High-Speed Rail and Connecting Transit

Noah Smith is skeptical about high-speed rail in the United States. He makes a bunch of different arguments against it, but I want to zoom in on the first, the issue of connecting transit, which Noah is far from the first person to bring up. It’s a genuine drawback of rail planning in the United States, but it’s very easy to overrate its importance. Connecting transit is useful, as is the related issue of city centralization, but its effect, serious as it is, is only on already marginal high-speed routes, like Atlanta-Memphis or Dallas-Kansas City. Los Angeles suffers from lacking connecting transit, but it’s also so big that nothing it connects to is marginal. Finally, high-speed rail and urban centralization are not in competition, but rather are complements, as in the history of the TGV.

Connections and centralization

Modal choice is about door-to-door trip times. This is why a large majority of people take a train that takes three hours over a plane that takes one: hardly anyone lives near the airport or has an airport as their ultimate destination. In practice, people are much likelier to be living near and traveling to a destination near a city center station.

The importance of connections then is that connecting urban transit extends the range of the train station. I didn’t live at Gare de Lyon or Gare de l’Est, but I could take the Métro there and it was a short trip, much shorter and more reliable than taking the RER to the airport, which made it easier for me to ride the TGV. With reliable connections, I showed up at Gare de l’Est four minutes before a train to Saarbrücken was due to depart, printed my ticket on-site, and walked leisurely to the platform, boarding still with two minutes to spare.

Regional rail has the same effect, at longer range. It’s not as convenient as urban rail, but it feeds the main intercity rail station and is timetabled, so if the system is punctual, passengers can time themselves to the main train station. In Switzerland the connections are even timed, enabling people who travel from smaller cities like St. Gallen to points west to transfer at Zurich Hauptbahnhof within a short window. However, this is completely absent from France: the regional trains are unreliable, and Paris has through-running on the RER but no single central station that can collect connections from secondary centers like Meaux or Versailles.

Finally, centralization is important because the reach of an urban transportation system is measured in units of time and not distance. Even racists who are afraid of taking the trains in Paris and rely exclusively on cars can take a cab from a train station to their ultimate destination and be there shortly. The average speed of the Métro is low, around 25 km/h, but Paris’s density and centralization mean that it’s enough to connect from the main TGV stations to where one lives or works.

But the US doesn’t have that, right?

What Noah gets wrong is that the US has connecting transit as in Paris in a number of big cities, and nearly every even semi-plausible high-speed line connects to at least one such city. Here’s Noah on New York:

The best thing about using the Shinkansen in Japan is that you can get to and from the high-speed rail station using a dense, convenient network of local trains. In America there is no such network. Thus, when I imagine taking the train from SF to L.A., I imagine taking a scooter or an Uber to and from the train station. In L.A., which is so spread out that I probably won’t stay in a small area, I imagine I’d rent a car. That’s a very different experience from using the Shinkansen in Japan. And in NYC, it would mean dealing with the nightmare that is Penn Station — a thoroughly stressful and inconvenient experience.

Let’s discuss New York now; Los Angeles deserves a separate section in this post. Noah lived on Long Island for years; he could connect to any intercity train by taking the LIRR to Penn Station and changing there. It’s this connection that he describes as a nightmare. But the question is, a nightmare compared to what? It’s clearly far less convenient than the timed Swiss connections, or even untimed connections between the Berlin S-Bahn and intercity trains. But the LIRR is a timetabled train, and while delays happen, they’re measured in minutes, not tens of minutes. Passengers can time themselves to arrive 10 minutes before the intercity train departs, even today.

All of this gets easier if a minimally competent agency is in charge and track numbers are scheduled in advance and printed on the ticket as they are here or in Japan. Penn Station is crowded, but it’s not a stampede crush and people who know their commuter train arrives on track 19 and the intercity train leaves on track 14, as written in the ticket, can make the connection in 3 minutes.

The secondary transit cities of the US are dicier. Their modal splits are all in the teens; San Francisco (excluding Silicon Valley) is the highest, with 17.5%. In that way, they’re comparable to Lyon, Marseille, Nice, Bordeaux, Toulouse, Strasbourg, and Lille. However, the way non-New York transit systems work in the US is, the system is usually semi-decent at ferrying people to and from city center, it’s just not strong for other destinations. In Boston, for example, people could transfer to the subway at South Station or Back Bay and cover a decent chunk of urban destinations. It’s nowhere nearly as good as the options for Paris or Berlin, but it’s not the same as not having any connecting transit.

Destination centralization

The connecting transit critique of high-speed rail in the American discourse goes back at least to the Obama era; Richard Mlynarik used it to argue against what he views as inflated California HSR ridership expectations, and everyone who commented on transit blogs in 2008-9 had to address the critique in one way or another. In 2012, I posted about the issue of destination centralization, that is, that destinations are more centralized than origins, especially at long distance. For example, at the time Manhattan had 22% of New York metro jobs, but 36% of jobs involving out-of-county commuting – and the longer the trip, the likelier one’s destination is to be in Manhattan.

The data I looked at was the distribution of five-star hotels, which are incredibly centralized. Depending on data sources, 50 out of 56 such hotels in metro New York were in Manhattan, or perhaps 36 in 37. In Boston, either all are in Downtown or Back Bay, or all but one are and the one is in Cambridge, a few Red Line stops from South Station. In Philadelphia, they’re in Center City.

In New York, there are clusters of lower-priced hotels outside Manhattan. The biggest such clusters are in strategic locations in Queens, Brooklyn, or North Jersey with maximally convenient access to Manhattan, where tourists and business travelers cluster. Some hotels serve suburban office parks, such as the various Central Jersey hotels I would go to gaming conventions at, but they’re smaller and lower-end.

In the Bay Area, Richard argued in favor of the primacy of San Francisco over San Jose by citing broader data on interregional travel. San Francisco, per his dataset, absolutely dominated. More recent data can be seen here, measuring tourism revenue rather than visitor numbers, but San Francisco with 900,000 people is about comparable to Santa Clara, Alameda, and San Mateo Counties combined with their 4.4 million people. There is also a comparison of international arrivals to San Jose and San Francisco – there are several times as many of the latter; I cannot find domestic arrival numbers for San Jose that might compare with San Francisco’s 26 million visitors in 2019.

The upshot is that high-speed rail does not need to connect two strongly-centered cities to be comparable in ridership to existing lines in Europe and East Asia. It only needs to connect one. People may need to drive to a park-and-ride or take a taxi to the train station, but if their destination is New York or any of the secondary transit cities of the US, it is likely to be fairly close to the train station, even if most employment isn’t.

The Los Angeles exception

Noah is on stronger grounds when he criticizes Los Angeles. Even Los Angeles has 1.5 subway lines connecting to Union Station, soon to be augmented with the Regional Connector, but the city is weakly-centered, and a car or taxi connection to one’s ultimate destination is likely. Moreover, the destinations within Los Angeles are not centered on Downtown; for example, high-end hotels are the most likely to be found on the Westside.

However, there are two saving graces for trains to Los Angeles. The first is that Los Angeles’s transit ridership is so low because the city’s job geography is so decentralized that the network is bad at connecting local origins with local destinations. If it is guaranteed that one of the two points connected is Union Station, the city’s network is still bad for its size, but becomes usable. The under-construction Westside subway will open later this decade, providing decent (if not good) connectivity from the train station to high-end destinations in that part of the region.

The second and more important saving grace is that Los Angeles is huge. The absence of connecting transit is a serious malus for intercity rail, but people can still take a taxi, and that may add half an hour to the trip and a cab fare, but we know what adding half an hour to a three-hour train trip does and it’s a 1.5th-order effect. A 1.5th-order effect can turn a line that is projected to get a marginal 2.5% return on investment into one with a below-cost-of-capital 1.5% return. It cannot do this to lines serving Los Angeles, none of which are economically marginal, thanks to Los Angeles’s size. On my map, the only line connecting to Los Angeles that a straight gravity model doesn’t love at first sight is Los Angeles-Las Vegas, and this is a connection we know overperforms the model because of the unique tourism draw of Las Vegas.

On the same map, the other connection that everyone (including myself until I ran the number) is skeptical of, Atlanta-Florida, has the same issue as Los Angeles-Las Vegas: it connects to a very strong tourism region, and the train station would serve the biggest tourist attractions. (This is also true in the case of Los Angeles, where Anaheim is still supposed to get a station within a short shuttle distance to Disneyland.) So my model thinks it’s only 2.5% ROI, but the strong tourism volume is such that I am confident the model remains correct even with the malus for weak job centralization in both Atlanta and the cities of Florida.

High-Speed Rail and Connecting Transit

Noah makes a broader point portraying intercity and regional public transport in opposition:

Building high-speed rail without having a usable network of local trains instinctively feels like putting the cart before the horse. If I had a choice between being able to train around San Francisco conveniently, or quickly get between SF and San Jose, I’d choose either of those over being able to take a Shinkansen-style train to L.A. or Seattle. The lack of local trains and fast commuter rail simply limits my travel options much more than the lack of high-speed rail. A local train network without HSR is great; HSR lines without local trains seem like something that’s at best slightly better than what we have now.

And yes, I realize that money earmarked for “high-speed rail” sometimes goes to create faster commuter rail, and that’s good. But that doesn’t answer the question of what these maps are for.

Noah is pooh-poohing the connection between intercity and regional transit as “the money sometimes goes to create faster commuter rail,” but he’s underestimating what this means, in two ways.

First, on the Northeast Corridor specifically, any improvement to intercity transit automatically improves commuter rail. The reason is that the most cost-effective speed treatments there are shared. By far the cheapest minutes saved on the corridor come from speeding up the station throats by installing more modern turnouts and removing speed limits that exist due to agency inertia rather than the state of the physical infrastructure. Trains can save two minutes between South Station and Back Bay alone on a high seven to low eight figures budget for rebuilding the interlocking. These improvements speed up commuter rail and intercity rail equally.

Moreover, in higher speed zones, it’s necessary to invest in organization before concrete and schedule trains with timed overtakes. But this too improves the quality of regional rail. Boston-Providence trains need to be electrified and run faster to get out of intercity trains’ way more easily; even with trains holding twice for an overtake, this speeds up Providence-Boston travel by 15 minutes even while adding station stops. New York-New Haven trains had better run faster on both short- and long-distance connections – and the difference between improving intercity rail this way and in a way that is indifferent to integration with regional rail is the difference between doing it for $15 billion and doing it for $150 billion.

And second, in cities that are not traditional transit cities, high-speed rail is a really good catalyst for expanding a central business district around the station. The best example for this is Lyon. Lyon built a dedicated central business district at Part-Dieu, the Metro, and the LGV Sud-Est simultaneously. This was not sequenced as local transit first, then high-speed rail. Rather, the selection of the site for a high-speed rail station, within the city but just outside its traditional center, was simultaneous with the construction of the new business district and of an urban rail system serving it.

This is particularly useful for cities that, by virtue of size (Dallas) or location (Cleveland) could be high-speed rail hubs but do not have strong city centers. In Cleveland, demand for housing in the city is extremely weak, to the point that houses sell for well below construction costs, and demand for city center office space is likewise weak; but a train that gets to Chicago in 2 hours and to New York in about 3:15 can make the area immediately around the station more desirable. In Dallas this is more complicated because it would be the system’s primary city, but a location with convenient rail access to Houston is likely to become more desirable for office space as well. This is not in competition with local transit – it complements it, by giving existing light rail lines and potential commuter rail lines a meatier city center to connect suburban areas with.

Public Transport and Scale

Noah asks what the proposal maps are for. The answer is, they are proposals for improvement in passenger rail. There is a real issue of scale and details, which is why those maps don’t depict literally every connection. For that, there are smaller-scale maps, in the same way there is the TransitMatters proposal for Regional Rail in the Boston area, or maps I’ve made for timed connections in New England and Upstate New York between intercity and regional trains. At lower-altitude zoom there’s also the issue of local connections to buses.

A roadmap like Google Maps or a national planning map, shown at such zoom that the entirety of a continental superstate like the United States is in the field of view, will only include the highest level of the transportation hierarchy. In the case of roads, that’s the Interstates, and the map may well omit spurs and loops. At lower altitude, more roads are visible, until eventually at city scale all streets are depicted.

The same is true of public transit – and high-speed rail is ideally planned as public transit at intercity scale. A continental-scale proposal will depict high-speed rail because it depicts all cities at once and therefore what matters at this level is how to get between regions. A state map or regional map such as for New England will depict all regional connections, and a local map will depict bus connections around each train station. At no point are these in competition for resources – good integrated planning means they all work together, so that improvements in regional rail also enable better bus connections, and improvements in intercity rail enable better regional connections.

