perpignan and if SNCF figures out how to set up Bordeaux-Toulouse-Marseille-Nice as a decent-frequency trip with a timed transfer at Toulouse or Marseille. The issue with radial-orbital network layout is that on metro system it lives on untimed transfers, which require very high frequency. On RegionalBahn (as opposed to S-Bahn), orbitals kind of suck, which is why Berlin isn’t running continuous service on the Outer Ring.

          • Eric2

            Toulouse-Narbonne is low hanging fruit – short distance and flat ROW, and quick connections from Toulouse to Marseille+Lyon+Barcelona.

            Lyon-Nancy and/or Mulhouse-Strasbourg might make sense for similar reasons, though I think they are more questionable.

            None of the other radial routes seem plausible.

          • michaelrjames

            Given that Bordeaux-Toulouse-Montpellier is bound to be built, and thus connections to Marseille, Nice and Lyon, then if LGV-Paris-Orleans-ClermontFerrand-Lyon gets built you will have a grand orbital for all of France south of Paris. There is already a line Orleans to Tours–in fact I would guess TGVs travel it to Nantes on the Atlantic coast–and even without upgrading they could run TGVs Tours-Orleans—Lyon etc.

          • df1982

            I know the French are allergic to Swiss timetabling principles, but there’s no real reason why a HSR orbital would have to have untimed transfers. Could the system not be retooled to work to a Takt, and the orbital segments built to take multiples of 30min (or whatever the Takt should be), since they only need to run as fast as necessary rather than as fast as possible.

            As in metro networks, the orbital line need not stack up by its own ridership alone, but it may prove to be of value by making a wide range new connections possible, and it would also relieve pressure on the Paris termini. At the moment you have anomalous situations where the fastest connection between somewhere like Limoges and Lyon is via Paris, which is pretty ridiculous.

          • Herbert

            The only other country with significant hsr and an overall similar geography to France (roughly circular shape with the capital roughly in the center) is Spain and there it ought to work even better as centrifugal tendencies are much stronger and the capital is much more in the geographical center…

          • Herbert

            Of course one of the problems with France is that everything is centered on Paris and lines that won’t hit Paris will elicit a great big “now why would you do that?” From graduates of the Grandes Ecoles…

          • michaelrjames

            Herbert, I think these tropes die hard. In fact devolution policies came in, by those same Enarcs*, in the 60s to counter the gravitational effect of Paris which was threatening to overwhelm the city. It was part of the reason for the TGV network. Where the government has some control they made it more difficult to remain in Paris, eg. teachers compelled to teach some years in the provinces (“turboprofs” because many kept their Paris base and commuted weekends via the TGV), and scientists because INSERM and CNRS positions were very hard to get in Ile de France, and explicitly made easier in various provincial centres along with research budgets. A friend of mine was even a turbo-doc, getting a good position in Clairmont-Ferrand research hospital, while maintaining his Paris base. I won’t bore you with the whole story which I have told here before but my entire Paris lab was given an offer, that we should have accepted, to relocate to the Montpellier med school which had been designated to be a research pole. (I say ‘should have’ but it was at the same time we got an even better offer to relocate to Oxford-Wellcome Trust which at that time had become the world’s wealthiest private charity.)
            And the vibrancy of French provincial cities is also testament to these policies. Compare to the UK. The Lille region was rescued from what could have been a downward spiral that we have seen in rust-belt US and UK. For much of the past 3-4 decades Toulouse and Montpellier have been the fastest growing cities in France (and sometimes Europe).

            *Of course it was the case of those Enarcs pulling up the ladder behind them.

          • Sascha Claus

            Somehow the graduates of the Grandes Ecoles didn’t notice the construction of the A89 (“La Transeuropéenne”) and the upgrading of the E62 (“Route Centre Europe Atlantique”) east of the A20, so it seems possible. They even gave poetic names to these freeways!

          • Max Wyss

            @Eric2: Marseille – Lyon – (Mulhouse) – Strasbourg – (Frankfurt) is already an existing TGV connection, using the LGV Rhin-Rhône. (anecdote; I have never before encountered a train in France where the announcements in French, German and English, by the Chef du Train, were pretty much accent-free … amazing). (another anecdote: the TGVs on the LGV Rhin-Rhône run smooth… even at 330 km/h)

    • Henry Miller

      As a very rough number, just looking at states that would be covered by this map, the Midwest has around 55 million people (counting the population of the state if there seems to be service to the sate, even if population centers are outside the service zone), and France around 65 million. Thus the midwest has a bit less people, but close enough to say there should be about the same amount of track in both.

      Other than the Chicago west to Des Moines route, if you squint hard you can see Alon’s map – just tweak the major lines a bit to where the cities are. Alon’s map lacks all the not major lines that France has. They don’t exist in the midwest, though if there was good train service they could be built next.

      What the midwest lacks is good public transit – this will drive HSR usage down a lot. Though building HSR should be an aid to local transit – I wouldn’t hold my breath about local transit managers getting a clue about what good service means, and even if they had a clue I’m not sure if they will get the budget.

      As to your question, when you are starting at zero you should look for what works. France’s dreams of a larger future are a good guide to the Midwest’s dreams for once the first lines are running.

      • adirondacker12800

        Midwestners and Northeasterners fly to Florida in great big hordes and manage to leave the car at home. They’ll figure out how to park at the high speed rail station and take the subway in Chicago, Boston, Toronto, New York, Montreal, Philadelphia and Washington. Or the commuter trains. Or the bus.

        • Henry Miller

          When you fly to Florida you rent a car. It used to be a regular things at diseny world for someone ti get into the wrong car and the key worked so they drove away. Even though the odds agaist it were one in 20000 or something, there were so many rental cars that it was a regular honest mistake. (Today cars have electronic keys that are unique to one car )

          Midwesterners will most likely rent a car, but at that price in HSR distances they will drive as it is only a bit more time and it saves a lot of money. Plus it saves a lot of stress compared to the airport (I’m not sure how stressful getting your train is ).

          Sure we can figure out local transit. These are Midwestern cities though. Only a few will find local transit that works for their needs. Most will look at the route map and see no service to their destination, or such slow service (1.5 hours to get there…) The rest will run into missing weekend service, or get stranded because the last bus left before they are done. Those last will lead to stories back home leading even more people to drive.

          What I’m saying is local service to sprawling Midwest cities is terrible and should be assumed not to exist for purposes of HSR.

          • AJ

            I think his point is that Midwesterns already know how to get to an airport and go to destination where they don’t need a car at the other end, so HSR should work the same. If the destination is a downtown Chicago or NYC, there is no need for a car; in a smaller Midwest city, someone picks you up in a car, whether that’s an uber driver, family/friends, or a colleague, just like they do at airports all across the Midwest.

          • adirondacker12800

            They wouldn’t be building tracks across the mountains of Pennsylvania for a twice a day train. Or a high speed train to Green Bay. If they decide to use the nice straight ROW that passes Cleveland Airport on the way in and out of downtown it would okay to have another station there. With parking and rental cars and local service on the Red Line to intermediate points. Looks like there is an almost abandoned railroad yard across the Interstate like highway they could use to park trains overnight.

          • Henry Miller


            That point isn’t useful though. We are talking about HSR distances – so the personal car that you already own is competitive. Sure the car is slower, but the incremental costs (most of the cost of a car is the same if you drive it or leave it sitting in the driveway) is about the cost of one train ticket, and you can put more than one person in the car for no extra cost. Plus you get the advantage of your own car once you get there instead of having to rely on friends or pay extra to get around. Sure the HSR trip might be 2 hours, but then you have to get to the station, and from the remote station to where you are going, so 5 hours by car door to door isn’t that much worse.

            When we are talking about distances that people fly, of course people are smart enough to figure out options. However those are not HSR distances and so the fact that people can doesn’t mean they will.

            You will of course get business travelers. They have an expense account, are not allowed to take a personal car (It is cheaper to rent a car than pay IRS mileage on a personal car, plus insurance gets weird about these trips), and are probably going someplace close to a station.

            You will under perform the model though because the lack of a good transit system means you are either paying a lot of money to get around via rental cars or taxi (uber is a taxi), you need to rely on someone agreeing to drive you, or you have to arrange your life around what passes for local transit: just driving is looking like a good option.

          • Herbert

            In sane countries there’s literally no less stressful way to travel than a direct train ride. You get to the station (depending on how you get there with a margin from anywhere like fifteen to thirty minutes or five if you’re familiar with the connection and can trust your mode of getting there enough) get on the train… And then you get off at your destination. How could it be any easier?

  4. Onux

    “In fact, a gravity model with exponent 0.8 predicts that the combined TGV ridership from Paris to all the above cities, excluding Lyon, is nearly twice the ridership on Paris-Lyon.”

    What is the actual split on LGV Sud-Est between Lyon proper and surrounding cities? If it isn’t double that calls into question use of the gravity model.

    • Alon Levy

      I don’t know. SNCF reports ridership per station on all SNCF trains, not just TGVs, and double-counts transfers, so the busiest station that isn’t a Paris terminal is Juvisy, the RER C/D transfer in the suburbs. I can try to figure this out from TGV frequency, which is matched to demand rather than clockface, but that doesn’t work when a Paris-Lyon train also takes in passengers from Bourg and such. It’s only useful for things like Paris-Marseille vs. Paris-Toulon-and-Nice.

      • Yom Sen

        A Paris-Lyon(-Saint-Etienne) train will take passengers to mostly Lyon and Saint-Etienne (+Vienne) since all other significant cities have direct trains (that bypass Lyon so are much faster than a connection in Lyon). OTOH the eastern part of Lyon metro area is partly served by trains to Grenoble, Chambéry, etc. in Lyon Saint-Exupery. My impression is that number of trains is roughly proportional to the population served by these trains.
        By the way, there is 1 small mistake in your article: Bourg-en-Bresse is served by direct trains from Paris to Geneva/Annecy. Your comment probably applies to Vienne, small city 30km south of Lyon which indeed was never served by TGVs.

  5. Tom W

    As has been mentioned above, many TGV *services* continue off the high-speed *lines* to other destinations. This takes advantage of France’s (mostly) electrified regional rail network. This isn’t (yet) an option for the USA – a new high-ish speed service can only use the new high-ish speed line, without a viable option to extend service beyond the upgraded section.

    • Henry Miller

      This is both good and bad. Bad for the obvious reason: the lack of infrastructure means we can’t do any minimal service to low demand places, so they get no service.

      It is good because we are not stuck with legacy infrastructure that can be upgraded. Since most of the cost of building a new line is the same we should look into skipping traditional track for maglev (which is probably mature enough today that a commitment to build it would bring prices down to competitive with iron rail), or vacuum trains (clearly not ready yet, but if you are planning a new line you should keep track of it, if things go perfect it might be ready before you break ground). Legacy iron rail is still a good option as well, it is somewhat cheaper, but when considering the total labor and land costs there total difference isn’t that much so why not go with something faster if you can get ROI.

      If the US plays leapfrog right we can have the fastest high speed rail in the world in 20 years. That is a big if, but it is possible.

      • df1982

        The Midwest has loads of legacy rail line. And because the area is mostly flat, the lines around there tend to be pretty straight too. They’re not great for the stretches between metro areas, but the alignments can be made use of for the approaches to city stations. So no, I don’t think Maglev would be a preferable alternative to HSR in that region, or indeed anywhere in the US. The only places it would make sense are real legacy-free places like the UAE or Saudi Arabia, and even then only if the distances involved make 350km/h insufficient for convenient journeys.

        • Henry Miller

          Either it is already used for freight, or it has been torn up for bike paths (you won’t get them back for political reasons). In middle of nowhere places it has been torn up for nothing, but in general if there is legacy rail it can’t really be converted to HSR without kicking the freight off.

          Maglev has it stands today is a worse alternative to iron rail. However it isn’t that much worse, and there is opportunity to make it better. Speed is always important. Even if you don’t need 400 km/h, if you can get it for not much money you should – the faster you can go, the harder it is for cars to compete. See my other comments about why mid-westerners drive: the faster the train goes, the harder it is to make the argument to take the train.

          • Eric2

            I suppose it all depends on the price of maglev construction. My basic assumption is that it’s much more expensive than regular HSR, otherwise Europe and Asia would be building maglev already (right now it’s a very niche method). So whatever routes Alon thinks are profitable in this map, may be unprofitable even at higher speeds if built as maglev. Notably the strongest line in the whole country, the Northeast Corridor, already has HSR more or less for free, but would be a very expensive new maglev build.

          • Henry Miller

            @Eric2 the US is big enough that commiting to maglev automatic forces scale that brings down overall costs. You need a few years to design the tracks and equipment, than another year to build the track factories and jigs, another year to work out production bugs. However on the scale of a national HSR network it would reduce costs in the long run.

            While doing the above I hope we invest in standard stations. For aboveground stations we can save a ton of money with a standard design designed for repeatability.

            There are two problems. That need to be considered.

            First, I don’t know how much costs can be reduced. Steel rail is something easy to build, maglev track will be more expensive for the foreseeable future. I don’t know how much, but this must be considered. There is risk that it won’t get to as cheap as you predict.

            Second, the above plan can’t be done in less than five years, and we haven’t put down any track yet. Biden may not be in office and whoever replaces him may not continue the investment. Though mitigating this, just getting right of way will probably take a few years and this is something we can start now.

        • Onux

          Maglev makes a lot of sense in the US because of the distances involved. Berlin-Naples-Seville-London encompasses effectively all HSR in Europe outside of Russia (Turkey is as far away). In the US this roughly corresponds to Boston-Charleston-Dallas-Detroit, leaving out all of Florida, Canada, and even Chicago. A lot of the area between those cities in European is water, and there is of course no HSR service London-Naples or Berlin-Madrid, etc.

          Alon’s national nap from a few days ago included lines that couldn’t really support end to end travel but were there because the tails of reasonable lines touched (i.e. Chicago-NY, or DC-ATL). There were also lines that maybe just didn’t cut it such as Nashville-ATL-Jacksonville. But at Maglev speeds means that it could compete with air for NY-Chi, Bos-Charlotte and ATL-Houston travel. At that point Texas is no longer an isolated system (connect Chi-Dallas too) and the eastern half of the US can really look at making a dramatic change in travel patterns.

          Although this is all true in the abstract, Maglev carries a high cost of entry because everything has to be built from scratch. Instead of the map at the top of this post, which shows the LGV network (the high speed TRACKS, and still not all of them) look at this map, which shows the TGV network (the routes taken by high speed TRAINS, even if only part of the journey is at high speed):

          As you can see, the black lines (TGV trains on regular track) is quite extensive. Despite what Henry Miller says, there is extensive in-use or available ROW across the US (how does Amtrak get from Chicago to Quincy, for instance) that could be used to either build high-speed track or run high-speed trains on a slow route for continued service. But if Maglev is chosen, you get no service unless you commit the capitol expense of a brand new line.

          • adirondacker12800

            The tunnel from Union Station D.C. to the northern suburbs of Baltimore would be almost as expensive at the tunnel from New York to the southern suburbs of Philadelphia. A bit south of Wilmington Delaware. Wilmington is halfway between New York and D.C. give or take a few miles.

          • Henry Miller

            Most of the row isn’t wide enough to add HSR track . Unless you are going fully elevated so you can be over the slow tracks, which alot more expensive and not worth doing in rural arras with few grade crossings to close. You can make the ROW wider, but what does that save you over all new ROW?

          • Onux

            @Henry Miller

            There is normally plenty of width to add HSR track. Most US railroads have 100 feet ROW, legacy of 19th Century acts creating them. This is plenty of space to add two HSR tracks. France allows 350kph operation with 4.5m track centers, just under 15 feet. Other nations require 5m, which is 16.5 feet. That means you could get four tracks of HSR (or anything else) in just 66 feet.

            The big issue isn’t taking land to make the ROW wider, but taking land to make the curves wider, since both HSR and Maglev require very gentle curves. But this effects Maglev more than HSR since it needs even larger radius curves.

            There is an argument for the US to adopt an east-half-of-the-nation Maglev system due to distances involved, but “brand new Maglev ROW is cheaper than reusing existing ROW for HSR” is not that argument.

          • michaelrjames

            @Onux: “both HSR and Maglev require very gentle curves. But this effects Maglev more than HSR since it needs even larger radius curves.”

            I understood the exact opposite. First, maglev vehicles are ‘glued’ to the track magnetically so technically they could turn the tightest curve that the size vehicle allowed. Second, maglev vehicles are very, very light compared to standard rail that partly depend on gravity to keep them on the rails. (Incidentally this has allowed Bögl to use elevation spans about 50% longer than standard rail which obviously impacts cost and construction time.) Third, passenger safety doesn’t allow maglev to turn as tight as they technically are capable but they can handle any angle of cant–indeed they could probably hang upside down–and so the turns and maximum cant be tuned to the designed speed for that part of the track and within pax comfort limits (avoiding spilling their wine or coffee).
            Another related engineering advantage of maglev is that they can handle much steeper grades–again more limited by pax experience than physics. Also a maglev track doesn’t have to have any safety margin due to other types of trains using it. For these reasons a maglev route design would be more flexible than standard rail, and if using an old ROW into a city-centre station maglev could approach at higher speed than standard HSR.

            Here’s something citing a Transrapid curve radii graph that is now dead: (it was

            Curve radii
            The transrapid web site gives data four points of data for curve radii <ref
            here. Plotting this as a graph enables estimation of minimum curve radii for any given speed.
            Note that at 200 km/h maglev is able to turn through curves with a radius of just 705 meters. This is one of the key benefits of this technology.

          • michaelrjames

            It even applies to lower-speed maglev. This is the Japanese metro system:

            Linimo Metro maglev:
            At 100 km/h (62 mph), it is sufficiently fast for frequent stops, has little or no noise impact on surrounding communities, can navigate short radius rights of way, and operates during inclement weather.

          • adirondacker12800

            North American trains are going to be wider than TGVs. “Amtrak” loading gauge is closer to Shinkansen than TGVs.
            People in Pittsburgh are going to want to go to Chicago or New York. It would be silly to have two kinds of trains to do that. And it would mean the people who want to go from Detroit to Philadelphia or Cleveland to New York would have to change trains in Pittsburgh. Chicago is going to have the same trains as New York. Once you overlap everything it’s everything in the Eastern and Central time zones. Not the most exquisitely good choice but considering the all of the existing commuter platforms in the Northeast and Chicago, it’s good enough.
            Legacy railroads used to run equipment that spewed stuff all over the place. 100 foot wide kept most of it within the ROW.

          • Alon Levy

            Shinkansen track centers are 4.3 m. I think in Germany it’s 4.5 (and nominally trains are allowed to be up to 3.15 but in practice they don’t go there), but Richard knows this better than I do.

            Click to access spargeometri.pdf

          • Onux

            It appears I may have been wrong about maglev curves. A 705m radius for 200kph would be less than standard rail (1800m) or tilting trains (1300m). However, the point of maglevs is speed up to 500kph, and you cannot ignore physics – as speed goes up curves must widen (as you note for passenger reasons even if not for technical ones). It appears that at top speed maglev curve requirements are not much different than HSR at its top speeds. The Chuo Shinkansen is not 60% in tunneled because JR Central is afraid of the sun, it’s to keep the track (guideway? channel?) as straight as possible for high speed.

            Maglev’s are not glued to the track, just the opposite, the magnets are pushing it AWAY from the track to get it to levitate and reduce friction. No current maglev design could run upside down. You would have to specifically design it to do so, like the Wuppertal hanging railway vs regular track.

          • adirondacker12800

            This is the Japanese metro system:Linimo Metro maglev

            Has been open since 2005 according to Wikipedia. No one else has gotten the urge to build another one.

      • Herbert

        China in 2007 had more Maglev than “conventional” hsr. They still went with the latter rather than the former…

        • michaelrjames

          @Herbert: “China in 2007 had more Maglev than “conventional” hsr. They still went with the latter rather than the former… ”

          Which primarily should be a lesson on how quickly China can do stuff, really big stuff. And that the oft-reported death of maglev has been much exaggerated (including by Docteur Herr Levy). Even putting to one side low-speed urban maglev (about 6 operating lines in Japan, Korea and China) China never stopped work on maglev, based on the Transrapid technology (of ThyssenKrupp & Siemens). I thought it was particularly feckless of your nation’s corporates to abandon Transrapid but the technology lives on in several forms including cable-free elevators and variable-speed fast travellators but still some in transit, namely Max Bögl’s intra-urban system. Not exactly high-speed but at 120km/h the Chinese city of Chengdu are investigating its use in a RER-type service (I recall it might even be inter-urban). They are building a test-track and the trains were shipped (actually flown on an Antinov) from Germany in 2020.
          The Chinese claim to have significantly reduced the costs of their maglev HSR (ie. based on what Germans claims to have been stolen from Transrapid). But here’s the latest weird news (and no, the date of the article is not 01 April):

          German maglev test track set for revival?
          Keith Fender, 06 Apr 2021.
          INDUSTRIAL Plant Operating Company (IABG), which is responsible for the former Transrapid Emsland test track near Lathen in northwest Germany, has confirmed to local media that it has been approached by CRRC about re-opening the site as a test area for CRRC’s new maglev vehicles.
          CRRC has continued to develop the German Transrapid technology under license from Thyssen-Krupp. CRRC’s latest CRRC600 series maglev train will be used on the planned 400km, $US 14.1bn maglev route connecting Shanghai with Ningbo via Hangzhou in China’s Zhejiang province. The Shanghai – Hangzhou project was announced in 2020 and is expected to be completed by 2035.

          • Eric2

            There is at least one niche where maglev is the best solution: augmenting successful mid-length HSR routes already at capacity – Chuo Shinkansen is an example, and China has a number of routes suitable for this. India will too, in the more distant future when it is much richer.

            Another possibility is routes which are too long for HSR to compete well – whether NYC-Chicago where HSR is marginally competitive, or LA-Dallas where HSR is completely uncompetitive. The problem here is that precisely because the routes are long, the costs are high, even before you add the maglev per-km cost premium. So even with higher ridership than HSR, I think it is hard for such a line to become viable. (Unless you institute carbon taxes high enough to drive flights over land out of business)

            I have no idea why anyone would build low-speed maglev. You get the higher costs, incompatibility with existing rail, and vendor lock-in, with no significant benefits over rail as far as I can tell.

          • michaelrjames


            The thing about high costs of maglev, it is another trope, more established on lack of firm experience than actual real-build evidence. While it is unreliable and inapplicable to anywhere else, maybe in China itself, the Shanghai maglev came in at US$43.6 million per km including trains and stations, and it was fully elevated (and there were cost over-runs due to trouble with the soft watery Pudong soils. It also needs about half the number of trains to achieve the same capacity and frequency of service, though this doesn’t scale to long-range quite so much.
            It is also not true for low-speed maglev which ranges from $40m/km (Changsha), to $56m/km (Incheon) to $100m/km (Linimo Tobu-kyuryo).
            The German experience in the Munich airport proposal was much more cost escalation due to other issues quite unrelated to it being maglev.
            The high cost of the Chūō Shinkansen line is that it is in a tunnel for more than 60% of the entire line, and 40 m underground (deep underground) for a total of 100 km in the Tokyo, Nagoya and Osaka areas. If a forensic accountant had the tools to disinter the different factors maybe the maglev component wouldn’t be so high.
            Then the new Bögl medium-speed urban/regional maglev is claimed to be around 30 to 50 million euros per kilometre of double-track line with elevated track. No one is going to believe that until the Chengdu line is built, and even then … special deals on the first build and Chinese opacity on costs etc.

            But the real cost advantage–which becomes a huge operational advantage–is the sheer reliability and ultra-low maintenance. The Shanghai maglev has been running non-stop for about 17 years (since 2004) and engineers have said it has had the equivalent of about 2 weeks maintenance. Both the trains and the track simply don’t age in the same way as steel-wheeled system which is brutal for heavy conventional HSR. It is this feature–ultra-high levels of operational reliability–that makes it attractive to urban transit which is punishing on trains and track, and of course unforgiving of any weaknesses as the NYC subway shows. It is obligatorily fully-automated, and in operation has a near-perfect record. The only incident in all that time of operation of the Shanghai Transrapid was an onboard fire caused by a passenger. The crash of the German maglev on their test track was due to someone parking a maintenance vehicle on the line! Of course even steel-wheel HSR has a very good track record; I think France still has to have a single death of a pax from an operational accident (there were two deaths in a non-service test). For these reasons the economics of maglev needs to be subject to normalisation. There’s also the incalculable effect of time-saving, and making longer routes more viable and more attractive to passengers. This would be relevant to bigger countries like the US, and China, even Australia. (Australia’s three main cities, Melbourne-Sydney-Brisbane are awkwardly ≈900km apart. The 4th city is 4,000 km away from them!)

            Also, FWIW, look at the absurd costs of HSR in CaHSR and UK-HS2. Or London Crossrail at current $40bn and ≈US$339m per km. Naturally, even if the Chinese and Germans or Japanese build maglev at sensible or affordable costs, the Anglosphere would manage to make it much more expensive, create endless psycho-dramas and stretch it out unnecessarily to decades instead the 4 or 5 years the Chinese/Japanese/French/Spanish build longer, more complicated lines. But at least with maglev one would have a future proofed best-possible solution, most likely good for the next century.

          • Onux


            Augmenting mid-length HSR lines at capacity is not where maglev is the best solution. For capacity the solution is more tracks, not increased speed. You are suggesting the equivalent to NYC building HSR from Harlem to Wall Street because the Lexington Ave subway was crowded.

            The reason you switch to higher speed is because of greater *distance.* For “mid-length HSR routes” conventional HSR speeds are fine.

            Re: the Chuo Shinkansen, there is need for additional capacity, but maglev is overkill for the distance. Chuo will be 438km Tokyo-Osaka; at an average of 318kph (achieved Beijing-Nanjing) this would be 1:23, a full hour faster than current HSR. This is more than fast enough to capture 100% mode share. Despite higher speed, the maglev will only be 16 min faster than this (1:07).

            Chuo Shinkansen demonstrates the downside of incompatible technology. If it were standard gauge rail, trains from the Sanyo Shinkansen, south of Osaka, could route onto it. This would cut Tokyo-Fukuoka time from 5 to 4 hr, dramatically increasing ridership.

          • Eric2

            “But the real cost advantage–which becomes a huge operational advantage–is the sheer reliability and ultra-low maintenance.”

            I have not seen anyone in the industry cite this as a reason to pick maglev despite its disadvantages in other ways.

            “Also, FWIW, look at the absurd costs of HSR in CaHSR and UK-HS2. ”

            I think we can assume that however the Anglosphere manages to multiply the cost of HSR, they will manage to multiply the cost of maglev by the same factor.

          • Eric2


            But a gravity model means that even if you get a higher *proportion* of riders going from 5 to 4 hours, you get a lower *number* of riders than going from 2:07 (or whatever) to 1:07 on a more populous route. While also having much lower costs because of the shorter route.

            I did not realize the difference between regular HSR and maglev on this route would only be 16 minutes. Perhaps then it would have been better to just use regular HSR on the route. Of course riders could take HSR from Fukuoka to Osaka and switch to maglev there, which would be a bit faster than through-running HSR at the cost of a reliable transfer.

          • Onux


            This is where the concerns about Alon’s gravity model raise their head. All of the empirical evidence I have seen is that once you get HSR travel time below 2:00 (or maybe 1:30) ridership essentially plateaus, for people making intercity journeys being any faster doesn’t actually generate more trips.

            And remember even with conventional HSR the Tokyo-Osaka time drops from 2:24 to 1:23, so you would still be getting the advantage of higher ridership there plus the boost to destinations south of Oaska.

          • Eric2

            Maybe they want to prove maglev on this route so they can eventually market it to other countries like China, India, or the US. Or maybe they plan to eventually extend the Chuo Shinkansen all the way to Fukuoka. Or maybe HSR (and not maglev) is more speed limited in tunnels and Chuo Shinkansen is mostly tunneled. Or maybe you are right and using maglev here is a dumb idea. I don’t know.