Is all of this absolutely necessary? No. France manages to make certain connections work without it, and when I try to model this as a door-to-door trip, it’s a factor of 1.5-2 question, not an order of magnitude question. But a factor of 1.5 question is still serious, and it’s one that resolves itself with good public transit planning, rather than with not building high-speed rail at all.


  1. Benjamin Turon

    Great response! I never understood why some critics of High Speed Rail think people won’t park n’ ride at the train station and then utilize taxis, rental cars, or carpool with a friend/colleague/family member at the destination as they do for air travel. I know that from my readings of British books on their Intercity services that park n’ ride suburban stations were seen by the railway as being particular important in attracting ridership, and as being an advantage over air travel in that Intercity served both city centers and suburban stations.
    This is why I have advocated for an Albany-Colonie Station on the former brownfield site of NL Industries as a great location for a new Amtrak station. It’s a short drive from the interchange of the I-90 and I-87, served by CDTA express buses on Central Ave, and is close to SUNY Albany. Albany-Rensselaer across the Hudson River is the 9th busiest station in the Amtrak system in large part because as an Upstater you would have to be crazy to drive your car into New York City. Much better to take Amtrak and then use the subway, taxi, or your two feet.
    Take a train or plane into LA, and you have the same issues of getting to your final destination, except that an intercity train can serve several stations in an urban region, so one might get you closer to your destination than the other. Some large cities do have multiple airports, but I think the advantage here is still to a rail system with a tiered service and frequent stations, like Amtrak’s NEC or Japan’s Shinkansen. Boston has South Station, Black Bay, and Route 128. California’s system will have five stations in the LA region, at Palmdale, Burbank, Union Station, Norwalk, and Anaheim.
    The fact is that a lot of the big cities in the United States which have been touted as part of a city pair do have some level of decent public transit and dense CBD, and a high speed rail station in the CBD will only reinforce ridership of connecting transit and property development around the station, just as TODs have been planned and develop around subway and rail stations in for several decades in North America.

    • Benjamin Turon

      I’m curious why the Texas Central Railway doesn’t have suburban park n’ ride stations in Houston and Dallas, a “South Dallas I-20” station for example. Brightline has/will have a lot of stations making a convenient for travel across Florida. I wouldn’t be surprise if Texas Central doesn’t want to encourage property speculation, and that beltway stations indeed might be built later, as with the new stations on the Taiwan Shinkansen.

      • Herbert

        To my knowledge they simply do not own the real estate…

        Brightline is in part a real estate speculation business… But to speculate on real estate, you need to own it (or do weird financial wizardry with it…)

      • Phake Nick

        All three planned stations for Texas Central are already located near highway interchange to facilitate easy transfer from/to automobiles.
        Additional stations would cost money to build, lengthen journey time, and operation cost, which isn’t worth when you can just direct people into your main station

        • Eric2

          Money to build – should be minimal for 2 platforms, an underpass, and a parking lot.
          Journey time and operation cost – stop penalty should be a couple minutes maybe, negligible compared to ridership gain.

          • Henry Miller


            The time lost to stopping and starting is 2 minutes (116 seconds – close enough to 2 minutes), plus whatever the stop time is. That stop time can be anything from 45 seconds to 20 minutes, but 2 minutes is reasonable to allow people to get off and new people to get on, without being excessive for people still on the train. Thus each stop costs 4 minutes for someone riding through (assuming we don’t have so many that we can’t actually reach top speed). That is the easy part.

            Cost is tricky. If the station is just a platform – no rain shelter that is vastly cheaper than a NYC sculpture that a train stops in. Those are both extremes though, I’m going to assume you want a rain shelter the whole length of the train, and probably some rest rooms, and some ticket booths (or machines?), which will put us at 300-500k. A parking lot can be dirt (free), concrete, or something in between – at different costs. Something reasonable is in the 100k range for the pavement. I’m assuming the underpass is to let the train go over this station, but on the other hand we are at a freeway exit so the underpass is needed anyway to let cars through, so overall we are looking at just making the overpass perhaps 10 meters longer, so about $400k is a reachable price. We also need to buy a bit more land to put it all one (this is Texas, so land isn’t that expensive and we can go to the next freeway exit if someone is unreasonable) $150k for a couple acres of land. Overall it is reasonable to figure a million for each station, but if they are not careful it can easily get to half a billion each.

            Of course I assume you will place the stations near a town with some small population (100k). The town might be willing to give you some “prime downtown land” (this will have some sort of empty building on it – don’t read too much into the word prime) along the route – it is a great investment for the town as the surrounding property will go up far more than the land costs. Chamber of commerce often does do these sorts of deals in small towns (meaning the tax payer isn’t on the hook – an important point in Texas). You probably can’t get this deal with a big city, but enough small towns along the way are willing to make this sort of deal that you should be looking at potentially useful places for a station to see if any deals can happen.

            You can take the above numbers and play with them. There are a lot of choices, but considering the short distance I’m not sure if it makes sense to add more stops – every time you add 4 minutes the alternatives look a little better (mostly the car), which the the trade off.

          • Phake Nick

            JR Central’s bare-bone Chuo Shinkansen station (elevated) is expected to cost 46 Billion Yen or 460 Million USD each, with no shops, no restaurants, no waiting area, no ticketing office. It being maglev would make the cost higher than regular high speed rail station, but should still be within the same order of magnitude

          • Alon Levy

            S-Bahn infill is 10 million euros, or 20 in really difficult areas. MBTA infill starts at $20 million and can go up to $45 in extremely constrained areas like Newton, and the stations are 240 m long, vs. 400 for HSR, so that’s not what makes the difference.

          • Eric2

            Marco Chitti recently posted something about infill Italian stations costing under ~1 million IIRC. Can’t find the twitter link offhand.

          • Henry Miller

            How much of that station cost is track that needs to be built anyway? If the cost of tracks is included in the station, at grade sbahn should be about 6 million for tracks at Spanish prices (if it understand their costs ), and HSR would be even more. I intentionally didn’t count that though because we are building the track either way. I did include some track costs, but only because I assume we would be at grade most of the way except the overpass to get over the road which I lengthened a bit for station use (this might be a bad assumption, but it seems reasonable )

            Also, this is a Texas station, they only need walls for privacy around the toilets, and a roof for rain /sun protection. On cold days people will need to bring their own coat and deal with freezing – if there is even power to run the train, as this winter showed there probably won’t be so even more reason not to bother with a heatable shelter. Add a million to that price if you want something that can be heated and air conditioned (and even more if you want to install HVAC )

            Even still, I’m sure my estimate is very low. I still stand by it, construction has plenty of opportunity to blow up costs for what us ultimately cosmetics. You will have no problem adding another zero on and getting people who don’t have to pay for it to support the idea. The key here is I’m thinking as if I was personally going to have to pay back the loan and that means I’ll cut corners where it isn’t pound foolish.

      • Oreg

        I think I remember Alon arguing that P&R inhibits TOD and the sprouting of a local transit network around the station. That argument might be weaker in Texas than in Europe…

        • Benjamin Turon

          You can do park n’ ride and TOD at the same station depending on the land available, on outlying greenfield site you could build a big parking garage/surface parking lot on one side and new town on the other side. Look at the office parks, shopping centers, hotels, and apartments built around the Route 128, Metropark, and New Carrollton stations on Amtrak’s NEC. You can also go with building a high-rise tower or two or three to add density. And you can adopt a “New Urbanist” design so that you can walk to nearby destinations. Yes, building a station costs money but the combination of new ridership and real estate could make it worth while. The Texas Central would cross the I-20 very close to its interchange with the I-45, there is still a lot of undeveloped land around that area. My guess is if Texas Central has such plans, they keep them close to the vest.

          • Oreg

            Right, but if you combine P&R and TOD the way you describe it, you still end up with only half the TOD you could have without the parking. Yet in Texas that might be a good trade-off, given how hard it is to build decent transit for sprawl. Presumably, the P&R would vastly increase the attractiveness of the train in the eyes of such a car-bound society.

          • adirondacker12800

            Last I heard the waiting list for parking at Metropark was two years. Metropark does have dedicated ramps from the Garden State Parkway. ‘I’m sure it all worked out fabulously more than 40 years ago when they consolidated Colonia and Iselin at Metropark. Not so much today. Jersey Avenue opened before it, Wikipedia says 1963, to keep the parking out of downtown New Brunswick. They opened Hamilton in 1999 to keep the parking out of Trenton and the edge of downtown. The people around Hamilton complain about the parking desert. And the traffic. They complain about the traffic and parking desert in Ronkonkama too.

    • Richard Gadsden

      In the UK, we conventionally make a distinction between a park n’ ride, which is a suburban station (rail or bus) with a big car park, and a parkway station, which is a long distance station built outside a city centre with good road connections and a massive car park.

      It is, honestly, not a distinction that really needs to be made – the main practical difference is that park n’ ride car parks don’t normally allow overnight parking (or charge enormous fees for staying overnight) and parkway ones always do – but it is confusing to hear “park n’ ride” for what I’d call a “parkway”.

      Parkway stations are named after the first, Bristol Parkway – which was, in turn, named after the motorway that it is built alongside.

      • Matthew Hutton

        The list of UK parkway stations:
        * Aylesbury Vale Parkway – 200k passengers a year – urban area 72k – no close major roads
        * Bodmin Parkway – 240k passengers – no nearby settlement – close to A30
        * Bristol Parkway – 2.4 million passengers – urban area 670k – close to M32/M4 junction
        * Buckshaw Parkway – 455k passengers – urban area 100k – close to M6/M61
        * Coleshill Parkway – 265k passengers – village ~10k but close to Birmingham – close to M42/M6/M6 toll
        * Didcot Parkway – 3.34m passengers – urban area 27k – fairly close to A34
        * East Midlands Parkway – 340k passengers – no nearby settlement – on A453 and close to M1
        * Ebbw Vale Parkway – 44k passengers – village 20k people – fairly close to A465
        * Haddenham and Thame Parkway – 930k passengers – village 4.5k – close to A418, fairly close to M40/A40
        * Horwich Parkway – 670k passengers – village 20k people – close to M61
        * Liverpool South Parkway – 2.66m passengers – urban area 860k – close to A562, fairly close to M62
        * Luton Airport Parkway – 4.20m passengers – urban area 250k people, close to M1 and London Luton Airport
        * Oxford Parkway – 1.02m passengers – urban area 150k people, close to A34
        * Port Talbot Parkway – 545k passengers – town 37k people, close to M4
        * Southampton Airport Parkway – 1.59m passengers – town 24k people, close to M27 and Southampton Airport
        * Stratford-upon-Avon Parkway – 110k passengers – town 27k people, close to A46
        * Tiverton Parkway – 510k passengers, no nearby settlements, close to M5
        * Warwick Parkway – 610k passengers, no nearby settlements, close to A46 and M40
        * Worcestershire Parkway – unknown number of passengers, no nearby settlements, close to M5

        So all in all other than Aylesbury Vale Parkway and arguably Liverpool South parkway and Stratford-upon-Avon parkway they are either close to motorways or expressways.

        In terms of transit oriented development, perhaps Aylesbury Vale Parkway and Liverpool South Parkway could have some, but I suspect for the others the existing places would be overwhelmed or you’d be building a massive amount of new housing close to a noisy motorway.

        • Matthew Hutton

          The most successful parkways that aren’t in a big city or by an airport? Oxford, Haddenham and Thame and Didcot. All in the Thames Valley around 40-60 minutes from London. All with decent road access but other than Oxford not on the motorway/expressway network.

        • Rational Plan.

          Not all parkways are equal. Some are in existing suburban areas or have been built in areas of suburban growth. A lot have been built because of road congestion in the centre of towns and cities where existing express services stop and their is limited parking available. Not something that is a problem in most US cities.

          Alyesbury Vale has been built in an area where new suburbs are being built and is really just a commuter railway station so people can avoid driving to the town centre station.

          Didcot is an old station in an old railway town on a major junction of the mainline. ( so had lots of old industrial land for parking). It’s just become a centre of massive suburban expansion for Oxford.

          Though many towns in the UK have good rail service most have limited parking. People often drive to major towns to catch trains to towns or suburbs with high frequency/express service. It this ‘rail heading’ that lead to the growth of parkway stations.

          Bristol Parkways the most successful. What with already being on the Great Western mainline, just where the line splits from London towards either Wales or the South West. That with being next to the M4 and M5 means good access and this meant that North Bristol became the focus of suburban sprawl, And is now home to many large corporations and government departments, university campuses and hospitals.