  6. Matthew A da Silva

    Agreed the Midwest (with the exception of Ohio) is missing the plethora of small cities in Europe, but aren’t there small cities all over the Northeastern US that make for solid HSR or conventional-speed branch markets? Here’s a list of these small but decidedly independent metros by state, all of whom are either on or near your proposed Northeastern HSR network and could be served by an 80-100mph branch if not directly:

    Maine: Portland, Bangor
    New Hampshire: Manchester, Concord
    Massachusetts: Worcester, Springfield, Pittsfield/the Berkshires
    Rhode Island: Providence
    Connecticut: Hartford, New Haven, New London, Stamford-Bridgeport
    New York: Poughkeepsie-Newburgh, Albany, Schenectady, Syracuse, Utica, Binghamton, Ithaca/Finger Lakes
    New Jersey: Trenton, Atlantic City
    Delaware: Wilmington, Dover
    Pennsylvania: Scranton, Lehigh Valley, Lancaster, York, Reading, Harrisburg, Altoona-Johnstown, State College
    Maryland: Frederick, Hagerstown, Salisbury/Eastern Shore

    • Alon Levy

      Yes, the Northeast has a lot of smaller cities, and New England is unusually German in having many satellite cities rather than contiguous urban sprawl on the Asian, French, and American-except-New-England model. However, the Northeast also has New York, which means that intercity rail ridership is dominated by the strongest city centers in ways it isn’t here.

      • adirondacker12800

        You really have to start looking at population numbers before you blurt things out.
        “(state name) statistical areas” can be helpful, all neatly organized by the information gnomes at Wikipedia who cite U.S. Census Bureau data.
        Nice round numbers: Springfield Mass. is as far from Boston as Allentown Penna, is from New York. Allentown has more people than Springfield. Allentown is about as far away from Philadelphia as Springfield is from New Haven!! !! Get to Philadelphia Allentowners can go to Lancaster which is a bit smaller than Springfield but it’s on the way to Harrisburg. Springfielders can go to Portland and then a few of wide places in the road until the potato fields peter out into forest. Or from Philadelphia Allentowners can get to Baltimore or Washington D.C.
        ….. Bangor? the only reason Bangor gets rail service now is that shipping the lumber and frozen french fries out by rail makes more sense than trucking it. Also very round numbers there are as many people in metro Richmond Virginia as there are in the state of Maine. The whole forested potato field dotted expanse of it. It’s passenger railroad, It should go where there are people to use it. Not because there is a dot on a map.
        …Give Pittsfield and Atlantic City a whirl sometime. There is a nice flat straight ROW from Atlantic City to Frankford Junction in Philadelphia.

        • Matthew Hutton

          If you want it to pass congress realpolitik says you need to hit as many states as possible.

          • adirondacker12800

            Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut have as many Senators as Wyoming, South Dakota, North Dakota, Montana, Idaho and Utah. New York, New Jersey, Pennsylvania, Delaware, Maryland and Virginia have as many as New Mexico, Colorado, Kansas, Nebraska, Iowa and Missouri.

          • Henry Miller

            True, but a line between Kansas City and Omaha should pick up Missouri, Kansas, Iowa, and Nebraska votes. The line itself won’t have a good ROI, but it should be cheap enough to build and run as a loss leader. Stations need to be carefully located right on the border to pick up each state, but the population centers on the border anyway. There are a number of other city pairs where a small investment can pick up a bunch of votes, pick the ones you need to get your results. (those were the first two cities pairs I thought of, they might or might not be the best ones)

            Marketing is important though. You need the people in the cities served to be excited about it. You can’t just announce a line and hope people will see the value of it and so write their congressman. You need to ensure the congressman voting feels like he risks his job by not supporting the line. Right now most passenger rail in the US is a badly run waste of money, and should be killed, we need to market a better rail that is actually worth building and running.

          • adirondacker12800

            It’s 185 miles between Omaha and Kansas City on I-29. With 3.5 million people in the two of them. It’s realllllly reallly flat west of Utica New York, 198 to Buffalo on I-90/NY Thruway and would be equally cheap to build. With 3.3 million people. With Toronto, Detroit and Cleveland beyond. Or Montreal, Boston, New York and Philadelphia beyond going east. Keep going north of Kansas City you get to Sioux City with a metro area a bit bigger than Glens Falls NY. Glens Falls is part of Albany’s CSA which you would have to go through to get there. Which one should be built first? ….. Albany’s MSA is the same size as the Nebraska portion of metro Omaha… They could keep going to Lincoln!. With less people than the state of Vermont. Vermont seems to be aiming for something that is faster than driving to get to Saratoga Springs NY. Which would also be the station to serve metro Glens Falls, isn’t that tidy, and cheap? Where if things work out for faster instead of slower they could be in New York in well under two, Philadelphia in a bit over two and Washington DC in three. Or Boston in well under two. They could keep going north from Sioux City for Sioux Falls with a bit over a quarter of million people, Fargo with a bit over a quarter of a milion and Grand Forks with 100,000!!

          • Eric2

            You don’t have to build KC-Omaha. Just use the existing low speed tracks. Red state voters won’t know the difference.

          • adirondacker12800

            Half the riders because it’s slow and twice the labor and equipment costs per passenger because it’s slow. Sounds like a plan.

          • Eric2

            Almost no riders because it’s slow, and almost no labor or equipment needed to serve them.

            Doesn’t matter. You already got their votes.

          • adirondacker12800

            You give them another reason to screech that Real Americans(tm) drive everywhere and why are you wasting the money of valiant taxpayers (in some other state because we are subsidy sucking welfare queens)

          • Henry Miller

            This is about politics not usefulness. People in New York state already see some good transport so they won’t scream real Americans Drive. People in Nebraska don’t have other options, and so do scream Real Americans Drive. The politicians in New York already know Rail is useful so they won’t hear that from any voter they care about. This doesn’t mean that the congressman from Boston will vote for the system, but his vote is worth less than all those we hope to pick up out west who otherwise get nothing out of it, and so can vote against it and say they voted against Pork.

            You can’t do the slow train: it will just prove the Real Americans Drive point. Slow trains might work if we had a HSR from Kansas City all the way to NYC, because at least people who take the slow train can connect to something worthwhile. This is only a might – it is hard to see them not flying or driving, but at least it is possible to imagine “real Americans” taking a slow train if it connects to something fast.

            You will note that I didn’t connect the line to South Dakota. The goal is only to pick up some votes in congress without spending too much money. This line is ideally placed to get some votes from 4 different states. Your line might get more riders, but it isn’t a compromise to get a better system. Most of the money should be going to a DC-NYC-Boston line that doesn’t pick up either of the cities you mention (Once that is fully funded we can pick them up as part of a NYC to Toronto line) because that is where the low hanging fruit is in the US.

            Of course New York Politics is part of the problem we have with rail. A well run KC-Omaha line could do better than a Utica-Buffalo line run by the people who are screwing up transit elsewhere in NY. I can imagine transferring the most incompetent out of NYC, which would seem like a promotion for someone with political connections, but ensure that this system makes even more mistakes while allowing the useful people in NYC to do something right – in fact this might be one of the better justifications for that: a politician who cares needs political cover to get rid of incompetent yet powerful people, promoting them to a out of the way line gets rid of them without the political baggage of firing them. Without the NY construction costs (are they only in the city, or the whole state?), and good operations KC-Omaha could have a better ROI than the line you mention despite the much better connection situation that means it should do much better.

          • adirondacker12800

            Utica-Buffalo actually has passengers which seems to me to be doing much better than the non existent train between Kansas City and Omaha does. I couldn’t resist. In fiscal year 2019 Utica had more Amtrak passengers than all of Kansas or all of Nebraska. Buffalo-Depew, Rochester or Syracuse had more than both of them combined. Hudson and Rhinecliff had twice as many. In nice round numbers. Albany had more than eight times as many.

            Click to access NEWYORK19.pdf

            Click to access KANSAS19.pdf

            Click to access NEBRASKA19.pdf

            ….makes me wonder where trains would be a better value. Especially since building trains through subsidy sucking states in the Midwest in the middle of nowhere comes from places like New York.

          • adirondacker12800

            I had some time to toy with some other numbers. Try this stuff on for size.
            There will be a local contribution of some sort. Will the 19 million richer people in New York have an easier time coming up with it than the 14 million poorer people in Missouri, Kansas, Nebraska and Iowa? You want some political optics, how much easier will it be to get the Missouri legislature to come up with the money to connect Kansas City Missouri with the people in Saint Louis Missouri? That would leave the 8 million people in Kansas, Nebraska and Iowa to come up with the money to go to Omaha. Once the Kansas City folk get to Saint Louis they can get to Chicago. Once they get to Omaha they can go to Sioux City. Hmmm. Perhaps you had Des Moines in mind?

          • Henry Miller

            @adirondacker12800 If New York or Missouri want to come up with the money to build their own system what is stopping them – not me. Let Missouri build something stupid like Naylor (pop 632) to Elmo (pop 168) non stop (580km) for all I care – I’m not paying taxes on that boondoggle. At Spanish costs New York could fund the Buffalo-Utica HSR line for less than their share of the costs at NYC prices.

            This is a national proposal, and that means it needs national support. In turn that means politically we will be doing some national projects. DC to Boston should be the main focus of this. Florida with a serious long term plan to get to DC gets those in between states on board (probably we need to build some stubs in that direction to prove we are serious, but there are a few useful stubs to work with). Chicago to Detroit picks up Indiana and Ohio.

            Alon’s map picks up less than half of the states, you get a few votes from those not affected because it is the right thing, but you also get opposition that you need to bribe with some pork. KC to Omaha picks up 3 states (Missouri gets St Louis so I didn’t count them, though this probably pulls in a couple votes that were leaning the other way) at a fairly low cost, while the route is marginal it is at least useful to enough people that the trains won’t be empty.

            I’d prefer to pick up Des Monies (personally – I’ve been transferred there), but that skips Nebraska so there are less votes. If the Des Monies line eventually gets to Minneapolis it is possibly better ROI, but the total subsidy is still higher for less national votes (Minnesota needs the Chicago line more and so we pick up their votes on that). Although Des Monies has extremely outsized importance in national politics, so maybe it is better anyway?

            I’m trying to get useful non-car transit without spending too much money on pork. That means we need to bring in enough state votes that we can pass congress, without wasting too much money on things that don’t work out. If we can get the votes without this line, then so much the better – we can hold for round two when there are lines, but we need to complete the network to get ROI.

  7. SB

    Doesn’t Midwest have some small cities which doesn’t have direct service in your crayon but could be serviced with spurs(either HSR speed or more likely upgraded legacy)?
    I’m thinking of Peoria, Champaign–Urbana, Fort Wayne, Akron-Canton, Lexington.
    You stated that out of small French cities you listed only Valence has direct HSR service and rest gets spurs.

  8. ericson2314

    I wonder whether a federally funded HSR in the US is more likely to come with a requirement to upzone around the stations, or some push for local public transport. I also wonder how how much US cities would desprawl with just one of those + HSR, and maybe some sort of airline fuel tax. (All 3 or 4 is a lot to hope for…. :/)

    • Henry Miller

      I’d settle for just changing HUD, Freddie Mac, and Fanny Mae rules so that single family housing zoning wasn’t encouraged. I haven’t checked lately, so maybe that has been done, but a large part of the current mess is it is a lot easier/cheaper to get a loan in an exclusive residential zone, with a bunch of other anti-TOD features. If it is just possible to build TOD in the station areas that will help.

      Actually though I don’t worry about this much. HSR stations are most likely to be located in either downtowns, or the existing shopping district, and probably has what passes for the a knot to the local transit system (though this might not be located in a good spot to run rail). Either already has some ability to up build, and any residences close are probably apartments. These are by no means ideal (nobody actually lives near the line). This of course assumes we are not stupid about where stations are placed – I’m not confident in that.

      • Bobson Dugnutt

        @Henry Miller, it depends on local input. One of the challenges of California’s high-speed rail project was the local consultation of Central Valley cities. For cities supportive of the project, residents and city councilmembers all wanted stations in their existing downtowns.

        France’s high-speed rail company offered to build HSR for California but would not accept the all-downtown route as designed. It wanted the train to parallel Interstate 5, which is mostly agricultural land and far away from the populated Central Valley cities. These are 20 miles or more west of their downtowns.

        • michaelrjames

          @Bobson Dugnutt:

          France’s high-speed rail company offered to build HSR for California but would not accept the all-downtown route as designed. It wanted the train to parallel Interstate 5, which is mostly agricultural land and far away from the populated Central Valley cities. These are 20 miles or more west of their downtowns.

          I don’t know if that is quite accurate both for the Alstom proposal and the 20 miles (which users of the I5 to those cities have to cope with). But the French were making recommendations based on their experience of real-world HSR. As it happens this issue overlaps with the other discussion on this thread. The “airport-like” TGV stations in France, of which two have been identified (so far), namely Avignon and Aix-en-Province which are 6km and 16 km from their respective town centres. Both towns are of moderate size and already have mainline rail stations close to their centres but the ROW and geography etc didn’t fit in with building LGV directly into those stations. Just as with CaHSR the vast majority of passengers will be travelling end-to-end so that adding the huge extra cost into the centres plus the impact of the diversions on the overall transit times is a poor compromise. Further, just like Avignon and Aix, many of the passengers will be drawn from all over Provence (unlike Central Valley, this is filled with Peter Mayle’s prosperous expats who will use the TGV to get to bigger cities like Marseilles, Lyon and Paris). This is largely why those stations are built “airport-like” because these patrons will mostly be driving to the station rather than any other means. This will be even more true for Central Valley residents, so driving 10-50km to the centre versus the same distance to the HSR station just outside the town is really no difference. Note that Aix TGV station (not the downtown station) is also an interchange station for TGVs headed to various places north, south (Marseilles), east (Cannes, Nice, Monaco, Genoa) and southwest (Montpellier, Narbonne, Perpignan, Barcelona, Madrid). Forcing all these trains into a diversion into old Aix doesn’t make any sense unless the geography was compatible, and the cost reasonable.

          Another example is the Paris to Bordeaux GV (TGV L’Océane. For decades the TGV travelled about one third of the route on the LGV Atlantic (serving Rennes, Nantes, Le Mans) down to Tours then the rest was legacy track that passed thru several towns on the way to Bordeaux city. Other than upgrading the ≈300km of track to LGV it also involved building bypasses for about 3-4 towns en route. Here’s the relevant outcome (wiki): “The journey between Tours and Bordeaux is shortened by around 50 minutes. 302 km of high-speed track was built together with a further 38 km of conventional tracks that connect to the LGV.[3] The new line is expected to increase annual ridership by about five million travellers.” The previously 3 hour trip became two hours, quite a lot of the saving coming from bypassing those towns. Nevertheless, those trains are still served by trains (some of which are probably TGVs) and in fact benefit from the new faster LGV; this includes Châtellerault, Poitiers, Angoulême and La Rochelle. Not to mention the future further down the line, ie. Toulouse, Biarritz and into Spain. These future destinations and expansion of the HSR network are the main beneficiaries of that saving of about one hour. (Last time I did Bordeaux to Paris (522km) it took 3h05m and I thought it amazing, while 2h10m is stupendous!)

          Like those French towns, if there is an existing ROW, then some TGVs could serve them, ie. without compromising the operation for the majority of pax. For Avignon and Aix there are frequent minibus shuttles to their town centres. Yes, just like an airport!

          • michaelrjames

            Previous post:
            “Nevertheless, those trains are still served by trains”
            Obviously “towns”.

          • Bobson Dugnutt

            CaHSR’s routing is fraught. Before the 2008 vote, the public hearings formulated the plan that stipulated that the stations be in the downtowns. This is what had been studied, and the California courts narrowly interpret direct-democracy measures. Voters approved the bonds for the business plan, which is modeled around stations in the downtowns of Bakersfield, Fresno, Merced, Modesto and eventually Stockton and Sacramento.

            Few Central Valley residents have ridden the SNCF system. They know the San Joaquins very well, though. Supporters want that, but really fast. There’s another preference of San Joaquin riders: They want to do away with Thruway bus service. San Joaquin passengers destined for San Francisco or Los Angeles have a pained moment when they arrive at Emeryville or Bakersfield. The. Damned. Bus. Ride.

            The Bakersfield L.A. corridor is the busiest of the California Thruway services. The state owns and buys the motorcoaches, very high-end and comfortable MCI or Van Hool touring buses, and contracts Coach USA to drive and maintain them. There are abundant transfers to warrant more than one bus, and sometimes there are enough riders to bypass L.A. Union Station and continue to suburban bus stops.

            The forced transfer is the most loathed portion of the trip.

            The beetfield station approach will aggravate this issue. “We’re paying a trillion dollars for a bullet train and I still have to take a bus?!” is insurmountable if “Fresno” means the pasture in the westernmost reaches of Fresno County. You already have a mostly hostile congressional delegation in the area. They’re pretty well-known, too. One will become House speaker when the Qanon wave captures the House next year. Another undermines high-speed rail when he … check notes … is not preoccupied suing a Twitter cow.

          • michaelrjames

            @Bobson Dugnutt

            I understand that, and appreciate the politics. However, you said the stations would be 20 miles (32km) from the city centres which, first is not that far; second, probably could achieve within 10km without incurring excessive costs/disruption at the same as not incurring penalties on the HSR diversionary distance and third, most people there will be driving to the station whether it is in the town centre or 10 or 20km outside. However for some of these places what does ‘outside the centre’ even mean. They are some of the most sprawled places in the world. As Gertrude Stein said, there’s no there there.
            I understand the ‘last mile’ problem and messing around with buses. I hate it myself but that’s why I don’t live in such places. But their topography means that even if a station is in the dead centre most pax will still be travelling those 10-20km, and that’s just the people in the city proper, not to mention those in the metro area or further out still on farms etc. And of course these places don’t have any public transit worth zip; Bakersfield is notorious for it. So it begs the question of what those locals were complaining about.

            Having said that, I’d be surprised if any plan by the French did not include central stations in the bigger cities like Fresno (metro popn 972,297) and Bakersfield (839,631), inasmuch as there is a clearly defined centre. Any French city of that size won’t be bypassed by an out-of-town station (if one calls 6km out of Avignon as “out of town”). Avignon and Aix-en-Province have 92,000 and 143,097 residents respectively. Avignon is in the department of Vaucluse with 559,014 residents; Aix is in Bouches-du-Rhone but its 2m population includes Marseilles (of which Aix is really just a suburb these days; centre-to-centre is about 30km). However these are bigger destination cities because both these have a substantial summer influx of tourists, particularly Avignon for which one can catch a Eurostar in London non-stop all the way (except the reverse journey where the Brits insist you pass thru Immigration control in Lille). BTW, that’s about 1,100 km.

          • michaelrjames

            @Bobson Dugnutt

            People should not be confusing beetfield stations with the out-of-town stations like Avignon and Aix. That Haute-Picardie TGV station was a sop to locals so as to neutralise their objections to the LGV. This is the LGV-Nord that serves Eurostar and Thalys to Lille and Brussels. Most TGV services don’t stop there, and indeed it is a rare case of where the thru-trains scream past the platforms rather than on a passing track, such that they don’t allow pax on the platforms until a stopping train is stopped at a platform. Still, it is not completely useless as it gets 400,000 pax pa. destined for the two nearby towns, Amiens (popn 134,000) and St Quentin, which of course the LGV bypasses completely. Probably a fair number of those are tourists because it is in the middle of the Somme WW1 battlefields and monuments. Apparently two Eurostars per day stop there, 2h49m from London.

          • fjod

            michaelrjames – City-edge stations and city bypasses are emphatically bad practice though, as Alon has written in the past. Bypasses minimise frequencies to the bypassed station and potentially the peripheral terminus as discussed in that post. Meanwhile, city-edge stations make local transport harder and less effective by introducing another transfer/destination node far away from the city centre. Specifically, this screws up transfer-based timetabling and makes demand less concentrated; now all your local transport has to serve two main stations (central/local and high-speed) instead of just one. There is a reason that Japan, Germany, Benelux, Italy, Switzerland etc make minor routing sacrifices to have HSR serve cities’ main stations. France is the outlier here.

          • Alon Levy

            In some cases, tangent stations are okay, like Valence and Reims; this is for when you don’t and will never have much of a connecting regional rail network to worry about. But true beet fields are terrible, and in particular there’s a lot of value removed in Lorraine-TGV vs. stopping in or near Metz on the branch to Saarbrücken and Frankfurt and in or near Nancy on the mainline to Strasbourg.

          • Eric2

            China is like France in putting HSR stations at the edge of smaller cities, rather than slowing the line to serve the core.

            I don’t think there is a single answer to core vs edge stations. Obviously a small town should not get a diversion which slows the line down, but a large city should. Maybe the rule of thumb should be that if most trains on the line will stop at a city, then the station should be central, but if most trains will run express without stopping, the station should be on the line even if that means the city edge.

          • Alon Levy

            China is worse, it puts even the big-city terminals on the Moon, like Hongqiao. France only does this in the middle of LGVs at very small cities like Avignon and Nancy; the only times a major-city TGV station was not the historic one, i.e. Lille-Europe and Lyon-Part-Dieu it was in a central urban neighborhood and the plan involved extensive commercial TOD turning it into the city’s primary center.

          • fjod

            To Alon’s point – I am also concerned about non-train local transport connections (buses, trams, even bike infrastructure), and simply the ability of accessing the station without using antisocial and polluting (even electric cars produce brake particulates) motor vehicles. There are many positive externalities from stopping at the existing main station (or city centre) that go beyond rail operations and ridership.

            To Eric’s point – Yeah there is definitely a cost/benefit calculation to be made, but my intuition is that, in most cases, the central station should win out for the reasons I’ve given. What infuriates me is that it is obvious France is not doing this calculation, otherwise they would not be siting the new Montpellier station out of town – to your point, pretty much every train on this route is going to be stopping at a station of that size! The alignment through the city is also straight and wide enough (bar the station approaches, which matter less if you’re stopping anyway), so there’s no real construction cost/complexity excuse.

          • michaelrjames


            it is obvious France is not doing this calculation, otherwise they would not be siting the new Montpellier station out of town – to your point, pretty much every train on this route is going to be stopping at a station of that size!

            Isn’t it obvious that they go thru this agonising choice all the time. Like with the PACA route, it is an impossible choice unless money was unlimited and even then … In the real world they must make choices. For Montpellier, surely it is all about optimising for the many thru trains they envisage for the future, those to Narbonne, Perpignan and especially Barcelona, which will be the longest such line in Europe and one of the longest in the world (Paris-Barcelona ≈1040km; London-Barcelona ≈1500km; Brussels, Amsterdam, Munich …), and they are trying to squeeze every minute out of the line. So as to reach the far corners of the hexagon within a defined threshold (it seems they finally have thrown in the white towel on Nice). Plus, such a line needs 4 tracks, ie. +2 passing tracks, which is often a problem in inner-city stations.

            The thing is that in this discussion people seem to ignore that the old line and city-centre station still exists for all stopping services. Is anyone really arguing against Avignon-TGV station, esp. when some TGVs still go to the old station … I imagine Montpellier will be similar. Some long-term some urban reshaping, TOD etc, will happen; there is a certain Haussmannian* angle to this (and I’d say, fair enough; btw esp. in China). The comparison with other Euro countries is not really valid; not only do most of them not have comparable HSR, they are also nothing like the size of France which means more expenditure on more km, and more destinations, and pressure to reduce endpoint-to-endpoint times. Japan, especially has none of this, in that the bulk of their population lives in a very narrow corridor that spans not much more than Paris-Lyon; combined with the fact that the Japanese don’t care about running a ugly elevated line wherever. It’s no accident that it is the bigger countries, China, France, Spain, that tend to commit these ‘sins’ more than those others.

            I’m sure they make poor decisions on some of this but equally one might wait until the network is closer to completion to see how it works in the future.
            *On their blogsite wiesman writes: “In Zürich, a 6200 ton building was moved by 60 meters to make way for railway tracks”. Haussmann moved the Tour St Jacques a similar way to better align it visually/aesthetically to its park and the adjoining streets (while re-leveling the rue de Rivoli). An outrageous use of scarce resources! Off with his head! Shoulda built a megastation Hbf for central HSR there … perfectly straight route up Bvd Sebastopol to build an el for that HSR from Gares Nord & Est …

          • Bobson Dugnutt

            @michaelrjames, the push for Central Valley cities to site a high-speed rail station downtown specifically comes from a mix of sentimentality and economic practicality.

            Central Valley cities are really big small-towns. Most of the landform is single-family residences; multifamily residences and businesses tend to be along arterials. The arterials are stroads platted about a mile apart. They look like the prototypical American suburb, but function as the principal cities. The largest city is the county seat, and also has 50% or more of county population, jobs and economic activity. Very little commuting is done on highways and people go to and from work on the city stroads.

            Their downtowns are akin to small-town city centers rather than the central business districts that you’d find in San Francisco, Los Angeles or San Diego. The downtowns are really the seats of city and county government. The city hall, county hall, social services offices, jail, library and general hospital are near one another, and all of these are relatively close to the train station. Existing or former train stations serve as the hubs where all of the buses come together to make transfers possible. They might also have Greyhound stations.

            They also contain a lot of small businesses, many often run by racial and ethnic minorities able to make a go of it by relatively affordable rents. Fresno’s shop gear primarily to the Latinx community, though there’s also a Hmong presence as well.

            There are some big-city amenities, too. Each of the big Central Valley cities has minor league baseball. There are 4-year universities in each of the planned CaHSR stops at Bakersfield, Fresno, Merced, Modesto/Turlock and Sacramento/Davis. Stockton has the private University of the Pacific.

            You have to see it both ways. These communities see their downtowns not as a place of wealth generation, but as the heart of their civic lives. These preferences pose cost and design challenges weaving high-speed trains and rights of way into a built-up fabric. It means a costlier high-speed rail project, and a somewhat slower end-to-end time — to the chagrin of L.A. and Bay Area riders.

            Economically, though, the Central Valley shows some shrewdness in a centrally placed station. These downtowns are not impressive, but do have great bones — historic buildings, lots of free or paid parking lots that can be transformed into higher-value uses, and some anchor functions like the aforementioned civic buildings.

            Also, they have history as a guide to see what happens when an infrastructure project transforms its built environment. Look at the decrepitude along the fabled Route 66, not to mention the old main street highways throughout America after the Interstate Highway System was finished. There’s always the temptation to surround a beet field station with some development beside a parking garage. Maybe an office building, a hotel, or a lifestyle district. Mazel tov for the high-speed rail station, but that development diverts investment from already built-up urban environment. Investment follows the newer, shinier, younger favored quarter. The existing urban environments’ residential and commercial buildings will roll into decline as they come to the end of their life cycles.

          • Herbert

            Chinese cities are still growing at breakneck speed – what is a beetroot field today may be downtown tomorrow… Which is why they also extend metro lines to beetroot fields…

          • Alon Levy

            Shanghai has a population cap because the snob burghers don’t want people from rural China to be able to move in. Population growth in the last few years has been glacial.

    • Bobson Dugnutt

      @ericson2314, no mandates will be imposed. No mandates should be imposed.

      Density and local transit access are contingent, not conditional. Important but not essential. They pose problems that should be thought of independently and iteratively. Conditionalizing an urban form and/or required transit service in order to optimize train ridership is setting yourself up to fail. Everything will work itself out.

      Think about this. Look at the state of Amtrak right now. Aside from the Northeast Corridor, California and a couple of state-funded corridors, most of America’s lived experiences of Amtrak are one train a day or thrice a week. Many riders are even forgiving of Amtrak’s shortcomings when they ride. Many riders don’t access the train from a palatial Union Station, smaller towns have to make do with a wooden shack or sometimes just a slab with an awning.

      Despite these realities, Amtrak is the U.S.’s seventh- or eighth-busiest airline.

      • adirondacker12800

        Amtrak is the U.S.’s seventh- or eighth-busiest airline.
        Looking at fiscal year 2019 numbers:
        Because a third of it’s passengers use trains titled in it’s schedules as “Acela” or “Northeast Regional”. Throw in the state supported services that have NEC stations in their schedules it’s half.

      • Henry Miller

        Amtrak is a tourist attraction outside the NEC . People arn’t forgiving it because the point is to ride the.train. next year you take a cruise ship to Alaska and get a different vacation. The other group of people who ride amtrack don’t have other options. Minot is forgiven because the city is too small to get good airline service (unless Minneapolis has a connection to where you want to go, you get bad service Denver, and no service at all to any other major hub)

        • Bobson Dugnutt

          For many small towns and rural areas, Amtrak is essential transportation. The last attempted murder of Amtrak was in the latter years of the George W. Bush administration, and there were a lot of rural, conservative areas that mobilized against it.

          • Herbert

            Amtrak is great for “flyover” states precisely because a train does not, in fact, fly over anything…

  9. AJ

    I recall listening to a podcast interviewing French planners who talked about how the TGV distinguishes between ‘urban’ and ‘suburban’ HSR stations, where the suburban stations look like an airport without the runway, with a bus transfer facility, a rental car facility, and lots of parking. 1) is this a thing in Europe, and 2) is this something that would port well to America, particularly in the Midwest? A HSR line would certainly need to serve downtown Chicago directly, but for many smaller Midwest cities wouldn’t the ‘suburban’ approach be more appropriate given the absence of good transit in all other Midwest cities (MSP probably merits two good urban stations). If there is an existing ROW that facilitates a good urban station, then great, but for a city like Madison, it seems the suburban ‘airport’ approach is going to be the best solution.