    • Lee Ratner

      I think many HSR fans and advocates don’t want HSR park n’rides because they want HSR to be a fundamentally anti-car system. So they want people to take transit and/or walk to HSR stations and take transit/walk to their destination or event.

      • Herbert

        a parking lot cannot possibly get large enough to generate the ridership a proper urban HSR stop gets…

        • adirondacker12800

          Click to access NEWJERSEY19.pdf

          There are buses in Woodbridge where Metropark is. The ones in Iselin, the neighborhood Metropark is in, are small and infrequent. There are far less parking spaces than commuters who use NJTransit too. I don’t know what is going on with that. They are figuring things out. Trenton is someplace between all of it. The River Line is a one zone bus fare all the way to Camden. Trenton’s local buses don’t suck. They aren’t great but they don’t suck. Masochists take SEPTA there. There’s parking but they built a parking desert next to the Interstate in Hamilton. The only way to get to Newark Airport station, unless you are are vigorous hiker, is Airtrain. Parking in Downtown Newark is expensive. Newark has buses that don’t suck.

          There is a parking desert in Rennsalaer. With the cheaper lot farther away.

          Click to access NEWYORK19.pdf

          With a delightfully uncongested bridge to I-787 in Albany built in anticipation of I-590 or whatever it was going to be. That will never get built. People whined enough and the bus to the station is less sucky than it used to be.
          If I’m remembering correctly a bit less than French levels of ridership, for more or less hourly slow trains, and about half of Japanese. With a bus that sucks.
          ,,,, I can speculate what is going with Hudson. There isn’t much parking there. I doubt very many of them are taking the bus.

      • Phake Nick

        Most intermediate station of JR Central’s Chuo Shinkansen in Japan are located near Expressway interchanges rather than local train stations for some reasons.
        And I believe Kyushu Shinkansen attracted quite a number of park and ride users from Eastern Kyushu?

        • borners

          It would be western Kyushu, and only with some stations like Shin-Tosu. And not just the Kyushu Shinkansen, the Tohoku (Nasuoshiobara) and Joetsu/Hokuriku Shinkansen (Honjo/Karuizawa) have stations outside main towns that have lots of parking.
          Curiously park and ride in Japan is called Paaku ando raido….and is seen as some new thing which is why they nabbed the English word.

          • Phake Nick

            I mean passengers as far out of the way from the current West Kyushu Shinkansen route, located on the other coast of the island, are still willing to drive some quite a distance to reach their nearest Shinkansen station on the other side and change for the train

            Looking at data the model share of Kansai to Miyazaki is 4.2% on train versus 95.8% aircraft, despite using the Shinkansen on such route would require 2 hours of driving or bus connection in addition to 3.5 hours high speed train, or 4 hours of high speed rail plus 2+ hour of local train

    • adirondacker12800

      Colonie? The village or the town? The eastern border between the village and the town is Wolf Road. The border between the town and Albany or Guilderland is the tracks more or less. I’m going to assume you aren’t fantasizing about building different tracks. State employees are convinced they are woefully underpaid. And have flex time. A CDTA station at Fuller Road and shuttle buses to the offices might attract enough riders for three car rush hour only service. Being woefully underpaid the ones who work in downtown Albany might even put up with having the high frequency buses extended to Rensselaer. It could be three cars every half hour! With a Saratoga County express once or twice at peak. Nah, there wouldn’t be enough stations for an express.
      They went and tore out almost all of the railroad on the Albany side of the river for I-787. It would be powerfully expensive to build platforms hovering over it. Fantasize about it for a moment if you want to but it is very unlikely to ever happen.

  2. Telso

    Fyi, SNCF is adopting the Thalys rule of being able to deny boarding to anyone not on the platform two minutes before departure. They are enforcing this with faregates that they are building at more and more stations, and this causes all kinds of needed exceptions for platform access for group tickets, children, aides for less abled people, etc. I used them two years ago in Rennes and they worked, but seem like an unnecessary waste, especially as they may still do ticket checks on board, though maybe that’ll stop? After all, a few days later I took a OuiGo from Lyon airport and they had a station attendant check tickets before going down to the platform and there was no conductor seen for the whole two hour trip to Paris.

      • Phake Nick

        China requires 15 minutes befkre boarding iirc, and in Hong Kong they require 45 minutes before boarding for border and custom control

        • Richard Gadsden

          Eurostar is 30 minutes for border, customs control and security. 10 minutes if you pay for first-class. They couldn’t actually process everyone that fast; there aren’t enough lanes through the system, but they can operate a priority lane that quickly. Certain disabilities entitle you to the (same) priority lane without having to pay first-class prices (blind passengers and wheelchair users are the ones I know about; I expect there’s an official list somewhere).

          The security checks are similar to pre-9/11 European airport checks (ie metal detector gateway and luggage X-ray, no millimetre body scanner, no clothes or shoes removal).

          • Phake Nick

            The border and custom procedure itself in Hong Kong is easier because local residents and Chinese residents can use electronic border crossing which only take like ten second or so and there are dozens of lanes, but then there are also airport style security check with X-ray scanners, and then passengers need to present their ticket before entering the safety check arena, where sometimes your luggage will also be checked for size and weight to ensure they’re within the limit, and the ticket itself would stop getting printed 45 minutes before the train depart. Even if you have purchased the ticket digitally you need to hit the counter and exchange for a paper ticket before you can use it to board the train. And then before entering the train station you also need to present train ticket purchase permit to prevent non-welcomed individuals like protesters from entering the train station, and then if you entered the train station quickly and cleared the border and custom check quickly need to spend some time, the waiting hall for the high speed train have zero amenity because they cannot figure out how to deal with things like taxation, currency, and safety law as Chinese law are being applied inside that area instead of Hong Kong domestic law with exception but not specified enough. Then, in line with the Chinese railway network requirement, sometimes trains departing fro Beijing will require a second security check before passengers board the trains, when important political meetings like NPC or CPPCC are being carried out in Beijing, or in other times that are of political sensitivity.

          • yuuka

            The bit about electronic border crossing is a side effect of Lo Wu and cannot be replicated anywhere else, maybe apart from the Malaysia-Singapore border.

            But ultimately, it would be fair to assume that no other HSR operator will ever have or need the paranoia of CRH.

          • Phake Nick

            Hong Kong also have electronic border crossing arrangement with a number of other countries, that I wouldn’t say it is exclusive to non-sovereign borders

  3. michaelrjames

    Many good arguments.
    But I reckon one of the potential biggest effects of HSR in the US is on sentiment towards transit in general. It’s because of the particular type of user who will use HSR and inevitably be converted to its usefulness, and rapidly become dependent on its conveniences and comforts. You don’t really want to fly again if you can take a HSR instead, which also explains why the maximum travel time on HSR deemed acceptable re flying, continues to lengthen. Noah himself proves that in his spiel. And not just time per se–that is a serious under-rating of the HSR experience–but uninterrupted quality time which is in great contra-distinction to air travel. Essentially, even the cheapest seats on HSR are better than any premium seat on a plane–especially as more time is actually spent getting to, and getting processed thru, the airport in a whole series of staccato irritations, outside of your control, with the delightful lottery of scoring a full-body search (or at least a thorough search of luggage and your person). And no (for the naive Americans), that doesn’t happen on HSR, even on international journeys, even with the obsessive Brits on Eurostar; IIRC you were still allowed to begin passing thru customs etc with as little as ten minutes before departure.

    The point is that all kinds of people that wouldn’t dream of using city transit will be using HSR, and before long, loving it. For many Americans, against their lifelong concept and hatred of the very idea.

    I believe this would have a very significant effect on their attitudes to city public transit, or at least their willingness to either try it, or to have their taxes fund it, or at least not to fight it as if it were an affront to American exceptionalism, personal freedoms and God’s way …

    • Eric2

      “You don’t really want to fly again if you can take a HSR instead”

      Only if the HSR trip time is reasonable.

      You can take HSR from Amsterdam to Granada with only a couple short non-HSR segments on the route (which will become HSR in a few years anyway). Travel time is over 10 hours, I’m willing to bet the mode share for rail is extremely low.

      • Herbert

        Part of the issue with pan-European hsr trips is that while booking a flight is damned easy (now, booking the cheapest flight is a different ball o’ wax altogether…) booking a trip from – as you imply – Amsterdam to Granada involves several different tickets (with no guarantee of carriage if connections misalign) several different conditions attached to the tickets and almost certainly schlepping from one station in Paris to another…

        Ironically at the height of the Cold War, it was easier to buy such a ticket – the TCV would allow a station attendant at Goslar to sell you a through ticket to Marakesh (maybe you’d have to add on a separate ferry ticket, but often that was included, too) or Beijing and have it honored all along the way…

        Now if there were a single train (or maybe one train with a reasonable connection at e.g. Madrid) that did the 10 hour trip in one night, of course there would be demand… But alas in the age of “privatized” state railroads everyone jealously guards their turf and the only international trains are those allowed by bilateral agreement….

        • Phake Nick

          First problem is maintenance window, and utilization rate of a train designed for night train service
          Second is such aervice wull still cater to relatively less percent of travellers, making it hard to accommodate them while sacrificing daytime service level
          Third is different electrical system, signal system, and loading gauge across different countries
          Fourth is operarional cost of overnight high speed train against airplanes

          • Herbert

            The EU is supposed to make most of those issues disappear. As for loading gauge: Virtually all of continental Europe fits inside “Berne gauge” and that is enough for modern HSR…

            Now the fundamental issues of night trains exist regardless of speed and ÖBB just ordered a new batch of them (iirc with 230 km/h capability making them on par with the fastest daytime trains owned by ÖBB) so certainly some people in the industry think they can make it pencil out…

          • df1982

            None of those are issues.
            1. ON trains only need to occupy a single track since there is only one or two a night, so maintenance can be done on the other track. 8-10 hours a night is not a bad level of utilisation. Probably the best procedure is to have loco-hauled sleeper carriages, then the locomotive can be used for other duties during the day while the carriages sit in a depot (for which there will always space during the day since way more trains are on tracks in the day rather than overnight).
            2. Why would you be sacrificing day time service levels? You only need as many passengers as will fill the train to make it viable. But you’re providing a service that other transport options don’t cover well (you get a night’s sleep then wake up in the morning in another city, earlier than you could on a morning flight or HSR train without waking up at an ungodly hour).
            3. This problem has already been solved for other cross-border services like Eurostar, Thalys, etc. At worst you have bimodal locomotives and run them as diesel across incompatible electrical services.
            4. Operational costs may be somewhat higher than a plane, but tickets can also be more expensive, since for a lot of people a night train means saving the cost of a night in hotel, so you still come out ahead. Austrian Railways don’t seem to be having any trouble making their routes work. What Europe really needs though is more point-to-point night trains covering city-pairs like Berlin-Paris, Paris-Madrid, Berlin-Zurich etc non-stop. London to the south of France would also be a popular market if they can get the border issue sorted out.

          • Alon Levy

            In Europe, it’s not normal for work to be done while trains might run on an adjacent track. In the US it’s very normal, but there are special rules for this that slow down both the train and the work.

          • Herbert

            It’s always easier to just switch locos than to diesel under wire, especially if the only reason you do it is incompatible electric systems…

          • Phake Nick

            >ON trains only need to occupy a single track since there is only one or two a night, so maintenance can be done on the other track.
            Even for high speed operation? As like China only allow such one track maintenance, another track in operation procedure to traffic less than 200km/h for safety reason iirc

            >Probably the best procedure is to have loco-hauled sleeper carriages, then the locomotive can be used for other duties during the day while the carriages sit in a depot (for which there will always space during the day since way more trains are on tracks in the day rather than overnight).
            For multiple reasons it seems like there are more and more EMU instead of loco-hauled high speed trains around the world..

            >Why would you be sacrificing day time service levels? You only need as many passengers as will fill the train to make it viable. But you’re providing a service that other transport options don’t cover well (you get a night’s sleep then wake up in the morning in another city, earlier than you could on a morning flight or HSR train without waking up at an ungodly hour).
            I mean there is a sort of perception from like New York Metro where time can be spent on cleaning and maintenance seems to be sacrificed due to around the clock operation. But it could be simply because of New York Metro have not enough budget to do proper cleaning and maintenance work in its stations and trains instead.