    • Autolycus

      I think that sort of idea would definitely work in a lot of the US markets served by Alon’s concept map. I know the Southeast far better than other regions, so I will speak to that. Atlanta would be best served by: 1 station attached to the airport at least by the SkyTrain, 1 downtown station adjacent to our Five Points MARTA station and ideally part of a Multi-Modal Passenger Terminal, and 1 station heading each direction north that is somewhere near I-285 (which is our perimeter highway and the outer limits of our current transit service). “Each direction” meaning 1 on the route NW toward Chattanooga-Nashville-etc. and 1 on the route NE toward Greenville-Charlotte-DC. Another station south of town might make sense, but the airport itself is already fairly accessible to most on that side of town. The airport station would already be connected to loads of parking and rental car facilities. The stations north would probably need parking decks and possibly small rental car facilities.

      As far as planned stations in the suburbs of ATL, GDOT’s Tier 1 EIS for “high-speed ground transportation” heading to Chattanooga calls for stations at the airport, downtown ATL, a stop at Cumberland (which is basically the interchange between I-75 and I-285 and where the Braves’ new stadium is), a stop in the far suburb of Kennesaw, and then a couple of smaller-town stops about 1/3 and 2/3 of the way from ATL to Chattanooga. The stop in Kennesaw would probably have a parking deck and small rental car facilities. The stops in Cartersville and Dalton would probably have just surface parking for locals to catch the train to somewhere else.

      Chattanooga would have stops at the airport outside of town and downtown. Obviously the airport station would leverage existing parking and rental car facilities. Downtown would probably have small facilities for each, driving more of the long-term parking to the airport.

      You’d see a similar setup heading toward Greenville-Charlotte, although I think there would actually be an extra “suburban” stop in Atlanta metro before heading further afield to Athens, Toccoa, or Commerce, depending on the final route selection. Greenville, SC might see a setup like Chattanooga with a downtown and airport station, but one likely routing bypasses downtown and just heads straight to the airport. That’s unfortunate, IMO, since the airport is well outside of the “urban” part of Greenville. I would prefer a station at least closer to downtown, but Greenville is a pretty typical American city anyway, with a very small footprint of real urban environment which is bordered by substantial sprawly suburb. Chattanooga has a far more noticeable urban downtown.

    • Alon Levy

      1. I don’t know if it’s a thing in Spain? In Germany it’s rare. We sacrifice speed to reach city-center stations. Occasionally there’s a French-style station, as on the Frankfurt-Cologne line. One thing we do that is similar to Japan and prewar American that France doesn’t do is stations that are neither suburban nor city-center but neighborhood-based to stay on a mainline if the main station is a terminus, for example Kassel (cf. Shin-Osaka, or North Philadelphia for PRR through-trains from New York to Chicago).

      2. Depends. At smaller cities, like maybe Mansfield, absolutely. But the bigger the city, the higher the value of a downtown station. Often it depends on rail approaches – Youngstown probably gets a suburban-style station because reaching city center requires dealing with curvy legacy lines but the Interstate is straight, but Toledo should get a city center station because it’s easy to do so on a Chicago-Cleveland line.

      • adirondacker12800

        North Philadelphia was much more useful back in the day. Shorter walk and a faster trip to use North Philadelphia and North Broad if you lived in a suburb served by the Reading. More of a meh today because Reading trains stop on the upper level at 30th. There are enough people on the West Trenton line to scare up a through train once an hour NY-DC. Vaguely the Royal Blue but all electric with level boarding and higher speeds. Toledo could be on the Detroit-New York line and the southern route Chicago-Boston line. Not on the Cleveland-Detroit-Chicago line because there will be enough demand for Chicago-Detroit-Toronto and Detroit-Toronto-Montreal for that to have a separate line east of …. Gary?

      • Henry

        Isn’t Naples-Afragola basically this?
        HS2 will build these at Birmingham and Manchester Interchanges as well as at the East Midlands Hub.
        In the US, Texas Central’s College Station and Houston stops will be at highway junctions outside city center.

        • michaelrjames

          And dare I ask where all these so-called airport-like HSR stations are in France? I can think of is the one about 16km outside Aix-en-Province. I haven’t been to the Massy station in suburban Paris so maybe that counts. (Incidentally is Wiki wrong when it says “Thus certain trains serve at Montparnasse Station and this station simultaneously, although it is not permitted to use the TGV to travel between those two stations.”?)
          Other than Interconnexion-Est, which is actually at an airport, all the others I can think of are in city centres. Are people confusing new purpose-built stations as “airport like” when it is really just simple convergence of function? A criticism of the renovation of Gare du Nord is that it will be more “airport-like” which seems to mean that it will have far more in-house facilities for travellers including retail, and adopting arrangements to smooth the flow of the ≈900,000 pax expected in the 2030s. Didn’t the St Pancras renovation adopt very similar structures–like the opening up of the lower levels so it was a series of mezzanines (very airport like), though I have read complaints that while the station is a magnificent structure, it is poor for pax with just the usual British tat retail and very poor for restaurants & food compared to the Gare du Nord quartier.

          • SB

            “And dare I ask where all these so-called airport-like HSR stations are in France?”
            TGV Haute-Picardie station aka sugar beet station is one.

          • michaelrjames


            Airport-like station? Nah, that is a token station, barely used, amongst the sugar beet fields purely to appease some locals to get the line built. It’s almost the opposite of what was intended, as far as I understood.

          • fjod

            Avignon TGV is another, with all the car hire and parking described above.

          • Max Wyss

            @fjod: Avignon TGV has a local train connection to Avignon Centre, running every 20 to 30 minutes (and some of these trains are extended to Carpentras). OTOH, bus service to Avignon TGV is a story for itself…

            FWIW, Aix-en-Provence TGV is similar, as it primarily serves the hinterland of Marseille.

          • Yom Sen

            Here is the list of all 17 airport-like “TGV” stations, with number of passengers in 2019. I would exclude Massy TGV though since it is also a major RER hub with RER B and C and soon tramway T12 and metro M18. OTOH Nimes Pont-du-Gard has been added last year.
            Valence, Belfort and Nimes have been built at a junction with regional trains. Champagne-Ardenne, Besançon and Avignon TGV are connected by an extension to other regional lines, Aéroport-CDG, MLV and Lyon St-Ex are served by local transit extensions, RER or tramway. All other are only served by car or bus.
            Lorraine-TGV was planned to be in Vandières at the junction with Nancy-Metz but was finally located in Louvigny at a cheaper location. It could have been connected by train to not only Nancy and Metz, both 400K metro areas, but also Thionville, Luxembourg, Epinal, Dijon, etc. but has only buses instead.
            For Avignon, the “virgule” connection to Avignon Centre was built at a later stage.
            Le Creusot TGV is just a few hundred meters from the regional railway and could have had services in 4 directions (Dijon, Nevers, Paray-le-Monial, Autun) but why bother? In this kind of small urban area, “everybody” drives…
            Haute-Picardie you describe as a “token station” has actually more passengers than many small regional airports.

            Aéroport Charles de Gaulle 2 TGV 15’227’840
            Marne-la-Vallée Chessy 5’442’426
            Avignon TGV 4’108’767
            Aix-en-Provence TGV 3’594’151
            Valence TGV Rhône-Alpes Sud 2’597’471
            Lyon Saint-Exupéry TGV 2’392’925
            Massy TGV 2’214’961
            Champagne-Ardenne TGV 1’154’599
            Montpellier Sud de France 908’651
            Belfort – Montbéliard TGV 739’793
            Le Creusot – Montceau-les-Mines – Montchanin TGV 691’000
            Besançon Franche-Comté TGV 657’453
            Lorraine TGV 601’834
            Mâcon Loché TGV 468’049
            Vendôme – Villiers-sur-Loir 433’234
            TGV Haute Picardie 375’044
            Meuse TGV 235’255

          • michaelrjames

            @Yom Sen.

            Good, thanks for that. Dare I ask a supplementary query: where do you get the pax numbers? I failed miserably to find a good source, other than the few random data given by Wiki. For example I tried to compare Avignon-TGV with Avignon-ville.

            Quite respectable numbers really. CDG gets good patronage even though I am always reading that it is poor (half of what all Amtrak gets; according to Politico this morning that is 32m in 2019–can that be right, looks like a typo?). Of course that line was built more for bypassing Paris north-south and NE-south as much as serving CDG. Aix is good too, as it serves suburban Marseilles as much as Aix-ville, and presumably Marseilles Marignane airport which is only about 8km away. Likewise the Montpellier TGV is only a few km from its airport, halfway between it and the city; and actually only ≈5km from the city centre. Again its raison-d’etre is the bypass of Nimes & Montpellier to speed up that LGV to the south, but the station may pick up pax because that is a popular airport for Languedoc-Roussillon.

            Anyway with this list one can see the strategy fairly clearly. I’d have to say it makes sense in balancing speed versus coverage in a national, and increasingly international, network. Incidentally, Alon’s common complaint that it is the TGV’s slow pax egress that cripples its ability to serve lots of intermediate/smaller stations is not really true. The slowing and acceleration consumes about 8 minutes, and even more with the non-LGV routes into these towns/cities. I think Herbert cited a diversion into Lyon-Part-Dieu adds 20 minutes for this reason.

            Haute-Picardie: yeah, it gets more pax than expected (though 2nd lowest on your list). I reckon it has to be WWI tourism, more or less confirmed by those two Eurostars per day. Brits can do 11 Nov and 1 July homage as a daytrip! The main Somme battlefields and memorials are just ten km to the north (Thiepval, Poziers etc) and the Australian memorial at Villers-Bretonneux is about halfway between the station and Amiens. (It was here that the Australians fought and won the Battle of Amiens, losing 11,000, and marking the beginning of the end for the Germans.)

          • michaelrjames

            Gasp, shock horreur, it’s the end of the Enarques!
            This article claims it was provoked by the Gilets Jaune. Whatever next, Boris abolishing PPEs. Who knows, maybe because unlike Macron who is an Enarque, Boris was a classics major with no discipline to master anything, and probably hates the PPE-ers like David Cameron.

            Macron to close elite school that hothouses French leaders
            Institution has been pathway to power for country’s elite, including four French presidents
            Kim Willsher, 8 Apr 2021
            Emmanuel Macron is expected to officially announce on Thursday the closure of the École Nationale d’Administration, the elite French finishing school for the country’s leaders where he studied.
            Known as ENA, the grande école has been the hothouse for France’s top civil service and a pathway to power in the public and private sectors. Four French presidents, including Macron, have passed through its doors as have dozens of ministers and business leaders.
            Founded by Gen Charles de Gaulle in October 1945 with the idea of breaking the upper-class hold over France’s higher echelons, ending nepotism and making the civil service more democratic, it has instead become a byword for an establishment elite that critics have accused it of encouraging groupthink.

            FWIW, I’m not sure the problem is the ENA or the Enarques in quite the same clearer way that PPE and its illuminati graduates are guilty of creating and perpetuating their toxic neo-feudal economic caste system. But ENA does need to widen its intake. Macron has made significant changes to the French two-class tertiary system so he’s serious about this stuff. It’ll take a generation or longer.

          • michaelrjames

            @Yom Sen

            Thanks so much for that.
            I should have known to use French search terms, but even then I’m not sure I’d have hit the correct combination. I especially wouldn’t have thought of using “Fréquentation en gares”.

            The answer to that immediate query:
            Avignon TGV 4 108 767
            Avignon Centre 2 301 960

            Not clear if some of the Centre traffic includes many of those 4m TGV pax taking the TER into Centre? I’ve slightly lost my thread on what I was intending to infer from these data … in fact, I guess we’d need to see a time series to tell if the TGV station stole traffic from Centre or substantially grew traffic overall. I’m guessing it is the latter. Though I suppose those Brits taking the London-Avignon summer run don’t contribute much, stats wise.
            Funny enough I am more inclined to arrive in Avignon at approximately one hundredth the speed (≈4 to max 8 km/h):

          • Yom Sen

            Yes, number of passengers for Avignon Centre is very low which shows the weakness of regional rail in France, especially in Provence. Avignon is a small city with low density, however it is surrounded by many towns of 20 or 30k connected by rail to Avignon Centre (Orange, Cavaillon, Carpentras…), together they make a metro area of half a million people.
            Avignon TGV is connected to the regional network to the north only so for example from Arles (40km south) it is faster to go to Paris via Nimes Centre in spite of the detour. A better location could have made of Avignon new station a major hub for both national and regional trains, making Nimes new station unnecessary.

          • michaelrjames

            @Yom Sen

            Yes, as I wrote earlier Avignon is the capital of the Vaucluse department with population of >500,000 which comprises the Mayles heartland of hilltowns etc. As I also said, I suspect mostly they drive to Avignon ville for all the usual, and drive to Avignon-TGV for distant destinations. But for those near destinations you mentioned (Arles, Orange, Cavaillon, Carpentras etc) they drive rather than drive to Avignon-Centre for the TER etc.

            A better location could have made of Avignon new station a major hub for both national and regional trains, …
            What did you have in mind? Right-bank (Rhone, not Durance)?

          • Yom Sen

            I had in mind a location on the west bank that would allow a connection between the LGV and the PLM legacy line, maybe the rail yard 1km north of Avignon TGV. We would have had a bridge on the Rhone a bit further north and the tunnel under the hospital to the east a bit further north as well. This way, it would have become the main through station for both TGV and TER at walking distance from the city center (1.5km). It’s too late anyway… Today maybe the best would be to divert the legacy line to Marseille a bit to the west to serve the current TGV station, but that would probably be quite expensive.

          • AJ

            That would be a bit further than I had in mind, as that is also outside of the city.

      • SB

        On Madrid-Barcelona line there are Guadalajara–Yebes and Camp de Tarragona station.

      • Mikel

        Yes, there are plenty of airport-style HSR stations in Spain, but more like a provincial airfield: little more than a bus shelter and a few dozen parking spots — a bit like French beetfield stations but with more utilitarian architecture. Many of them are intermediate stations in rural areas where a 4-track section for stationing and overtakes was needed anyway, or near cities not big enough to be deemed worthy of a city-center station. Here’s a list with context for some of them:

        -Madrid-Barcelona line: Guadalajara-Yebes and Camp de Tarragona. The latter is supposed to serve a polycentric region but currently has no rail links to most of the urban stations
        -Barcelona-French border line: Figueres-Vilafant.
        -Barcelona-Valencia line: Cambrils and L’Hospitalet de l’Infant.
        -Madrid-Valencia line: Cuenca-Fernando Zóbel and Requena-Utiel.
        -Madrid-Alacant/Murcia line: Villena and Elx AV. The latter is to be connected to Elx and Alacant via a branch of the conventional Alacant-Murcia line.
        -Madrid-Sevilla line: Villanueva de Córdoba-Los Pedroches.
        -Córdoba-Málaga/Granada line: Puente Genil-Herrera and Antequera-Santa Ana.
        -Madrid-Galicia line: Olmedo AV, Medina del Campo AV, Sanabria AV and Porta de Galicia. The first two are located close to the respective towns, and the last two are in sparsely populated areas but within short driving distance of the Portuguese city of Bragança.
        -Madrid-Burgos line: Segovia-Guiomar and Burgos-Rosa de Lima. The first one sits just outside the Guadarrama base tunnel and has an awkward missed connection — the HSL and the classic line that connects to the city proper cross 2.5 km away from the station; the second one is outside the city as requested by the local government.
        -Basque Y: Ezkio-Itsaso. Again, missed connection to the existing line, since the station was added well after the alignment was fixed in place.

        In general speed is always prioritized over centrality because 1) the politicians don’t use local transit to connect to the AVE, and 2) if you reuse an existing city approach you run into the track gauge nightmare. Zaragoza and Lleida have urban stations and a bypass for nonstop trains, as will Miranda de Ebro and Loja in the future. Madrid and Barcelona have freight bypasses, and another one is under construction around Valladolid.

        • AJ

          Thanks for the examples. Those ‘beetfield’ stations seems a lot like the Lacey, Washington station, which functions as the intercity rail station for Olympia. Even with HSR (or HSR-light ), that station location is likely sufficient and doesn’t need much more than a better bus loop to serve Olympia directly. I’d imagine many Midwest cities comparable to Olympia WA would also do just fine with a simillar station placement.

    • Bobson Dugnutt

      I don’t know if both Minneapolis and Saint Paul both need a high-speed rail station. A light rail line already connects both downtowns with the university in the middle and takes about 40 minutes. Unless there’s a need to get between both CBDs in like 5 minutes — doubtful since the area between both is fully built up and an unobtrusive right of way doesn’t exist — the train can serve two points of the triangle of downtown Minneapolis, the airport and the capital.

      The Minneapolis and St. Paul points make sense is the train is due east toward Madison via Eau Claire. Most likely, though, the corridor will be due southeast, serving Minneapolis and the airport via tiny Rochester and La Crosse to Madison. Rochester will produce bountiful ridership because of the Mayo Clinic’s national reputation. Many patients and visiting health workers (Mayo is a teaching medical center) will likely arrive by train. La Crosse has a University of Wisconsin and an enrollment of 10,000 students.

        • Bobson Dugnutt

          If you need to compete with planes those 40 minutes are going to be a dealbreaker.

          @Henry, don’t kick own-goals.

          The Green Line is a thing that exists right now.

          Its ridership has three destinations: Downtown Minneapolis in the northwest, the university just outside downtown MSP, and downtown Saint Paul. There are a dozen stations serving the neighborhoods and connecting with local buses.

          It has its own travel patterns, and a connection to a not-yet-existing high-speed rail line would boost ridership to some degree. It’s improbable that it will drive ridership so much that it will eclipse the ridership patterns of today.

          I am not advancing the argument that HSR’s ridership is conditioned upon Green Line ridership. HSR ridership will also come from:
          1. Foot traffic from workers or residents in downtown Minneapolis.
          2. T bus ridership.
          3. Rail ridership from the Blue Line and North Star.
          4. Uber, Lyft and taxis.
          5. Rental cars, either from station-adjacent locations or a prearranged rental that’s brought and parked near the station.
          6. Cars, either driven by their owners or rides given by friends and family.
          7. Courtesy shuttles run by hotels, offices, parking garages, etc.
          8. Intercity buses.
          9. Evolved activities (i.e., any of the previous 8 items that don’t exist right now but will be induced by the availability of HSR).

          Also, T’s Blue Line is a thing right now.

          This line connects downtown Minneapolis, the airport and the Mall of America. An airport to downtown ride is 40 minutes, like the Green Line’s St. Paul to Minnesota leg.

          I offer you a bad-faith question: Which agency should be assigned failure? Metro Council for its slow train, or the airport authority for not attracting more light rail ridership than it gets? (The correct answer: I literally told you it’s a bad-faith question. You don’t take the bait and dismiss my framing. 😉 )

      • adirondacker12800

        doubtful since the area between both is fully built up and an unobtrusive right of way doesn’t exist
        I-94?… That would depend on how much money you want to spend.
        There appears to be what used to be a very wide railroad from downtown St. Paul to the “wrong” side of the river in Minneapolis. It ain’t what it used to be because the tracks end just short of I-35W. It’s got at least a track and freight cars on it. And something that looks suspiciously like it used to be an alternate into downtown Minneapolis on the “right” side of the river that has been built on from the funny shapes of the buildings where it used to be.
        Rochester has an airport. I don’t know why a high speed train from Chicago needs to stop at the Minneapolis Airport when Chicagoans can fly out of O’Hare or Midway.

        • Bobson Dugnutt

          @adirondacker12800, my concern with using I-94 as a viable high-speed rail ROW is that at least two lanes will be taken from the highway. A satellite view shows I-94 has three traffic lanes in each direction, a median and a shoulder. The median is less than a lane wide and is a short K-rail barrier. High-speed rail is ostensibly a highway widening project. The law says a lane that is taken from mixed-flow lanes, like for a carpool lane, must be replaced with a new lane.

          I-94 will still be three lanes, at least. There’s probably a local constituency that might say, why not four lanes? Five? Six? You might also need to add a safety cushion between the train tracks and the freeway lanes. A fast train is likely to be loud as well as whip up turbulence strong enough to push vehicles into the lane to their right.

          HSR will be a highway widening project, and the highway will be the largest cost driver that train passengers will have to pay for.

          I do see a right of way outside of downtown Minneapolis due toward St. Paul, but it looks like it’s a busy freight corridor. Expropriating it for passenger use has its own land use and takings challenges. You can avoid it by … oh, planting a bunch of pillars in the Mississippi River and have the train run over water … but this is its own environmental nightmare.

          Rochester has an airport. I don’t know why a high speed train from Chicago needs to stop at the Minneapolis Airport when Chicagoans can fly out of O’Hare or Midway.

          Rochester has an airport. But does it have flights? I’m looking at RST’s departure board. You can fly directly to three destinations: Minneapolis, Chicago and Denver. In a small plane. At great expense. With just 8 flights total per day.

          The day HSR comes, two of those flight routes are toast — Minneapolis and Chicago. A flight will win on time but will lose on ticket price, frequency and user experience — the train will have ample legroom, interesting ground-level scenery, and passengers regain the minutes of time lost to take-off and landing safely. Passengers won’t have to turn off their electronics, latch their tray tables or wait to go to the bathroom until the train reaches its cruising or approach speeds.

          Minneapolis-Chicago flights will remain viable even upon HSR’s arrival, but their frequency will be reduced. Again, over such a large distance, HSR can’t complete with a plane. It will however compete on price, frequency and user experience. The market will segment between price- and experience-sensitive passengers opting for the train and speed- and status-sensitive passengers opting for the plane. But overall, all passengers will have improved options.

          Fundamentally, what doesn’t change is that the cost basis of air travel is the highest of any transportation mode. There is also the physical constraint unique to aviation: plane seat capacity and gate capacity.

          When comparing air and rail city pairs, the train wins on cost basis even when return on capital is factored in. Consider how much an unsold seat costs the operator. A train operator can afford to not sell a seat — and in fact can hedge against it through unearned revenue (IOW, prepaid ticket books or daily, weekly, monthly, quarterly or annual passes; or bulk sales to major employers, tourist attractions, universities and the military). An unsold seat is brutal on a plane.

          Why? Network effects. A Chicago-MSP flight is dependent upon flyers going between those two airports. High-speed rail has the advantage of the formula n(n-1), with n being the stations along the line. On cost basis, the train has two advantages: more station pairs, and more seat turnover.

          Hypothetically, the Illisconsota Corridor (working title 🙂 ) will have:
          1. Minneapolis CBD
          2. MSP Airport (served because of the Be on the Way Principle)
          3. Rochester
          4. La Crosse
          4.5. Some politically determined stop(s) between 4 and 5
          5. Madison
          6. Milwaukee
          6.5. Milwaukee Mitchell Airport
          6.75. Racine
          6.875. Kenosha
          7. Chicago O’Hare
          8. Chicago CBD

          The non-integer numbers are optional stops. Milwaukee has both a CBD and an international airport, on the way, and relatively close together. The problem: To serve both requires an enormous right-angle turn in downtown Milwaukee. So for this Extremely Online proposition I am only proposing one or the other but not both.

          On Google Maps, the straight-line route is 400 miles. A 100 mph average speed gives a 4 hr. end-to-end time. At 125 mph, 3’15”. At 150 mph, 2’40”.

          With 8 stops, I get 8(8-1) = 8*7 = 56 possible stop combinations. The median distance is 200 miles, and out of the 56 possible pairs, the most pairs are most probably between the 25%-75% distances, or 100 to 300 miles.

          There will be some Chicago-Minneapolis riders, and the probability is higher with higher speeds. The point is, the train has 56 possible stop combinations, so the train can afford to lose a CHI-MSP passenger. The airlines have only 2; they cannot afford to lose any passengers.

          • Henry Miller

            You don’t stop at the Minneapolis airport, you stop a few miles away at the mall of America and so attract shoppers who like that sort of thing. You pickup the same light rail guide by the airport and complete the triangle to at Paul.

            The only question us how many people will commute to St Paul, now that there car beating connection.

          • AJ

            I don’t see a Chicago-Minneapolis run that has zero stops, so the hard right turn in downtown Milwaukee shouldn’t be an issue if the train is always coming to a stop there anyways? Other stops (Madison, Milwaukee airport) presumably will have overtakes, but I don’t see the need for that in downtown Milwaukee. That would be like running a DC to NYC express and not stopping in downtown Philly.

          • Car(e)-Free LA

            I’m not sure Chicago-Twin Cities has demand for express and local service. I’d just run an hourly Chicago-ORD Airport-Milwaukee-Madison-La Crosse-Rochester-MSP Airport/St. Paul-Minneapolis service and serve the burbs of Chicago and Milwaukee with separate regional rail. Every train should stop at Madison and Milwaukee no matter what so running express service only cuts La Crosse and Rochester–which only saves ~10 min or so of total travel time. Doesn’t seem worth it to me.

          • Bobson Dugnutt

            @Henry Miller, you stop at MSP and not the Mall because both planes and trains can enhance one another’s network effects. MSP now opens up to a pool of passengers in central and western Wisconsin and southeast Minnesota. The train now offers access to most of America and the world with just a single transfer. (This settles the argument over having a national high-speed rail network while not offering a transcontinental route.)

            You’re right, though, that Mall of America might be an attractive rail destination. It’s the Disneyland of shopping centers — with actual amusement park rides inside. It’s also mostly parking so it can function as a suburban collector station, a mall and jobs center with plenty of room for all activities. They’re also next door neighbors, though the mall is south of the runways and the two are connected by a highway. (For non-drivers, the Blue Line connects the airport and the mall. A ride between both is about 10 minutes. See the schedule at

            However, most of the Mall of America’s revenue is coming from the everyday shopping habits of people already living in the Twin Cities area. A high-speed train won’t be a ridership magnet, and it won’t outperform the Blue Line, which has a daily ridership of 30,000 or so.)

            HSR to MSP settles another argument, the one I hope will be settled so conclusively that humanity can shut up about it once and for all and find something else to think about.

            But, but, people can’t get to/from the station oh noez !1!!11

            A multimodal HSR/airport station is also stupidity disinfectant. The station access canard is both a bad-faith tactic to poison the well of discourse as well as immovable ignorance, in which a question keeps being asked or an argument keeps being advanced in order to stall an answer.

            If you follow a creationism vs. evolution debate, the discourse eventually leads to the insufferable question of “If human beings are descended from monkeys, why are there still monkeys?” The evolution side must constantly slay this zombie, and it frustrates them to no end. The evolution side had to develop a response not only to the question on its face, which is explained scientifically, but also to why creationists keep asking the question whose answer will not change upon repeated inquiry. The more savvy evolution scientists recognized that faith-based and evidence-based arguments have values that are incompatible and irreconcilable and their respective values and methods are in conflict and competition.

            Back to transportation. What happens if MSP has a high-speed rail stop? Well, we have the lived experience of an airport and a map to show that roads lead to the airport. That means rubber-tired vehicles can get to the passenger terminal areas. We can see parking garages around the airport, so we know that there will be a place to store non-moving rubber-tired vehicles. We can see cars that have a sign that says “TAXI,” a window sticker that says “Uber,” or a glowing pill that says “Lyft.” We see railroad tracks leading north and south of the station, as well as signs with a white T in a red circle. So there are buses and trains that will help circulate you through the Twin Cities. If we need to stay the night, there are hotels and short-term rentals available nearby.

            So, isn’t it fair to say that society and economics have solved the access problem in theory and practice? And while we have to create a high-speed train line, isn’t it fair to say that we could solve its access problem by … putting a station close to where people are traveling right now? It’s a hypothesis, but just might be crazy enough to work.

          • adirondacker12800

            A multimodal HSR/airport station is also stupidity
            It depends on the airport.
            Giving up and just using the airport in Rochester NY or Syracuse NY means the bus from the hinterlands to the transportation available in the big city can be more frequent. I’ve been in downtown Rochester and Syracuse. Meh. Consolidating everything at the airport means the locals who were going to drive to the station anyway because the bus system is vestigial means the enormous parking lots/garages aren’t downtown. Such as it is. And even though there isn’t any traffic, less traffic in and out of downtown. The outta towners arriving, who want to rent a car can do it at airport rates. For what it’s worth the the train station in Rochester isn’t downtown. It’s not far from downtown but it’s not downtown. Neither is the one that used to be in Syracuse, which is still there, still has the New York Central sign on it. Before they ripped out the tracks to build I-690. Albany or Buffalo airports aren’t near existing ROW. Utica doesn’t have commercial air service. HSR to Syracuse airport might be handy for the people who didn’t just give up and drive there. Utica’s spectacular station is still there and will be right next to the HSR tracks some day. It’s sorta kinda almost in downtown too.