            >This problem has already been solved for other cross-border services like Eurostar, Thalys, etc. At worst you have bimodal locomotives and run them as diesel across incompatible electrical services.
            IIRC it was said to need quite a lot of different system when one want to run a train from like Amsterdam to Rome through Switzerland, probably not impossible but still added complexity

            >Operational costs may be somewhat higher than a plane, but tickets can also be more expensive, since for a lot of people a night train means saving the cost of a night in hotel, so you still come out ahead. Austrian Railways don’t seem to be having any trouble making their routes work. What Europe really needs though is more point-to-point night trains covering city-pairs like Berlin-Paris, Paris-Madrid, Berlin-Zurich etc non-stop. London to the south of France would also be a popular market if they can get the border issue sorted out.
            Humm right, but isn’t there also competition with like overnight buses?

          • Herbert

            The only people who take overnight buses are extremely price sensitive people or aviophobics with no other choice. There simply is no good sleep to be had in a bus seat…

          • Phake Nick

            I guess it is an exception that Japanese overnight bus company can run overnight buses with low density seating or even business class service level that can cost more than flights and still attracted customers

          • wiesmann

            The electrical system issue is disappearing, multi-current locomotives are now the norm for freight in Europe, with modern electronics this is way less of an issue, and countries like France are (slowly) phasing out the low voltage DC system. Loading gauge is an issue, but there are gauge switching bogies, and, all in all, there are only a few outliers (Russia and Spain). Signalling issues are slowly going away with the adoption of ETCS.

          • Sarapen

            Fancy overnight buses that cater to more than low income people are also quite common in Latin America.

          • Herbert

            Loading gauge and gauge are different things…

            At any rate, Spain’s high speed rail system is standard gauge

          • Phake Nick

            Loading gauge roughly mean size of the trains allowed. There is now some common European standard according to my understanding but it could still cause problem when you enter like UK

          • Herbert

            HS1 is built and HS2 is planned to be built to a “continental” loading gauge…

        • Eric2

          Airlines used to have similar problems. Codesharing helped solve the problem, so did travel sites let you compare different airline routings side by side. I’m not really sure how much these factors currently exist, or could exist, for rail.

          • Phake Nick

            Problem is, due to HSR optimal travel distabce, and the focus of traffic from/to major cities, only very few passengers need to connect toward other train companies trains. This make the benefit of facilitating codesharing across rail companies minimal, and could cause the cost of doing so excessing the benefits.

            In fact, in Japan, the privatization of railway havs caused the national railway network be split up into multiple smaller companies, and while they still inherit and use the same ticketing system, those new companies are consistently trying to obsoletizing area of ticketing and operations that require cooperation of multiple JR passenger rail companies, with like JR Central stating that they want to detàch their upcoming new Chuo Shinkansen from the national railway ticketing system as much as possible

      • Car(e)-Free LA

        This question brings up a very interesting idea: what is the ideal coverage of a rail operator? I’d love to hear Alon’s take on this.

        On one extreme, you have Amtrak with everything from the Surfliner to the Downeaster operated as part of the same network, even as zero trips will ever use both lines. On the other end, you have a half billion different rail franchises in England alone. In Continental Europe, there is roughly one operator per country. It’s my instinct that Amtrak’s national network should be essentially dissolved as high speed and regional services develop while intercity lines around the country are vertically integrated with local, regional, and high speed rail. I can’t see any reason for Amtrak California to exist instead of being folded into a CalTrans operated portfolio of CalTrain, Metrolink, CAHSR, etc. How to organize services in the Northeast–especially the Tri-State area–is a particularly interesting and complicated question.

        In Europe, it seems to me that there are too many operators, especially as high speed rail networks integrate across national boundaries. Forcing a change from ICE to Thalys to TGV just seems unhelpful. In my view, broader successor to EuroCity needs to take charge of premium high speed services in France, the Benelux, Switzerland, Austria, Germany, Poland, Czechia, Slovakia, Hungary, Italy, and Denmark. There are just too many city pairs that cross too many national borders for the existing setup to work. On the other hand, including Spain and the United Kingdom seems foolish–in Spain’s case because there are basically no viable city pairs past Barcelona–except a single Paris-Madrid service, and in the UK’s case because of Schengen, HS2’s incredibly intricate timetable, and the need to flow all services through London anyway. This, of course, opens a number of questions on how regional and conventional intercity services should be organized.

        • wiesmann

          My understanding of the European law is that eventually, any operator can run trains on any link. For some countries like Switzerland this is already the case, on the tracks in Zürich you can see French-Swiss (Lyria) trains, german ICEs, Austrian Night-Jets and Italien trains…

          • Car(e)-Free LA

            Right. The issue is that this is really annoying for nice timetables, flowing past the first main city in a country, and giving a single operator a monopoly on a corridor–which they should really have for the sake of passenger comprehension. Consider Paris-Germany services. Paris-Frankfurt is split between ICE and TGV which is just unfriendly for passengers. Moreover, the TGVs don’t flow past Frankfurt to places like Berlin, even though that would make operations and passenger ease better, because the separate operators on Frankfurt-Berlin would just make system usability weird. At the very least, it would seem to me that Eurostar, Thalys, ICE, and TGV should all be one system.

          • wiesmann

            Switzerland has many public transport operators, yet there is an integrated time-tables.
            Still, I agree that getting this level of coordination at the European level will be interesting…

          • michaelrjames


            Isn’t that a bit of an exaggeration? Sorry, I guess you live there and should know but … In practice the vast majority of the mainline services are on the state-owned tracks run by SBB–about 70%? The second biggest operator has about 10% and the third under 4%. The dozens of other companies are running small private non-mainline non-standard gauge lines into resorts etc. Is it really a model for how to operate mainline trans-European networks?

            Isn’t Lyria the more realistic model, ie. where the two national operators (French, Swiss) form joint ventures for particular routes concerning those countries?

          • Oreg

            I really don’t see the problems you’re describing in practice. As a passenger it makes no difference to me who runs the train. I can book any French, German, Swiss or Italian train on any operator’s website. Deutsche Bahn and SNCF will both show me all connections, whether ICE or TGV. There are trains running from Berlin to Interlaken, from Hamburg to Chur, from Basel to Amsterdam, from Zurich to Vienna.

          • Richard Mlynarik

            Right. The issue is that this is really annoying for nice timetables

            Wrong. The infastructure owner can sell slots — “holes in a “nice timetable”.
            If your company wants to run a freight train or something at those times and you can meet the safety/speed/length/performance/etc criteria and can accept the price/guarantees/priority/etc you’re set. If you not, you aren’t.

          • adirondacker12800

            If the timetables aren’t “nice” it might mean consulting one or checking a smartphone.

          • Richard Mlynarik

            consulting one or checking a smartphone.

            I just checked Google Transit (I did have to upload my address book, passwords and browser history for convenience, but it so was worth it! Plus the “app” is so much more everything than the old-fashioned non-reactive “web” version) and it informs me the Clue Train has left the station.

          • adirondacker12800

            When I learned how to read the women in my family had already taught me how there are signs on the bus, that I could read, to determine if it was going where I wanted to go. I used those skills later in life to not have an emotional crisis when I found that I had just missed the South Amboy local and would have to take the Trenton Express. That wasn’t on the same track it was the last time I missed the South Amboy local. Back in the olden days when we didn’t have a smartphone to welded to our hand and I had a wad of change for the exact change bus and tokens for the subway.
            And the train to New York was on Track 2 instead of Track 1. And very occasionally 3. It’s amazing we all didn’t curl up and die.

          • Phake Nick

            I have learned form my local bus operation that unified schedule across multiple operators is too inflexible and can create many puzzling problems when any changes could be desirable, as any increase or decrease or shift in schedule would require careful consideration that neither side would potentially benefit more or getting their revenue hurt by any potential change

        • Oreg

          There are many single-seat cross-border services in Europe already, right? HSR and otherwise.

          > In Continental Europe, there is roughly one operator per country.

          Switzerland has many different regional operators away from the main line. Germany has competition for provisioning regional rail services. Italy even has competing HSR operators. And that’s just passenger rail.

          • Herbert

            The “competition” on regional rail is “who gets the contract”….

            And the Italo has been hemorrhaging money for most of its existence…

          • df1982

            Italy has the worst possible model. Instead of a train every 15 min on routes like Rome-Milan, it has two companies offering rival 30min services, making it far less convenient for travellers (particularly those wanting to make connections). A train route is a natural monopoly and should be served by a single company (whether this is public or private is a different issue). At the very least SNCF and DB (not to mention Eurostar/Thalys) should have shared ticketing for passengers and work out revenue sharing between them, since they should be playing complementary rather than competing roles. But their management probably doesn’t see it that way,

          • wiesmann

            @michaelrjames The integrated time-table is not only about trains, but also medium range (for Switzerland) buses, like the so called ‘postal buses’. The fact that local trains run on different gauge does not change the need for coordination. And in terms of coordination, all these connections increase the value of the main train network. Ditto for integrated fares.

            That said, I agree that joint ventures are a solution, in fact there are many such joint ventures even for local trains, public transportation in Zürich is a joint venture of 17 operators, which offer fare integration and coordinated time-tables. InterCity lines with Italy are a joint venture between SBB and Trenitalia, etc…

    • Richard Gadsden

      Eurostar, the worst security check that I’m aware of (Hong Kong to the PRC mainland may be worse, nowhere else is even close) is very similar to pre-9/11 European airline checks. Luggage through an X-ray machine, people through a metal detector gate.

      Your shoes stay on, you don’t have a full body scan. The rare cases where you can’t clear the metal detector gate however much you take out of your pockets (quite common if you have significant metal in medical implants) are a physical search, but that’s pretty cursory and no need to remove clothes. They’re really only looking for firearms or bombs; not that they won’t stop you for drugs but the searches are not really intended to find them.

      • Richard Gadsden

        Oh and: it’s normally 30 minutes before departure for Eurostar. 60 for skiing trips because they have to hand search skis (they don’t fit in the X-ray scanners!). The ten minutes is only for first-class (Business Premier, the top-level travel class) passengers, people with the top-level status on the frequent traveller scheme (Carte Blanche) and some disabled passengers.

        • michaelrjames

          @Richard Gadsden

          Yes, but you’re talking about recommended check-in times ‘guaranteed’ to get to the train on time. I’m talking about what you can get away with. I tend to leave it on the late side because, counter-intuitively, the lines are less. Also, I hardly ever travel at peak times–in fact I wish Eurostar would run later trains.
          Anyway, I was agreeing that the times and process are amazingly reasonable–even at the British end–for a post-9/11 world.

    • Bobson Dugnutt

      @michaelrjames, the good news is that American attitudes toward city transit are broadly supportive of it. APTA has taken on the role of the now-dissolved Center for Transit Excellence’s ballot tracker, and noting which regions put a transit spending measure before voters. The passage rate is in the low 70%s. There’s a 7 in 10 chance that a measure passes. It would be higher were it not for places like California where a tax increase requires a 2-1 margin to pass. San Diego and Sacramento had ballot measures that failed with like 62% support. And some of the figures are high because transit spending is part of an omnibus spending package on roads and highways, too.

      Even better, the consensus is broad for more transit spending. The 7 in 10 passage rate holds true in both “blue” and “red” areas, and the consensus is durable across all ages, races/ethnicities, educations and income levels. At the national level, polling — which is not binding as a vote and set policy — suggests Americans are all aboard with Amtrak Joe and his infrastructure proposal. Even Republican voters. That Mitch McConnell and the Qanon cretins in the House stand athwart this consensus will start to make rational sense when you realize the point is not to make rational sense. The Democrats are trying to play 4-dimensional chess with a party who is playing 6-dimensional go fish and still checkmates them every time.

      • michaelrjames

        @Bobson Dugnutt “The Democrats are trying to play 4-dimensional chess with a party who is playing 6-dimensional go fish and still checkmates them every time.”

        Dems always bring a knife to a gun fight.
        However, one refreshing thing so far about Biden is that he is ignoring the GOP in both HoR and Senate. Of course he still has to strategize on how to get stuff into law in the Senate but at least he isn’t blathering the bipartisanship b.s. anymore. There’s even a glimmer of hope of defeating the fillibuster (the only reason Manchin won’t abide it is because it would make his single vote somewhat redundant, or at least at risk of irrelevance). It’s a long-delayed benefit from the endless b.s. time wasting of the Obama first-term trying and failing to reason with McConnell et al. ie. Biden actually did learn some lessons. Indeed, new tricks. Funny, it’s the old school dems who are uncomfortable with Biden and so far not much if any negativity from the progs.