            Billions and billions of dollars to get to PHL is really stupid. If I’m someplace that doesn’t have SEPTA service I’m closer to another big airport and why would I even consider PHL? Especially since, if I’m originating someplace more distant I pass BWI from the south or EWR from the north. Both of which already have service on the Northeast Regionals. And the foamers who want to go to PHL are thrilled that there could be a station…………….perpendicular to Suburban…. A whole 15 blocks from 30th Street. That has three ways to get there. Including one – Suburban – that could be a one seat ride to Washington DC, Harrisburg or New York. They went and built a tunnel from the PRR “side” to the Reading “side” and it could be a one seat ride to Jefferson and Temple too. For free. Just have to buy a few trains. … we’ve gotten out money’s worth out of Amfleet cars, replacing them with another round of Avelias would be nice.

            St. Paul still has it’s spectacular Union Station with tracks just outside. Minneapolis tore theirs down and converted the bridge that gave it access to the east into a pedestrian trail. I see three ways to get from St. Paul to Minneapolis, one of them a very wide almost abandoned railroad ROW most of the way. But it would be impossible to use that. And it’s going to be easy peasy to get from the airport to downtown, whichever one, even though I don’t see a good ROW. If I squint at it, it’s possible for the existing commuter trains to go to St. Paul. They don’t but they could.

            The Mayo clinic is a destination. For rich people. I’m sure they take Blue Cross/Blue Shield and probably have a charity program but it’s a destination for rich people. Who get there in their own jet or a chartered one. If they are scrimping a bit don’t have a problem with the high fare to Rochester. That is a stroll across a concourse instead of hike to train station. With baggage handlers tending to the bags. I’m sure the TSA experience in Rochester is much more pleasant the the experience in Minneapolis or Chicago, on the return trip.

          • Bobson Dugnutt

            @AJ and @Car(e)-Free LA:

            We’re in agreement. The train line will initially serve all 8 stops until trains become so crowded that some sort of pattern for limited-stop service emerges. The weakest two stations would be the smallest cities on the line, Rochester and La Crosse. I surmise that La Crosse would be influenced by university travel. Rochester receives Mayo Clinic patients year-round, but probably less so during the winter months.

            The multimodal airport-train approach also provides the side benefit of not having to overbuild express tracks and overtakes. The planes will serve that function, though as I said in this thread that regional air service among MSP, MKE and ORD will be reduced because of the unsold-seat cost disadvantage of airplanes. Nonstop MSP-MKE flights might evaporate altogether. Doing a schedule search, I found the fastest flight is nonstop by Delta, with times ranging from 1’10” to 1’30”. The lowest one-way ticket price I obtained was $109. Delta flies 4 round trips daily; 1 flight in the morning, 1 in the afternoon and 2 in the evening. (There are 1-stop flights that are 3 hours or more and with the most conservative 100 mph land-cruising speed of a train; HSR wins hands-down.)

            For the train, I’m getting a straight-line distance (between each stop; this is not the distance as the crow flies) of 325 miles. At 100 mph average; that’s 3’15”. For the train to be price-competitive, the fare would have to be higher than $31 (lowest rate on Greyhound; one-way trip time is 9 hours) but lower than $109. However, at $109, the train would have to offer at least 4 trips daily. If we set a service policy of a minimum of clock headways (60 minutes), the train wins based on access. (The average service gap for the planes is 3 hours, and flights aren’t evenly spaced.) If you miss the flight, you’d arrive in roughly the same time if you hopped on the next available train rather than waiting for the next flight.

            Bump it up in speed and the overall trip times come very close to converging. At 125 mph, you’re looking at a 2’45” trip. At 150 mph, it’s about 2’15”. If you’re Delta, you’ve got three planes that could be redeployed to other routes. If you manage MSP or General Mitchell, you have three new gates to open up.

            AJ and Car(e)-Free LA, let’s talk Milwaukee. One stop (CBD or MKE), or both? I am using the existing Milwaukee Intermodal Station as my baseline. I initially wanted to bypass it because of that really sharp curve from west to south over Plankton Avenue. I also really hate its existing location, because at first glance it looks like the station is walled off from the CBD with the highways on the north and west and the rivers to the east and south. The street view is better than the map view. Each of the streets under the overpass connects to Clybourn, and all but one lead you to Michigan and Wisconsin avenues, with the convention center just three short blocks north. Today I learned Milwaukee has a toy streetcar.

            Personally, I believe that Milwaukee might be better served with a train station whose tracks orient north-south. Also, as Wisconsin’s largest city, with a highly attractive downtown no less, Milwaukee deserves a cathedral depot.

            My first choice would be to take a large industrial swath around where the Milwaukee and the Kinnickinnic rivers fork. It’s bounded by 1st Street, Florida Street, Water Street and National Avenue. Across Water Street, vessel slips would be built for the river cruise companies. With quick access to Mitchell, this MKE cathedral terminal would offer air, land and sea access! Heck, why not plan for the future in mind and leave an area for rocket launches and cover space while we’re at it 😉 ?

            My second choice would be the southwest tip of the Historic Third Ward. Take up all of that parking from the land’s edge to Chicago Street (immediately west of the freeway). You can have the same river cruise vessel slips, as well as a path to the Pier Wisconsin cruise dock, just a quarter-mile away. That land to the east of the freeway is the venue for SummerFest, which according to organizers attracts 800,000 goers annually.

            Either of these would make serving two Milwaukee stops worthwhile. Milwaukee would be a rail hub for an upgraded Metra UPN that would connect Chicago with Milwaukee via Racine and Kenosha. There would also be a line from Milwaukee to Green Bay, possibly along the Chicago-Milwaukee HSR corridor. There would be track space if Greater Milwaukee ever pursued commuter rail, and possible intercity rail with Rockford, Illinois as a focus.

          • adirondacker12800

            air, land and sea access!
            It quite far to Montreal where the tide goes in an out. I just checked, The next low tide in Montreal is at 9:17 PM. You forgot about the tower for dirigibles and stables for the stagecoach horses and canal boat mules.

      • AJ

        I think the real destination is Minneapolis, and St Paul gets a good downtown station if it is ‘on the way.’ Bobson’s alignment also makes sense. Serving either MSP airport or St Paul downtown directly seem like good outcomes.

        • Henry Miller

          Minneapolis and St Paul border each other. The downtowns are less than 10 miles apart. HSR won’t even get up to full speed before it puts the brakes on again. Only one needs a HSR station, though the local transit system may need to provide better connections to the station. Maybe you put more stations way out in the suburbs (The airport counts for this, though I don’t think it is where I’d put a station) so people getting to work can commute faster, but the two are too close for both to have a station on the same HSR line.

          Technically Minneapolis is slightly bigger and has more people, but the difference isn’t enough to be important. Put the station near whichever downtown where it makes geographic sense and let the other ride an express train. (The green line should be 4 track – I don’t think it is, but I moved out of the area before it was built)

          • Henry

            Re: getting up to full speed: Does it matter? Minneapolis is pretty much the end of the line for HSR in that direction; optimizing through speed through the MSP metro area would only be a big deal if a lot of people would be passing through, but the next biggest MSA past Minneapolis, Duluth is only a bit above a quarter million and farther away.

            All the Shinkansens stop at Shinagawa and that’s only 8km from Tokyo, but they make that decision so that someone is not spending 1h of a 3h intercity journey on a slow local connection. And that affects way more people than just whoever’s getting on at the first stop in Minneapolis. Alon is proposing 2tph for Minneapolis services at most, and you could squeeze in a regional service there the same way domestic HS1 squeaks in the gaps between the Eurostars.

          • Henry Miller

            Depends on how the line runs. If it comes from the east, then hitting both isn’t a big deal, if the train ever continues on it heads towards Fargo (I don’t see that ever happening, but that is the direction it goes). If it comes from the south (Rochester is south) as Alon’s proposal does curving to hit Minneapolis is going to need some extra expensive real estate, and the future extension is Duluth (this might actually be worth building – I haven’t run the numbers).

            I think the best is probably putting the station right in the middle, the land is a little cheaper (not downtown), and then put in an express light rail train the goes between the station and both downtowns. The university of Minnesota has some land in the area that might work, and then the future extension to Duluth has student riders to the Duluth campus (this part isn’t a serious idea unless the UofMN can put up some money). Downtown in both cities is important, but not that important, we really need to get people from all over the metro area.

            Though your point about who cares about full speed at the end of the line is well taken.

          • Car(e)-Free LA

            That seems like a worst of both worlds option. The overwhelming majority business and leisure travelers going *to* the Twin Cities will be going to the Minneapolis CBD, with most of the rest going to Downtown St. Paul and Bloomington. Being able to walk from the station to Downtown Minneapolis hotels/offices is *extremely* valuable. If anything, locate the station in Downtown Minneapolis and introduce suburb-Minneapolis-St. Paul-suburb regional rail that can do the trip between the two cities in ~15 minutes.

            Also, Fargo and Duluth are too small and too far from everywhere except Minneapolis to every be particularly useful rail lines. There are ~50 mid-sized cities in the Midwest worth serving before them.

          • Henry Miller

            I checked the map again, we are all wrong. There isn’t right of way between Rochester and MSP/Minneapolis. So the train goes to St Paul first because anything else costs a lot more in taaking right of way. We can perhaps turn and go to Minneapolis, but the airport is a u-turn no matter how you route it – at that point why not take existing light rail from Minneapolis and save building tracks to nowhere.

            As was pointed out elsewhere, Minneapolis doesn’t have a depot to work with, so I’m inclined to say let them ride the bus/light rail (the later probably should be an express train that makes the trip fast, but not HSR). If we go to Minneapolis it should be because we are going beyond that. Perhaps a stop in Coates (will be a suburb of St Paul soon, so we can gift them a station and great service knowing they will come – but this was picked as a random city on the Rochester ROW, you can pick other places in about that area if you want), St Paul, Minneapolis, Maple Groove, someplace else and terminate in St Cloud. The purpose of these stops is fast intra city travel, not HSR. I’d expect MN to pick up a large share of the costs of this part of the line, but when we are building it anyway they may as well get some useful local transit out of it.

            Duluth is close. That matters because the costs to serve the people in Duluth is a lot less. I haven’t run the numbers, but it is believable that you can actually get a ROI from that spur: the people only need to pay for 250km of track to get there, the rest is just bonus for trains that you are running anyway. Fargo at 400km is almost twice as much track. There are a few micro cities in between to stop at, but even still it is hard to see how you get any ROI. Neither spur will be great ROI, but by completing a line it can improve the center by enough to be worth it – maybe (again, I haven’t run any numbers). Of course if you go to Duluth you don’t go to Minneapolis at all.

          • adirondacker12800

            Minneapolis doesn’t have a depot to work with
            They could build one. On one of the many enormous parking lots downtown perhaps?

          • Bobson Dugnutt

            I checked the map again, we are all wrong. There isn’t right of way between Rochester and MSP/Minneapolis.

            There isn’t an extant right of way in the direction we’d need it, but it can be stitched together with some existing right of way and new track in a vast wilderness between Rochester and MSP.

            First, we have to decide on what constitutes the depot in downtown Minneapolis. The farthest one, but the one closest to the heart of downtown is across from Target Field (near the North Star platform). It would need 3 miles of subway, and it would likely consolidate a two-level track like under San Francisco’s Market Street with light rail and HSR beneath the street, so it’s a very complex megaproject.

            Another option is to put the depot at the University. There’s a freight rail yard near the stadium that could be repurposed for passenger rail and allow for access to St. Paul. It would need 2 miles of subway and cross the river.

            The third option, requiring the least amount of subway, 1.5 miles, would be a station between US Bank Field and I-35W. The land right now is parking and a low-rise office.

            The subway portal south of downtown would be around 28th Avenue next to the Hiawatha Highway. There’s a right of way to the east of the highway — the Blue Line right of way is to the west — that extends to a park around Minnehaha Creek.

            Here’s where we’re going to need another engineering marvel. At the creek, we’ll need a tunnel to get us to MSP Terminal 1. Under the airport, the tunnel will need to curve southeast toward Fort Snelling Park. It measures about 6 miles of tunnel. There’s some track near the I-494/I-35E junction, and there’s either a rail right of way or in/around Highway 55/52 that runs through mostly rural land all the way to Rochester. In Rochester itself, there is an at-grade track that could be claimed within the downtown for elevated or underground service.

          • adirondacker12800

            The train doesn’t need to be going the maximum possible speed until the instant it needs to start slowing down for the platforms. It would be okay if the last 15 miles, 10 of them between St. Paul and Minneapolis are slower.

      • AJ

        I think it’s reasonable to ask if HSR can do double-duty as an intracity alignment in MSP and consider if the Blue or Green corridors would be better complemented with the HSR alignment. In either case, the ‘slow train’ would still have good utility, with the HSR alignment providing express service. Same would likely be true in Chicago, as metro Chicago would have multiple HSR stations?

        • AJ

          I’ve had a simillar thought for Seattle. I agree with Alon that Cascadia doesn’t merit true HSR, just better intercity rail. However, for the segment between Seattle and Everett, it may be worthwhile to create an entirely new alignment (HSR or otherwise) because of the ability to provide great intracity express service, in addition to the improvement in intercity travel. I’d imagine this will hold true for other American cities, where new HSR investments can be leveraged into better intracity travel?

          • FDW

            @AJ, yeah. California is already well along in doing this, and you’re seeing this take shape in the Denver area (Where the Front Range Corridor is being used to finish/iterate upon Fastracks).

            And in Japan, there were a number of Rapid Transit/Regional Rail projects that were built alongside the Shinkansen, so it’s not like this stuff is completely without precedent.

        • Bobson Dugnutt

          @AJ, when I was eyeballing how to connect downtown Minny and MSP, I saw the best approach is a freight track just east of the Blue Line route on the other side of the highway. This would need to be re-engineered so obviously the HSR track doesn’t run at grade. The highway is a mix of interstate-class grade-separated freeway and then a classic surface-level highway with signalized intersections.