    • Herbert

      There are not two categories of travel (“business” and “leisure”) but three: Business, Leisure and Visiting Friends or Relatives…

      Of those Business of course clusters near businesses and conference centers which often have adequate public transit (and business travelers are found in taxis more often than their private cars anyway)

      Leisure is a very diverse category from package tours who get picked up by their hotel (at the airport in most current cases, but there is no reason why it couldn’t be at the train station) to individual travelers who like the “adventure” of figuring out the Barcelona metro or the Wuppertal suspension railway or what have you to all sorts of other categories including the penny-pinching young interrailer…

      Now VFR are of course not homogenous either but almost by definition they’ll have whomever they visit waiting to pick them up – and that is possible just as easily at a train station as at an airport…

      • Richard Gadsden

        One form of leisure travel that HSR suits very well is the “city-break”, where you travel to a city for a short period (often a long weekend), stay in a hotel and visit the city’s tourist attractions (museums, historic buildings, theatres, etc). This generally means travelling to a city-centre hotel, which are generally clustered close to the HSR station.

  4. Herbert

    If hsr were incapable of working where there is not good transit access, “beetroot stations” would not only underperform, their ridership would be barely there. Judging from Montabaur and Limburg Süd, the contrary is the case. In fact, hsr stations in the middle of not much are so popular in Germany that there is serious discussion of adding an infill stop on the Nuremberg-Erfurt line (roughly at the site of the Ilmenau-Wolfsberg overtaking station) for regional trains similar to what is already done at Allersberg (Rothsee) on the Nuremberg Ingolstadt line which is also served by “S5” of the Nuremberg S-Bahn or the plans for a new “Merklingen” stop on the Stuttgart Ulm line for regional trains…

    All those stops are, or are planned to be, equipped with huge P&R lots and adequate if not superb bus connections…

  5. Phake Nick

    Why the like of high speed rail option, like Kagoshima – Kansai in Japan, despite only requiring less than 4 hours trip time, still lose out to aircraft in 1:3 ratio?

    • Tonami Playman

      For a lot of international tourists heading to take a peek at Sakurajima, it makes more sense to change planes at KIX than to shlep for 1hr getting to Shin-Osaka before getting on the Shinkansen to spend another 4-5hrs to reach Kagoshima. For local traffic which is the bulk of travelers, it has to be a combination of price and time(Japanese airline boarding procedures are less restrictive compared to the rest of the world). A plane ticket from KIX to Kagoshima is about ¥12,000 while a Shinkansen ticket is ¥23,000

    • Andrew in Ezo

      Airfare up to one day before departure is slightly cheaper than the regular shinkansen fare, considerably cheaper if you use a LCC, though at the penalty of having to use Kansai AP rather than the more convenient Itami AP. Travel time is about the same between the two modes, when you factor getting to/from airport to city centers for air travel, check in, security, etc. There may also be the fact that air travel is still relatively pleasant/low stress in Japan, at least compared with the U.S.

    • Onux

      The shortest possible trip Osaka-Kagoshima is 3:55, so it isn’t really “less than 4 hours” it is 4 hours (many scheduled trains are 4:15, for instance). The flight is 1:10, so the Shinkansen is 3 hours longer. According to this chart (page 15: https://www.itf-oecd.org/sites/default/files/docs/dp201326.pdf) when a train trip is 3 hours longer than a flight, it usually gets around 25% of the split with air, or right about the 1:3 that Osaka-Kagoshima actually gets.

      HSR just doesn’t do great for longer trips, the highest ridership seems to come in the 1:30-3:00 trip range. Hence all of the discussion in Alon’s previous HSR posts about adding a ceiling to his ridership formula the way he has a floor, so that Charlotte-Miami or Nashville-Miami traffic isn’t counted as using the Atlanta-Jacksonville line.

        • adirondacker12800

          Youse have to decide on something else. Almost none of you are collective or anonymous. Or the Queen of England. Ya get people confused when yer using plural pronouns for a single person.

          • adirondacker12800

            for some amorphous/anonymous otherness that may or may not be a single person.

    • anonymouse observer

      I will add three more likely reasons to Tonami Playman’s and Andrew’s comments:
      1) Location of Shin-Osaka Station (sort of outside of Osaka CBD and almost a half way to Itami Airport if you are coming from “Minami”);
      2) Relatively low service level of one-seat Shinkansen service from/to Osaka (1 to 2 trains per hour per direction), and;
      3) Amount of time needed to go through the security at the airport in Japan, which gives more time advantage to airlines even with so-so level airport access in both Osaka and Kagoshima.

      In Japan, passengers can jump onto the flight – airlines let you board the airplane as long as you check in a half hour before the departure time, at the security 20 minutes before the departure time, and at the gate 10 minutes before the departure time. If you show up at the security in the last minute (close to 20 minutes before the departure), you typically get the highest priority at the security screening.

      • Phake Nick

        Then the question is, why on other routes like Kanto to Hakodate, with worse frequency, worse station access, and better airport access, the Shinkansen still manage to capture higher modal share than Kansai to Kagoshima? (Not significantly better but still it captured one third of the market compares to one fourth on Kansai – Kagoshima)

        • archie4oz

          Because nobody actually “goes” to Hakodate (I’m being slightly facetious here… I actually *like* Hakodate). Sapporo is the real destination of choice (and one of the busiest air routes in the world). Shinkansen dominates to Hakodate simply because it goes there, and you can get to downtown Hakodate faster from Shin-Hakodate by local train than a bus or taxi (your only options) can from Hakodate Airport. Also even if it’s just a bit further, if you’re going to Hakodate it’s more likely that the journey is a part of the trip as opposed to Sapporo where it’s all about “just getting” there.

        • Matthew Hutton

          Probably also like everywhere else in the world flights are super cheap for major city pairs (e.g. Edinburgh to London) whereas they are more expensive for unusual city pairs and the flight timings may also be worse.

        • anonymouse observer

          I am not sure exactly why, but it is probably because of how “Kanto” is defined and ease of access from/to airports in Tokyo from various areas within “Kanto”.

          Area defined as “Kanto” is relatively large in Japanese standard. Regularly speaking, it includes 6 prefectures (Tokyo, Kanagawa, Chiba, Saitama, Gunma, Tochigi, and Ibaraki). Trip to Haneda Airport from the capital city of 4 out of these 6 prefectures in “Kanto” region takes more than one hour. Because these 4 prefectures are located north of Tokyo and 2 (Saitama and Tochigi) of them are on the Tohoku Shinkansen alignment, it’s much easier to just hop onto the Shinkansen to Hakodate rather than heading south by 1+ hour to Haneda Airport and fly. I believe JR East still schedule Yamabiko trains and Hayabusa trains in a way that passengers can transfer between them at Sendai only with a short wait (this covers Utsunomiya). Also, from 2 other prefectures (Gunma and Ibaraki), the Shinkansen stations (Omiya or Ueno) is on the way to Haneda Airport, and you could easily save 30 minutes by not going to Haneda if you are taking trains to Haneda.

          I guess the mode share changes somewhat significantly if “Kanto” region is separated into two groups: “Kita-Kanto” and “Minami-Kanto”.

          Location of the Haneda Airport relative to “Kanto” region could be a disadvantage for airlines. Getting from/to Haneda Airport is harder than one thinks unless the origin/destination of the trip is on Keikyu/Toei-Asakusa/Keisei Line corridor or near the train stations served by frequent “limousine bus” from/to Haneda Airport, and there are not a lot of locations outside of the Tokyo CBD with frequent “limousine bus” service. For those going from/to Haneda Airport from the location with subpar “limousine bus” service, the trip would be 2+ seats – either going to nearby node with frequent “limousine bus” service to catch the bus or going all the way to Hamamatsu-cho (for Tokyo Monorail) or big stations on Keikyu/Toei-Asakusa/Keisei Line like Takasago, Aoto, Shinagawa, Keikyu-Kawasaki, Yokohama, or Kamioooka. In most cases, going to a Keikyu/Toei-Asakusa/Keisei Line station requires another transfer or extra short walk between stations (JR stations and Keikyu stations are not co-located except Shinagawa and Yokohama) if you are from the area west of Yamanote Line. I would not be surprised if people prefer catching the Shinkansen from Tokyo, Ueno, or Omiya to Hakodate if they need to make multiple transfers to get to the airport.

        • Alon Levy

          FWIW, JR East’s factsheet has Tokyo-Hakodate at 28% Shinkansen, 72% air. JR Central is here, PDF-p. 4, and JR West is here. Shinkansen modal splits seem really awful by global standards when people travel between islands, I think? Tokyo-Hiroshima is doing way better than Tokyo-Hakodate.

          • Benjamin Turon

            Tokyo-Hiroshima as been in service for decades so its a more established market than Tokyo-Hakodate I would think, and the overall population and number of tourist attractions may be higher in and around Hiroshima than Hakodate. People are creatures of habit, and perhaps the airlines are doing a real good job with marketing. The Shinkansen station (Shin-Hakodate is a bit outside the city center of Hakodate (17km), surrounded by farm fields. The airport seems a lot closer to city center than the Shinkansen. They really need to get to Sapporo I would think for their to be a big bump in ridership.

            Shin-Hakodate-Hokuto Station

          • Tonami Playman

            To add to what everyone has said regard Hiroshima and Hakodate, Hiroshima Airport is 48km outside the city compared to the Shinkansen station which is right in the middle of the city. The Limousine bus Sanyo line combo takes 1hr 41mins at best making flying into Hiroshima a big disadvantage. The Nozomi Shinkansen takes between 3:50 and 4:01 from city center to city center.

            Hakodate on the other hand has a Shinkansen trip that takes 4:22 to reach Shin-Hakodate from Tokyo, but takes a total of 5hrs including access time to city center. The trip from Shin-Hakodate to Hakodate station takes 32mins compared to the 25mins it takes to get from Hakodate Airport to Hakodate station. Add the 1:30 flight time between Haneda and Hakodate Airport, and Shinkansen to Hakodate is heavily disadvantaged both on trip time and Airport access/Station access time. Pricing on the route is very similar for both Air or Shinkansen, so most travelers are choosing Air based on speed advantage.

          • Andrew in Ezo

            As Benjamin says, and Hiroshima is just a bigger city with much more industry (Mazda Motors is HQ’d in Hiroshima) and likely tourism, as it is prob. the second biggest domestic school trip destination after Kyoto. Hakodate is mainly a city dependent on tourism and services with little industry, and not only is it distant spatially but also psychologically in the minds of most domestic travelers (the Tsugaru strait has traditionally been both a natural and cultural barrier). Flying would be the default choice for many.

          • Phake Nick

            Then I guess the data I was reading could be not the most updated one and still include extra passengers who simply want to try out the train

  6. Phake Nick

    > This is particularly useful for cities that, by virtue of size (Dallas) or location (Cleveland) could be high-speed rail hubs but do not have strong city centers. In Cleveland, demand for housing in the city is extremely weak, to the point that houses sell for well below construction costs, and demand for city center office space is likewise weak; but a train that gets to Chicago in 2 hours and to New York in about 3:15 can make the area immediately around the station more desirable. 

    One thing people might fear is the straw effect. In many Shinkansen stations in Japan, or with the opening of new expressway, that they have shrink the traffic time so much that passengers would bypass old regional center and head directly to bigger cities for their business, commerce, and daily life needs that the role of local regional center is weakenef

    • Bobson Dugnutt

      @Phake Nick, don’t discount the possibility that Cleveland itself would become a more attractive destination by virtue of being within a high-speed rail network. Cleveland is more than just a bad house along a good block. If you’re a city, your selling point can’t be “Hey, we’re an hour or two of places much more interesting than us!” You’ll only help your dwindling population and economic base light out faster.

      There is the Rock & Roll Hall of Fame. There is professional baseball, basketball and football. If you love classical music, you’d know Cleveland has one of the best symphonies in America, if not the world. If you’re a college student, you have Cleveland State and Case Western Reserve. As a patient or health worker, there’s the Cleveland Clinic. And so on.

      Cleveland’s virtue is that while its past is better than its present or near future, when it was a young city it made investments that stood the test of time. It doesn’t just have good bones, it has King Tut’s tomb.

    • anonymouse observer

      Straw effect also get triggered by slower-speed modes (conventional rail or express bus) and in much shorter distance in Japan, like Nagoya CBD killed Gifu CBD after JR Central improved Tokaido Main Service and cities on Honshu “sucked” from cities on Shikoku through bridges over Seto Naikai in last two decades or so.

      Because straw effect could potentially trigger the centralization of the destinations and populations, it could help turning the tide of the argument of high-speed rail not suitable for North America if it is built and used wisely. I am not sure how to do this without causing damages (straw effect is dangerous and could kill a mid-size city CBD like Gifu, Yamagata, or Kofu), and this might not support Alon’s point as someone needs to create the initial straw effect which is significant enough to get the centralization going before building larger-scale system like high-speed rail.