          This adds to the complexity of the project, but if the HSR tracks can be raised/lowered, so can the highway and light rail. The operational dividends to light rail are either time-savings — a one-way trip between downtown and the airport is 40 minutes — or rapid-transit service with 2-3 minute frequencies if fully grade separated. The light rail stops are a mile apart, so there isn’t much cruising speed savings, but there is savings if the train doesn’t have signal priority.

          I also proposed a downtown Minny subway. The most expensive project would be a 3-mile double-deck tunnel with HSR getting one level and the Blue and Green lines getting the other, with a cathedral style train station next to Target field. I’d also envision a regular-speed rail service, say 75 mph cruising speed, that would link the Twin Cities to the few population centers in the state. It would be akin to commuter rail service with clock headways, like what Denver and Salt Lake City have. These would serve four of the five Minnesota “points”. The Rochester-La Crosse connection would be the first point. Point 2 is an easterly line that would originate in Minneapolis and serve St. Paul, terminating in Eau Claire, WI (like La Crosse, it has a sizable state university). Points 3 and 4 would originate in St. Paul, serve the Minneapolis cathedral station and serve Duluth and Brainerd (3), and Fargo, ND (4) via a branch at St. Cloud, which is part of the Twin Cities CSA and original terminus of the North Star. Point 5 would also serve St. Paul and Minneapolis but head west and southwest to Mankato.

          • adirondacker12800

            If the high speed tracks pass LaCrosse why would you send a slow train there?

          • Bobson Dugnutt

            @adirondacker12800, to clarify: Rochester and La Crosse would be the high-speed rail line, they’re on the way to Madison, Milwaukee and Chicago. There’s no need for a slow train other than to share the Blue Line and highway right of way for a few miles to downtown. And the southern edge of the Twin Cities turns rural just a few miles south of the Mall of America, with little in the way to Rochester to need to stop.

            It’s the other 4 points (Eau Claire-east, Brainerd-Duluth northeast, Fargo-northwest and Mankato-southwest) that would get non-HSR trains. Ideally, they’d use legacy track and get slightly above car speeds along a comparable highway to be competitive with driving. The project is radial in nature, intended to draw Minnesota (and North Dakota) into the Twin Cities’ orbit. Since there’s nothing comparable to the Twin Cities’ size in any direction, other than the corridor the HSR will serve, the only other rail option that would make sense is to connect Minnesota’s next-biggest metros, and there are very few of them. Most of them are the collar counties around Hennepin and Ramsey and are already in the Twin Cities’ orbit, St. Cloud is in the Twin Cities’ MSA (which is why two train lines will serve it), and then you have Fargo (on the MN border), Duluth and Brainerd (the route goes out of the way because Brainerd is more populated than the rural counties along I-35), and Mankato. These are the only metros in Minnesota with >100K population. Eau Claire in Wisconsin is similar in size and function to La Crosse; both have a U of W with similar student enrollments.

  10. Herbert

    Chemnitz is definitely “on the way” for the unfortunately, the is dragging its feet on electrifying the southern section of it…

    And Rostock would be “on the way” if the is ever built… Granted, that is a bit of a taller order, but it does involve Nordic countries which have deeper pockets and (especially in the case of Norway) tunneling expertise…

  11. Herbert

    In Germany there is an implicit assumption that cities above 100k (“Großstädte“) ought to have hsr. There are 81 of them in Germany and some immediately abut the municipal boundaries of another (e.g. Nuremberg, Fürth, Erlangen). Of course such a way of looking at it is incompatible with one based on “metro areas” which are pretty much undefined in the German context

  12. bruce hain

    I think you might find the lack of several smaller cities (as opposed to the small towns and lesser things that exist) is due to the lack of passenger rail in the US – not the other way around – many of these otherwise prosperous and dynamic places having been dumped by the railroads in the last century and never able to fully recover: their population, and raison d’tre as places of business + commerce. I see the Nickle Plate as the most preferred route between Chicago and Detroit for passengers, with the part west of Toledo (about half – but Toledo is bypassed) having been abandoned for fifty years or so, so the places along there in Indiana are much less developed and populous than along the part in Michigan, where at least they have direct freight service. It would be a boon to all the economies of the affected towns in both states to have direct passenger service: It can’t be provided by air in such a way that would be regular and economical enough to make any difference otherwise. (One feels God really had this planned in the scheme of things and for some reason we didn’t get the idea.) Where strings of better-developed towns come along nearby – as with the routes of the NYCentral and PRR running N. and S. of the Nickle Plate Route in Indiana to the west – there would be great advantage to tying service to places like South Bend (NYC) and Valparaiso (PRR) which could also be easily done (like reviving the Nickle Plate, i.e. easier than any other option given the benefits) given the large number of leftover rail lines around in this not-very-developed region along the non-existent Nickle Plate – which would turn the Nickle Plate towns into TOD rail transportation nodes as well, for getting passengers to the fast line.

    • Alon Levy

      A couple points.

      1. “Prosperous and dynamic” is not how anyone in France refers to its smaller cities. On the contrary, there’s the same declinism as in the United States, positing that globalization, highways, and hypermarkets have hollowed out small cities, which are held to be inherently more moral than either Paris or growing cities like Toulouse and Bordeaux. (Of note: Southern France’s main cities are kind of France’s Sunbelt in terms of population growth, though not in politics.)

      2. The cities I bring up as vaguely orbiting Lyon are a lot smaller than Toledo – the only ones even vaguely comparable are Grenoble and Saint-Etienne.

      3. It’s not the US that’s unusual in its metro area population distribution, but Europe. In Japan, half the population lives in the Tokyo, Keihanshin, and Nagoya regions, and in Korea, half the population lives in the Seoul region. It’s notable that for all of the extreme capital centralization people speak of in France and the UK, their capitals are around half the size of metropolitan Seoul even with slightly larger national population than South Korea.

      4. At the metropolitan level, better regional rail makes it easier for urban areas to agglomerate, not harder – it’s not a coincidence that France, where public transportation outside the main cities is horrendous, has a much smaller median metropolitan population than Germany. But even then, there’s a Europe effect reducing the size of the larger cities. Israel is comparable in population to Belgium, Sweden, and Switzerland, but metropolitan Tel Aviv is a lot larger than Brussels, Stockholm, and Zurich, and that’s with very weak regional public transportation.

  13. bruce hain

    Interesting. So much seems to depend on some characteristic of the place – that would make people want to agglomerate or otherwise – rather than cultural or economical incentives impelling it. While Americans living in out-of-the-way places like what you find along the Nickle Plate in Indiana probably do indulge in some degree of provincial jingoism (“We’re all rednecks and the better for it!”) and might be inclined to rankle at the idea of having a major high-speed trunk line running through their back yards (actually, in my version of things, THE high-speed trunk line – since the NYCentral “UPGRADE”/replacement ((easiest/first to be achieved)) would detour to Canada ((Niagara Falls)) with an easy connection to Toronto, having the advantage of avoiding likely trackage disputes bet. Buffalo and Cleveland – and also serving Detroit with its big population, some of it well-to-do – whereas the the PRR “REBUILD”/replacement would serve Cleveland on it’s main line – and BOTH would funnel into the old Nickle Plate lines – at Detroit and Toledo respectively – to complete the trip to Chicago, taking advantage of a short ((less than a mile if I remember correctly)) new-build connection to the existing, once-B&O line to gain access.) (well there was a sentence back there somewhere) Oh, “rankling at rail lines” yes: Well, THAT I think would be easily overcome if the advantages of what would happen to the value of their property and economic fortunes of their… fiefdoms were convincingly put, with those interests trumping all objections.

  14. Bobson Dugnutt

    @Henry Miller:

    You will under perform the model though because the lack of a good transit system means you are either paying a lot of money to get around via rental cars or taxi (uber is a taxi), you need to rely on someone agreeing to drive you, or you have to arrange your life around what passes for local transit: just driving is looking like a good option.

    Conditioning high-speed rail’s ridership around the availability and abundance of local public transit works off of a flawed premise.

    Transportation for local circulation and for intercity circulation are related but independent of each other.

    Connections between high-speed rail and local transportation are one subset of demands for both, and they’d overlap at the station. However, among the total population of transit users, intercity travelers are going to be a small segment of overall ridership. This holds true if we’re talking about a bare-bones transit system in the Midwest or a global city like Paris, where local transit and high-speed rail service are excellent. A substantial portion of SNCF’s ridership might have arrived via RATP, but RATP’s ridership from SNCF passengers is substantially smaller than its everyday ridership of Parisians circulating within its service area.

    An important distinction to be made is how the benefits and costs of local and intercity transportation accrue. Local transit’s benefits of mobility accrue entirely within its service area. Ridership (how many people use a service) and productivity (how many people use the service in a given space and time) are influenced by many outside factors beyond supply alone, but the clearest relationship is 0=0. If there is no service supplied, there is no service demanded. In order to benefit from Paris’s excellent public transportation, you actually have to be within Paris. (It doesn’t work both ways, either. Do Boston’s commuter rail service or Las Vegas’s bus system compare to Paris? Clearly no. And Boston’s rail system and half of Vegas’s bus system literally have the French high-speed rail company operating them!

    Intercity’s costs and benefits are classic examples of network effects. The ridership and productivity aspects are derived from the network itself and the ability to exchange (people, goods or data) both ways. The localized benefits to a participant (a rider, a producer or a community member) derive from the ability to cooperate and compete within the network as well as proximity to other participants.

    We get into trouble when we fail to see the One Big Difference. It leads us to the cognitive trap called natural kind.

    It’s the same fallacy that conservative populists invoke when they implement austerity policies because “a national budget is just like a household budget.” It’s also like a dean telling a college’s sciences department, “We shouldn’t be spending money on both microscopes and telescopes. What’s the difference anyway? At the end of the day, they’re both lenses.”

    You can create a composition and evaluate whether it’s true that local transit ridership determines intercity (air/train/bus/whatever) ridership. For instance, draw a tic-tac-toe grid and categorize the three columns as A-Local transit agencies, B-Transfers btw. A and C and C-Intercity services (airports or train stations). The rows would be 1=Top third of boardings; 2=Middle third; 3=Bottom third.

    My hypothesis is that the categories are going to be so scattered that there are more exceptions to the rule than the rule itself. Atlanta has the world’s busiest airport, but MARTA would be No. 9 in ridership (combined bus and train). What we’d need to know is whether the airport train station is a major driver of ridership; Wikipedia says the station produces 9,100 daily boardings. MARTA’s daily train ridership is 175,000 — 5% of total boardings is insignificant. So MARTA would be A1 and C1 (top) for transit and air; but B3 (worst) for interaction.

    New York City is obv A1 for transit ridership, and if you combined JFK and LGA it would be No. 2 for air service, beating out L.A. (if you counted L.A.’s satellite airports too like BUR, LGB, SNA and ONT it would be back at No. 2 and spitting distance of Atlanta.) A1 and C1, but bus and train service to the airports underwhelms and contributes very little. Same with L.A., where transit and airport ridership are No. 2, but the connection between the two is minuscule.

    Chicago might be the only case where you have high transit ridership, high airport passenger traffic, and a clearly high relationship between the two. Wikipedia says O’Hare (Blue) is No. 5 and Midway (Orange) is No. 16 in use. So Chicago can claim top-row status in all three categories.

    The only other top-tier city in the U.S. I suspect might be Seattle. Its light rail touches a parking garage that’s disconnected from the main terminal, but it still maintains several bus lines that serve Sea-Tac, helping to put it in top or middle-tier status.

    In the Sunbelt, you’re going to have airports that have higher daily activity than their transit systems, and service interface is palliative mobility at best.

    • adirondacker12800

      whether the airport train station is a major driver of ridership
      People work at the airport…………

      combined JFK and LGA
      It’s easier to get to EWR for a lot of people.
      While it is possible to air freight automobiles few people, other than heads of state, do it. All these people the naysayers will drive everywhere manage to get to their airport and not have an emotional collapse when they find their automobile didn’t tag along.

      • Bobson Dugnutt

        @Adirondacker12800 wrote:

        While it is possible to air freight automobiles few people, other than heads of state, do it. All these people the naysayers will drive everywhere manage to get to their airport and not have an emotional collapse when they find their automobile didn’t tag along.

        Exactly! People will figure out their options based on time, cost, needs and wants. Life will more or less look and work the same as now, but with a better mobility option that, if we get this right, future generations will see as so banal that they won’t be curious about the metaphorical and literal mountains we had to move in order to make this world possible.

        • adirondacker12800

          I’ve been told that it’s much cheaper to drive to Buffalo from Ontario and fly to U.S. destinations from there. Buffalo’s airport had just over 5 million passengers in 2018. It’s also the airport for Niagara Falls which likely affects things some.

          Rochester airport, in 2017 according to Wikipedia, had 2.357 million passengers. It’s out in the middle of nowhere. Which, because it’s Rochester, not Los Angeles is quite close to downtown. It is, the cab ride is short, uncongested and fast. The train station a few blocks from the center of downtown, such as it is, had 127,203 passengers. Wikipedia has a top destinations table. Atlanta and Charlotte are on it. The rest of them are in HSR range of Rochester. Three of the eight are EWR, JFK and LGA.

          Syracuse Airport had 2.315 million in 2018. The train station is vaguely out that way. Syracuse’s article doesn’t have a top ten list but they also have service to EWR, JFK and LGA. Syracuse had 131,881 Amtrak passengers in 2017.

          Albany’s airport under performs. It had 1.518 million in 2019. And only had service to EWR. Amtrak had 806,960 at Rensselaer, 237,268 at Hudson, 62,180 at Schenectady, 41,611 at Saratoga Springs and almost none at Fort Edward or Whitehall.

          Each of them had, in very round numbers 2.4 trips per person. The ones in Albany are finding the train stations.

    • mindstalk0

      I’m not sure that addresses the objection.

      The distances that HSR excels at (competing with air) are a few hundred miles. Many Americans are happy to simply drive that distance. HSR might be tempting, but if they feel they would need a car at their destination, then it can make sense to simply drive their own car, rather than taking HSR and renting a car.

      For longer distances most people are flying even if they have to rent a car at the destination, because they don’t want to spend a day driving. Destination transit has little effect, and airport busyness is irrelevant. It’s not the people flying into Atlanta or O’Hare who are potential HSR customers, but the people within a few hundred miles who are currently driving (or not making the trip at all.)

      • adirondacker12800

        The places they need their car at aren’t the destinations HSR serves. Origins perhaps but not destinations.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.