  7. Herbert

    Isn’t it interesting how in countries where that isn’t the capital anyway rail and air hubs tend to converge? (Atlanta, Frankfurt etc.)

    • Eric2

      I would add Chicago. But maybe the reason is just that they are the biggest cities or economic centers in their regions?

      • Autolycus

        tl;dr: At least for Atlanta it’s a bit of coincidence and maybe a bit of correlation.

        As someone who is a little more of an AV geek than a train geek, and one who lives in Atlanta metro, I can say that there’s _some_ connection with the size and economic influence of the city, but there’s a fair question over which is the chicken and which is the egg–at least for Atlanta. Atlanta was NOT the biggest city or economic center of the SE until the latter parts of the 20th Century. That coincides nicely with the rise of Delta and UPS.

        The big advantage that Delta had in Atlanta wasn’t actually Atlanta itself. It was Atlanta’s location, airport layout, and weather. Atlanta is within a 2-hour flight of 80% of the US population and a whole lot of popular tourist destinations in Florida. The airport has a midfield terminal complex with taxiways to both north and south runways. It has multiple runways and importantly no significant need for any cross-wind runways. This layout facilitates a LOT of traffic and does so very efficiently for both planes and passengers. Atlanta is also less effected by weather, especially winter weather, than other markets that might otherwise have made good air hubs. Charlotte and Nashville, for example, is just a little further north and just a little more icing.

      • Herbert

        Yeah, Chicago definitely qualifies, but one could explain it away as the only major city (back in the early 20th century) not on the East or West Coast (but instead on the “north coast” of the Great Lakes)

          • Eric2

            Yeah, but St Louis and Cleveland declined long before the 1950s, so I explained why Detroit did well for a while despite the decreased importance of water transport.

          • Onux


            St Louis and Cleveland both had their population peak in the 1950 census, so I don’t think you can say they were in decline “long before” the 50’s.

    • Phake Nick

      I think that’s a sign that development are inherently uneven, places with better access to one will inherently attract others, and so on

  8. Oreg

    One could argue the exact opposite of Smith’s position: The gain in individual convenience is much greater with HSR than with local transit. HSR is more comfortable; you have more quality time, only rarely interrupted by connections; and there is a good chance that you reach your destination faster than by car, which is rarely the case for transit.

    Also, where is @adirondacker12800 when you need them to respond to the lack of local transit around HSR stations: “That’s why airports are such a failure.” 😉

    • Bobson Dugnutt

      @Oreg, arguing that HSR is better than local transit is just as wrong as arguing that spending money on HSR deprives public transit.

      A better argument: Break the frame that says HSR and transit investments must trade off and come at the expense of each other. Instead, argue for local transit improvements that need to be made and would be nice to have, and use some buy-in like a local tax to leverage state and federal support. At the same time, argue for federal support to seed construction and starting operations funds for high-speed rail.

      The flaw of the “studda argument,” like the “couldda argument” or the “shouldda argument,” commits the fallacy of the natural kind. See https://en.wikipedia.org/wiki/Natural_kind. (Studda is meant to be instead of, but the corrupted pronunciation is in service of poetic license 😉 )

      Each one of these arguments sees alternatives as essentially the same in kind, failing to appreciate the One Big Difference that has created the distinctions and interactions between them.

      In the studda argument, a new pot of money comes available. You could spend the money studying, constructing and building high-speed rail. But instead, you could take that money and buy yourself more local transit vehicles and staff their operations and maintenance. The natural kind fallacy is treating high-speed rail and local transit as essentially the same because they are both vehicles in which someone else drives you to your destination.

      The One Big Difference: Local transit’s benefits accrue locally; their benefits don’t extend beyond areas they do not serve. (i.e. New York City’s high ridership doesn’t translate to high ridership in other cities.) High-speed rail, and all other intercity infrastructure, are nodes in a network. The network effect is meant to facilitate exchange to the benefit of all participants within the network. Intercity trips are part of a network; local transit is a network.

      They are interrelated but must be planned differently. Think about local transit transfers at a rail station. There might be, say, 10,000 boardings a day at the station. For the HSR line at the station, those transfers might account for say a third of ridership on a 30,000-boarding line. That’s crucial for the HSR’s planners. The local transit system might be carrying 1,000,000 riders daily. For transit planners, that’s just 1% of overall ridership. The local transit network is producing 990,000 boardings that don’t involve HSR. Using Noah Smith’s logic, those 990,000 transit boardings outcompeted the 10,000 rail boardings and the service hours could be redeployed elsewhere throughout the system. HSR loses a third of its ridership, imperiling its business case and forcing it to reduce or eliminate services.

      However, redeploying those buses within the transit network might be a suboptimal solution. When ridership is already high, it’s very hard to grow ridership. If service is in single-digit frequencies (9 min or better), a service improvement will not be noticed by riders. Doubling 60-minute service to 30-minute service is a much more dramatic service improvement (reducing wait time by 30 minutes) than doubling 10-minute service to 5 (reducing wait time by 5 minutes). Growth opportunities come by adding service at times and places not offered (say, owl service, offering a weekday-only route Saturday and Sunday service, or extending service out to a new subdivision), yet the transit agency may have added coverage that has yielded fewer than 10,000 riders or grew ridership but dragged down productivity (you added a rider at more than the cost of serving a train transfer).

      Don’t play this game. Be for transit. Be for high-speed rail.

      • Autolycus

        This is my line of thinking as well. Supporters of HSR and transit need to stop talking about them as if they’re exclusively dependent on each other and that either one is a necessary condition for the other to succeed. That sort of discussion ultimately results in less total money for both of them. We need to advocate for doing both every chance we get. Having HSR is better than not having HSR regardless of local transit conditions. Having local transit is better than not having local transit, regardless of HSR conditions. Therefore we always need to advocate for more money for HSR any time its an option and we always need to advocate for more money for local transit any time its an option. Stop arguing that one has to happen before the other because you just allow those opposed to your 2nd one to defer spending on that one indefinitely.

  9. Autolycus

    It’s absurd for anyone to suggest that quality local transit is a necessary condition for quality HSR. We certainly don’t have any such prerequisite for investing in air infrastructure.

    As for your specific responses: I completely agree that you’re right for Atlanta-connected routes. Atlanta metro residents are dispersed widely, but there are two clear reasons that it doesn’t matter anyway:

    1) any HSR routes would include stations north of town that are FAR more convenient to a lot of the metro population than ATL is. ATL is south of the city and it can be quite a nightmare getting to the airport for residents of the northern suburbs, especially during rush-hour, but even also on weekends when roadwork is more commonly done on our highways.

    2) The VAST majority of people coming TO Atlanta is doing one of these 4 things:
    a) coming for business, including the large convention business that ranks in top 3 in the US, and spending all or nearly all of their time in one of several parts of town that are well-connected to MARTA’s heavy rail lines (and the airport, of course)
    b) coming for a city vacation which would again mostly put them in one of the same parts of town that are connected to MARTA now or within a very short Uber ride to anywhere they want to go
    c) coming to visit family or friends in the area. Those family or friends will happily pick up at the train station that is nearest their home. See above for convenience of likely HSR stations relative to metro population.
    d) coming to ATL from a smaller market that has few and expensive flights, heading to somewhere far away. All HSR routes proposed for Atlanta would have a station at ATL and likely one downtown adjacent to MARTA’s red and gold lines that connect directly to the main ATL terminal.

    Yes, I know Atlanta is a moderately decent transit-connected city by US standards, even if it needs to be significantly better. But I think it really proves your point about the 2nd-tier US transit cities. HSR has a ton of potential for people coming to and leaving Atlanta _if_ the travel times to/from places like Chattanooga, Nashville, Birmingham, Greenville, Charlotte, and DC are competitive. Improved traditional rail service has nowhere near the same potential because it would not be especially competitive with either driving or flying for the uses I described.

    • Henry Miller

      It isn’t a necessary condition, except when the line calculates out to marginal anyway. If the line is marginal, then local transit is enough a factor in people’s choices as to make up a difference between a useful line and a waste of money.

      How you get around is a factor in your choices. Your choices are, go someplace else, rent a car, take a taxi (uber is a taxi), drive your own car, take local transit, or get someone to pick you up. Each of these has different pros and cons, and you are allowed to mix and match. Taxi and ride with friends/relatives means you won’t be getting around much, because of the high costs for each trip (either $, or just calling in a personal favor). Local transit gives you the freedom of the city subject to the timetable and size of service. Go someplace else means some other trip that might have different compromises (how you get around might be a factor in what trip you decide to take when there are competing choices). Of course driving assumes you don’t take HSR at all, but once you are there it gives you freedom of the city.

      Note that HSR and local transit should feed off of each other. Even if you can afford a rental car or taxi, if there is good local transit people will tend to prefer it because of cost. Those who are thinking about using friends will use transit if possible, just so they don’t have to call in too many favors. This means HSR should get more riders to the local transit system, and in turn more ROI for that system, in turn making a useful expansion easier to justify. The other way, good local transit means people can think about selling a car, and that means they are more interested in HSR because even driving is not an option. The two together are powerful.

      That is why I keep saying we need good local transit before HSR, it isn’t a necessary condition at all, but it is a useful one. Those who car about the environment, or safety (I’m this group) know that long distance travel isn’t a big factor in either.

      Note that don’t take the above too fast. Even if there is good local transit and HSR it will take people years to figure out it is good enough and change. As they say about affordable housing: the best time to build it was 20 years ago.

      • Autolycus

        FWIW, I don’t disagree about the benefits of local transit, and I definitely agree that it is complimentary to HSR (and air travel, for that matter). My big worry about your suggestion that we “need good local transit _before_ HSR” is twofold:

        1) We need both sooner than later. Delaying any effort at either will just delay adoption of both. If improved transit isn’t a necessary condition for HSR, why wait to start on HSR in earnest until the local transit is improved? Why not work on both at the same time?

        2) Local transit is… well… local. The federal government can certainly provide funds and incentives to cities to offer better transit, but without a major power shift, the federal government isn’t going to be building transit for them. It might take 40-50 years before some of the cities on major potential HSR routes have decent transit, if they ever do at all.

        If you’re limiting your “wait until there’s transit” just to the very marginal routes (maybe Nashville-Memphis?), then our disagreement isn’t too strong. If you’re saying we shouldn’t build anything outside the NEC until there’s better transit elsewhere, then it’s a strong disagreement because there are some very strong potential HSR routes even with the current transit systems, and by the time HSR can actually be built out, transit should also be improving in some markets, and the appeal of HSR might even add to the local appeal of transit. As we agree, they’re complimentary systems. Having a HSR rail line from Atlanta-Nashville might increase the interest in local spending on transit in Nashville, for example.

        I suspect our disagreement here isn’t really that strong. I just don’t want to lose ground on any progress toward a real HSR network because we’re hoping transit will get better in some cities that we can’t control. HSR is a little easier to control nationally, IMO. Maybe I’m terribly wrong on that point.

        • Henry Miller

          Our disagreement isn’t that strong. My priority is on local transit, because my focus is on safety, and local car trips are the most dangerous (in large part because they are most common).

          As for your points, 1 is correct. I’d do a different mix of funding, but you are correct to say we should work on both. Both feed into each other, so I can’t get the best possible local transit without HSR links to other nearby cities.

          Local transit is local, but the federal can do more to encourage good systems. Transit systems already report to the NTD, just collecting different metrics goes a long way. If hourly buses are called ADA service and so disabled rider data is reported (this is tricky for privacy reasons!) on those, it pushed people to ask the question is this bus serving the disabled. If peak and commuter service costs are expected to be entirely covered by fare collection, while all day every day rides are given more generous allowances for subsidies… Lots of other ways to collect and report data to make useful transit something that is encouraged.

          That the NEC isn’t already HSR is incompetence in Amtrak over the last 40 years. You can’t blame the republicans for this – if the democrats hired a competent management when there were in charge you would at least already have a the sharpest corners on their current route taken care of. Instead of doing study after study on the potential, the same money could have bought the needed land to fix a corner and built it – the rest of the line would be slow yet, but a few I of these over time will make a difference. Likewise they would be running more of their faster trains that make money, (or at least fought the congressional metrics that made this impossible)

      • Bobson Dugnutt

        @Henry Miller:

        That is why I keep saying we need good local transit before HSR

        Don’t kick own-goals. Your argument establishes a condition. IF = public transit THEN = high speed rail. The relationship is one way and the first must be fulfilled in order to fulfill the second.

        This relationship opens itself up to failure in 4 ways, by having your own logic used against you.

        1. IF you don’t build out your transit system first, THEN you would also jeopardize a high-speed rail station you might have gotten by being along the way or otherwise having a ripe market.
        2. IF you do build out your transit system, THEN high-speed rail access is not built (the project gets scrapped, your local officials don’t buy in, or your town will only get tracks but not a station) = END you have to scrap projects in progress and eliminate service added since its earlier state because the lack of HSR destroyed your use case.
        3. IF you do build out your transit system, THEN high-speed rail access is built ELSE transit’s share of access to HSR is low = END you have to scrap projects in progress and eliminate service added since its earlier state because your transit system has failed to facilitate sufficient transfer activity that was promised.
        4. IF you do build out your transit system, THEN high-speed rail access is built ELSE HSR’s share of local transit ridership is low = END you have to reduce service to your HSR station because most service is local circulation; you are either overserving the HSR station and might be taking riders out of the way or your transit system is proving a costly amenity of little or no benefit to local ridership.

        Scenario 1 can happen anywhere and you really have no idea how a community reacts. Scenarios 2, 3 and 4 put you in an “even if you succeed, you fail” scenario. You’re likely going to run into Scenario 2 in the Sunbelt, where cities are oriented around high population growth but their car-centered suburban land use will hobble local transit’s effectiveness. It’s a bad idea to deny Sunbelt cities rail service and keep them out of the network because of a doctrine that believes HSR before transit is like putting your underpants over your trousers.

        Scenario 3 is going to be encountered in midsize to larger metros, particularly in much of the Midwest and the Central Valley stations on California’s HSR line. In the latter, one of the reasons why those cities designated their downtowns as their train stations was because that is where their local bus networks converge. In many cases, the existing San Joaquin train depots already function as their central transfer points — some also offer Greyhound and/or Mexican intercity bus service. Fresno’s buses currently converge around a courthouse and park a quarter-mile away from their depot but the buses will be rerouted to be on the way of HSR. In these scenarios, just a few HSR passengers would dramatically improve bus ridership. Yet most bus riders are going to the train depot to change buses, not to ride the train. Most train riders will be coming by car or taxi/Uber/Lyft, and there’s already a lot of parking in place.

        Scenario 4 is going to be encountered in the biggest cities, the likely American HSR hubs. Think Chicago, Los Angeles, San Francisco, Washington, New York or Boston. These have high ridership on buses and trains, growing local ridership would pose a challenge. We know by these metro areas’ size, travel patterns and economies, HSR supply alone will create its own demand. HSR ridership will be high, and transit’s share of connections to HSR will be high. The same cannot be said of HSR riders’ share of overall transit traffic. It will be a small segment of many; transit will continue to predominantly handle local work, school, shopping and recreational trips and HSR will not dramatically alter the patterns.

        Transit and high-speed rail are contingent arguments. A contingent argument doesn’t bear the burdens of cause, necessity and/or sufficiency that a conditional argument has.

        You would get better transit, better high-speed rail and better interface between the two if you made three separate arguments independent of one another.

        An argument for better transit is sufficient on its own merits. We know transit’s benefits will accrue locally. Argue for improvements that benefit the community right now. Argue for improvements that will succeed with or without high-speed rail.

        • Henry Miller

          @Bobson Dugnutt

          People will make those arguments, but there are all wrong because they assume the reason to build a local system is to attract HSR; when should be get local people around town: to work, church, shopping, and whatever else they want to do. If there is a HSR in town, then local transit should get you there, but if there isn’t HSR you won’t be using it anyway. Of course every argument you state can be put into a 30 second made for TV news soundbite, while the counter doesn’t, so TV will be filled with arguments that I’m wrong without actually presenting my side.

          My argument is about safety. I’m interested in HSR, but safety is my focus. Local transit has the most potential to get people to stop driving cars. It is not possible for a human to be a safe driver, even the “better than average driver” is too dangerous. Maybe self driving cars can do better, but they are not coming fast enough, a bus at least allows better than average drivers, and trains can be fully grade separated and automated 20 years ago. HSR has the same safety benefits, but since most trips are local it has much less impact per dollar spent (money is a strong second concern of mine).

          For some cities local transit is good enough that HSR will have more safety improvements than than diminishing returns of more local transit. HSR in the NEC should solve enough safety issues as to be worth more than improving local transit on the big cities it serves. There are a few smaller cities along the way that I wouldn’t give HSR, but because they are lucky enough to be on the way they get HSR – their local system might need a lot of work before they HSR is the most useful way to improve their safety. This same applies to some other proposed HSR routes as well.

          As to your arguments:

          1: This is the only one with a bit of truth to it. If the HSR line is marginal, then a good local system can attract just enough more riders as to make the line worth it. Note here that I said line, not station. Stations are (well should be, but Alon has discussed this plenty) cheap enough compared to the cost of the line that they don’t matter. Either the HSR will exist and so you get a station, or it doesn’t generate enough passengers and it doesn’t matter what you do.

          The above is good news for cities without good local transit that happen to be on the way between two cities. It doesn’t matter how bad South Bend’s local transit it, any Chicago to Detroit will stop there because any additional passengers it picks up are profit (unless someone figures out cheap underwater tunnels). Indianapolis isn’t on the way of that line, so no matter what they do they won’t get on that stop (though they are on other lines and so should get HSR by them).

          2 3 and 4 are false because the purpose of building local transit isn’t to attract HSR, it is to get locals around town. Even if HSR never comes, local transit is how you get to the bowling alley. HSR makes local transit better because there are more places to go, but it is still useful without. (This applies to HSR as well, even without local transit it should still be useful).

          The only exception is the line that is built only to get to the HSR station. Ideally the HSR station is downtown where there already is a lot of local transit and this doesn’t matter at all. However in a few cases the HSR station is outside of town for some reason, and so local transit needs to make a special trip to get there. In those cases you wouldn’t even start that one bus until the grand opening of HSR (or perhaps a week sooner for practice purposes). If you build rail (rail needing years to build needs to be done in advance) to the future HSR station and it doesn’t come – well in business we refer to this is a sunk cost – it happens despite your best efforts: write the rails off and forget about it. For the rail case you shouldn’t be doing much building until the HSR is under construction, if you buy right of way at least you can sell it back latter so there is no big loss from that investment.

    • Onux

      “d) coming to ATL from a smaller market that has few and expensive flights, heading to somewhere far away.”

      Most people transferring at ATL come from a medium or large market with many well priced flights – that is why ATL is a big hub. Even if HSR stops at the airport, for most it will be more convenient to get off a plane inside the terminal they will board the next plane, with bags transferred, rather than get off a train to go through security. For someplace small like Greenville it might work, but cities like Charlotte, Nashville, etc. are full of flights to ATL and dwarf small city passenger volume.

      This is why I am skeptical that an Atlanta-Florida HSR link would capture much ATL-ORL passenger volume. Most of those passengers didn’t originate in Atlanta, they are coming from all over the eastern US.

      • Autolycus

        I was also definitely not thinking about HSR as a potential replacement for Nashville or Charlotte connections through ATL. You’re right that those are competitive air markets with frequent and affordable flights (often non-stop or connecting once in ATL, O’Hare, DFW or CLT). They’re also too far from ATL to make driving or HSR an option for replacing the one flight. I was thinking about Greenville, Chattanooga, and Athens along likely HSR routes plus less-likely-to-ever-have-HSR Montgomery or Mobile. Birmingham also has some amount of people driving to ATL, but mostly for international flights rather than longer domestic flights since Birmingham has become a more competitive air market since the arrival of Southwest and Frontier. Until those airlines added Birmingham, it was much more common to hear of people making the drive to save money.

        My “d” was not intended to suggest that category is anywhere near as prevalent as the others. It is certainly the least common, but it was the last category I could think of for people coming to Atlanta. (business, tourists, family and friends being themselves quite broad and diverse categories).

        • Onux

          I guess my point is that the number of flyers from Greenville/Chattanooga/Athens/etc. combined is very small, probably smaller than the flyers from just Nashville or Charlotte, and thus would not drive any meaningful HSR ridership, unlike possibly your other points.

      • Herbert

        A six hour train trip is better than waiting for your connecting flight for six hours

        • adirondacker12800

          If you are waiting six hours for a connecting flight chances are good that your origin or your destination wouldn’t have a train station either.

          • Herbert

            Now why is that?

            I would not even break a sweat finding counter-examples to your claim…

          • Phake Nick

            My understanding is, the Atlanta airport is dominated by connection flights offer by Delta which usually organize their flights into different banks, most major origin or destination have multiple flights across the day to facilitate different connections throughout the day, having to wait six hours mean either your origin or your destination isn’t one of such bigger market, and is thus unlikely to see a high speed station being constructed there

          • Phake Nick

            If a destination cannot get multi daily flight from nearby hub, it mean the number of passengers it can produce/attract is less than, say, 70 passengers per day, which is capacity of a regional jet. I don’t think it’s generally worth building a high speed rail station (and slowing down some trains for the station) for a destination with such number of passenger.

          • adirondacker12800

            According to Wikipedia Fresno has service on Alaska, Delta and United to LAX. Unites also goes to San Francisco. If I was flying east I don’t have to go to LAX. Bakersfield has service to LAX but less service in general too. Which tells me they are driving to LAX or one of the other Los Angeles airports.

          • adirondacker12800

            Silly me I picked the ones that might be called “nearby”.

          • Bobson Dugnutt

            @Phake Nick, miss something?

            I don’t think it’s generally worth building a high speed rail station (and slowing down some trains for the station) for a destination with such number of passenger.

            The San Joaquin, which is what HSR will become, has Fresno and Bakersfield involved in its top 10 city pairings, including both cities. https://www.railpassengers.org/site/assets/files/3477/39.pdf

            The train attracted nearly 370,000 “activity” in 2019. So roughly 1,000 passengers passed through Fresno’s train station to or from a train. The stop is also the city’s Greyhound station. Ground-level intercity transportation is important, and Fresno has other intercity bus carriers like Flixbus and Tufesa serving downtown. They’re all going where the train is.

          • Phake Nick

            @Bobson Dugnutt What I mean is destinations that are only being served by about 1 daily commuter flight from Atlanta airport by the hub airlines, which I am pretty sure Fresno and Bakersfield isn’t one such destination?

          • adirondacker12800

            Atlanta is realllly realllly far away and it takes a realllllly reallly long time to get there which discourages people from going there. You do understand that the more widely spaced the dots on the map are the farther away those places are? Here on planet Earth anyway.

  10. Luke

    I think South Korea’s overall situation is a great example of many of these things. Sure, there’s massive, well-transit-served Seoul at the end of its high speed lines, and Busan is also fairly large and well served by urban rail systems. However, the Honam HSR to serves Gwangju at its distant end. Gwangju is not a very centered city, has just one metro line, and the high-speed rail station is at the far western end of that line, outside of most of the city. Despite all of this, the KTX has a dominant modal share from Seoul-Gwangju.

  11. R. W. Rynerson

    A couple of notes;

    + In years of experience, business travel may actually be broken into subcategories. At the least in North America, there is an urgent category that demands air travel and a more relaxed category that will consider alternative modes. Sometimes it’s silly to use air travel on a short enough trip but doing so conveys a sense of urgency for some. The more relaxed category includes people going to conferences, religious activities, continuing education, or combining business with pleasure. Marketing regional rail or HSR needs to take that into account.

    + It is like herding kittens to develop intermodal connections between urban and intercity systems in North America. There are little red squiggly lines under the word intermodal on my screen right now, as though there is no such word. I’ve been working on the subject since I came back from Europe in 1971 and even where there is a good set-up eternal vigilance is required. In Portland, Oregon, the first situation that I worked on, the last piece of an intermodal complex at Union Station went into place with the Tri-Met light rail line four decades later. Even though the tracks are set up so that a one-seat ride could have been provided between Portland International Airport and Union Station, a transfer is required and that frightens air travelers. Then real estate developers noticed that the architecturally prize winning Greyhound station had LRT service and a clamor for redevelopment was introduced. In the Seattle area, I worked on the idea of a rail stop at Longacres Race Track which evolved into the Tukwila station but it requires a bus to bus transfer to reach Sea-Tac airport. In Edmonton, which had the potential for a good set-up, the individual pieces were scattered to the (cold) winds in favor of massive parking lots. In Denver, after three decades of fighting we have a good set-up, but there are people who would like the under-study regional rail line to be routed through the airport instead of Union Station.

    + Then there are categories of engineers. About 1/3 of those whom I have worked with understood the timed-transfer focal point concept by experience, about 1/3 understood it in the abstract after discussion, and about 1/3 just want to build stuff. Like the construction unions, they think it’s someone else’s problem to figure out what to do with their project. Each HSR or other capital project needs to be examined with that warning in mind.

  12. Onux

    “This is why a large majority of people take a train that takes three hours over a plane that takes one: hardly anyone lives near the airport or has an airport as their ultimate destination.”

    No one has the airport as an ultimate destination (no one has a train station, either) but given urban sprawl in the US is it really true that no one lives near the airport? People in NY can usually reach Midtown easier than JFK, and CDG is 3/4 surrounded by woods outside of Paris, but the Phoenix & Las Vegas airports are almost downtown, Chicagoland swallows Midway and O’Hare, and DFW is more central to the Metroplex than downtown Dallas.

    This is not an absolute rule (someone mentioned ATL being on the wrong side from Buckhead) but it does seem urban geography is different in the US. This shouldn’t make HSR a failure (especially places like the NEC), but could it meaningfully affect ridership and require a different formula to evaluate it?

    For instance, @Benjamin Turon mentioned the advantage of multiple LA stations, but Burbank and Orange County also have airports, so for intra-CA flights they don’t have to go to LAX, giving HSR no consistent access advantage.

    Further, in places with developed transit it is quicker to get to the center than a peripheral location like an airport, but somewhere with no transit where the assumption is “drive to the station like you drive to the airport” the opposite is true – traffic is usually worst heading downtown but less on orbitals. Thus it could be faster to reach the airport than downtown in a car. I realize this can be reduced by placing an outer city station as many have suggested, but this just puts the train on a level playing field with the airport – some have a shorter drive to one, some to the other – and should thus still affect ridership.

    How large do you think this effect would be? I would think that, as you pointed out, for business travelers downtown access always wins, but for leisure and visiting friends/relatives will this be the case?

    • Matthew Hutton

      Most people who fly fly a lot, so anyone who gives the remotest shit about their personal carbon footprint should be taking an HSR journey rather than a flight when the HSR journey is up to 4-5 hours, and frankly we should be disappointed with people who don’t.

    • df1982

      That’s another one of the differences between tourist trips and visiting friends/relatives noted above. Like business travellers, tourists will also be mainly headed to the inner core of a city, so there is at least an advantage of HSR on the arrival side. The exception is major tourist draws like Disneyland, but if these are big enough they will have their own station (as is planned for Orlando).

      VFRs have much more dispersed destinations, so the relative advantage of downtown arrival is less. That’s probably the hardest market segment for HSR to win over, also because you will probably be more car dependent than a business traveller to a downtown office or a tourist staying in an inner-city hotel

      • Herbert

        VFR almost by definition have access to someone to pick them up.

        But one more important thing is that VFR tends to be more price than time sensitive…

        • Yom Sen

          Why by definition? Being picked up at the airport/station in a city the size of LA can be a big ask. Last time I was in London and Paris visiting family nobody offered to pick me up…

          • Yom Sen

            Yeah my friends in Regensburg don’t like me neither. Last time I drove there they didn’t offer to pick me up at the autobahn exit…

          • anonymouse observer

            Do you know how big the LA actually is? Even the City of L.A. is 10 times bigger than Paris, and it is a fraction of the “L.A.”:

            L.A. County is bigger than some of the Northeastern states:

            Also, traffic is unpredictable and heavy, so people living there avoid driving around beyond certain distances. Even on the regular days, they don’t go more than 10 miles even if they are picking up someone from family or close friends because of the traffic. They also stick on freeways even if the traffic is heavy until smartphone becomes a commodity because they typically don’t know the local streets outside of their neighborhood very well. SNL made a skit out of this phenomenon:

            Good luck making someone picking you up at LAX. It is already bad on regular days, but on the worst day, it takes a half an hour just to get into the airport if you are driving:

            I took the FlyAway to LAX around the midnight on one day after Christmas couple of years ago, and it took 45 minutes just to get from Sepulveda to International Terminal.

          • michaelrjames

            @anonymouse “Even the City of L.A. is 10 times bigger than Paris, and it is a fraction of the ‘L.A.'”

            A few misrepresentations there. Proper, oranges-versus-oranges comparison:
            Ile-de-France (Greater Paris urbanised region): 12,174,880 residents (2017) on 12,012 km2
            LA County: 12,447,000 (2020) on 10,510 km2

            (For purposes of accuracy, these land areas are not of urbanised land, and thus should not be used to calculate residential density as in most meaningful definitions of the term. Such a simple, if not simplistic, metric would be arithmetic density not urban density.)

          • adirondacker12800

            Some people, like Noah Smith, who Alon quoted in the original post, live in one of GM’s Futurama propaganda films where there never any congestion and there is always someplace to park. And the train station or airport is short drive away.

          • Benjamin Turon

            LOL on the Futurama reference 🙂 … It did have High Speed Rail 🙂

          • Bobson Dugnutt

            @herbert, the problem might be specific to the LAX airport. A former limo chauffeur who regularly had to drive passengers to and from LAX describes the experience as going to an IRS audit right after a root canal.

            It’s LAX’s architecture that makes it so agonizing to get to and be inside of. The more people who use it, the worse the experience gets. Remember that LAX is the Southern California gateway airport, so its ridership base is geographically nearly half of California. And flying overall is generally expensive, so even with the availability of the other airports, flyers will often choose LAX over the airport nearer them because of the abundance of flights, a lower ticket price over the same distance, or they are traveling internationally.

            There are nine terminals around two looping roads, both of which are one-way and oriented counter-clockwise. One level is for departures, the other is for arrivals. Three terminals are on the north loop. At the west is the Tom Bradley International Terminal. On the south loop are four more terminals. There are parking structures in the middle of the loops. So amid this scrum you have curbside drop-offs, people going through the loop, people coming from the parking lots, a dizzying fleet of shuttle buses … and we haven’t even brought up what happened when Uber and Lyft arrived on the scene. To solve it, LAX officials, Uber and Lyft managed to reinvent the taxi queue. Hilarity ensued.

            I don’t fly enough and I’m not in tune with the aviation community to know what would be a good way to fix LAX. I don’t like to fly LAX if I have the choice, but my preferred mode of travel is the FlyAway express bus to Union Station, which is actually a pretty great transit solution for the LAX problem. I’ve generally thought that the U.S.’s attempts at integrating mass transit into airports are executed poorly. BART to SFO is highly overrated. OAK’s people mover thing to the Coliseum is somewhat better, it’s more to do with OAK being a much easier airport to navigate.

            The two I like are Portland, because the light rail is actually connected to the terminal and the airport experience itself is very pleasant. Another pretty decent one is Phoenix. The airport built a people mover connection to the light rail platform with big signage that says, “Hey I’m a people mover! Take me to the airport!” Light rail is frequent and quick to go to downtown Phoenix or Tempe. Sky Harbor has a great flight selection and the airport is big without feeling big.

  13. Bobson Dugnutt

    @Autolycus wrote:

    Local transit is… well… local. The federal government can certainly provide funds and incentives to cities to offer better transit, but without a major power shift, the federal government isn’t going to be building transit for them. It might take 40-50 years before some of the cities on major potential HSR routes have decent transit, if they ever do at all.

    Another thing: Local transit is fundamentally operations costs.

    The bus-rail argument is fundamentally an operations vs. capital argument.

    The budget is about taking money you have right now and spending them on the costs of the past to keep what you have (operations) or spending them on the costs of the future and having nothing to show for it (capital).

    Capital costs are not just building projects — capital is also replacing vehicles at the end of their service lives as well as building maintenance facilities for the vehicles and the tools and technology the workers need to keep the vehicles moving.

    Capital costs are also definite. They have a certain limit of money and time. You also have the option of not spending on capital.

    Operations costs on the other hand are indefinite (they recur all the time), are consumed (you are paying for the time the driver already operated the vehicle, the fuel your vehicle burned, and the parts worn down through use) and are unavoidable. No spending, no service. You can only cut future losses but can’t recover past costs — you can’t give the driver their hours back, you can’t unburn fuel, the vehicle doesn’t “wear up” and improve its performance through use, etc.

    The US federal government paying for capital costs makes perfect sense because it a treasure house. It collects the most revenue, and it could afford to do so because major capital projects as a percentage of debt service are lower than if states and municipalities floated bonds themselves. A major infrastructure project, like California’s high-speed rail, is untenable if California (even as wealthy as it is) has to go it alone. It’s about 5% of its GDP, which must be divided among its current expenditures (budget), pensions and service of other debts.

    At the US level, CaHSR is practically paying out of petty cash. It doesn’t imperil California or the US fiscally to fund it nationally.

    So capital spending is worthwhile, and on the balance has been generally beneficial to the economy.

    Continuous federal operations spending, on the other hand, is not worthwhile. Public transit’s benefits are accrued locally, so should costs. The peril of fourth-hand funding (farebox revenues are first, local subsidies are second-hand, state subsidies are third-hand) is that the federal government imposes a one-size-fits all rule across transit systems. This is bad because the federal government will become the bargaining agent of management, and unions will want to impose an East Coast or San Francisco Muni CBA on all agencies. This makes suburban, small-town and rural transit untenable. And what happens when the other party, which for cultural reasons anathemizes transit, gains power? If they don’t succeed in eliminating transit outright, they’ll try to impose “shock treatments” (e.g. Thatcherization), or test out some theory that comes up against a transit organization’s capabilities or culture and riders will bear the splash damage of such failures.

      • Bobson Dugnutt

        @AJ, have you actually seen the schedule for a rural fixed-route bus? It’s something like Amtrak’s Sunset Limited. You get a bus, maybe one round trip, and sometimes the frequency is less than daily. Then it’s usually a long ride to the nearest town.

        This is representative of rural transit. Mendocino County runs a bus line that will get you into Santa Rosa, the largest city at the northern edge of the San Francisco Bay Area. https://mendocinotransit.org/timetables/cc-rider-southbound-mon-sat/

        You must get to the Denny’s at the cock crow, and stop along the way to pick up passengers in Ukiah. You arrive in Santa Rosa four hours later! If you choose the SMART train, you’d wait two hours until a 12:30 pm departure toward the Larkspur Ferry Terminal for another 1’10” ride. It should be timed to meet the ferry into San Francisco, and it should be another half-hour or so. You could also take the Golden Gate Transit bus across the namesake’s bridge into San Francisco for another 2’30”.

        The rural service is provided at great expense, and it’s not even functional as mobility. Does your life align with rising to the cock crow and going to the Denny’s that early in the morning? Why don’t we “buy everyone a car” or “pay Uber” for a lot less cost — Google Maps says that the Fort Bragg Denny’s to Salesforce Transit Center in the heart of San Francisco is about 3’15”? It is a lot less costly, but only if riders agree to the same conditions and only if the productivity of the bus is so abysmal that only a handful of people use it. If Uber or Lyft were more attractive options, like allowing a trip anytime, or capping the fare a rider must pay, then the costs would explode the transportation budget.

        The fiscal case for rural bus service looks bad, but in the bigger picture, the expense is something we as a society can afford. We’re looking at the option of one Mendocino bus or zero. There’s probably not much of a clamor for high-frequency service along that long of a route, but a few more trips per day might goose up some additional ridership and even bring the per-rider cost down. And the community would know what their demands are.

        So even if the per-unit cost of rural transportation is so high, getting rid of the option for rural communities cuts off their access and yields so little savings that you might only be able to buy a school tripper run a day in San Francisco, where ridership is so high that you wouldn’t even notice its effect on service.

        • df1982

          The cynic in me thinks that heavily subsidised transit service like rural or small-town lines are deliberately made to be as inconvenient as possible so that only those who have absolutely no other alternatives would ever countenance using it, while any normal person would just stay in their car, since more passengers only means potentially having to provide more loss-making service. Hence why these lines often have ridiculously circuitous routes and passenger-hostile timetables.

          • adirondacker12800

            The ridiculously circuitous routes are for the disabled people who can’t drive. Or walk across a parking lot out to a bus stop on the street.

          • Phake Nick

            I don’t know much about transit in small towns in the US, but from other part of the world, it is the unavoidable result when the density is low and everybody drive, so your transit is only going to be used by very few people and thus have limited fund and resources, yet you still need to cover as much area in your place as possible, and the result of that is guaranteed to be inconvenient timetable and circuitous route.

          • Alon Levy

            From talking to people in that orbit, I don’t think that’s true. Rather,

            1. The planners never ride the buses, so they have no idea what works.
            2. They think that circuitous routes are okay for serving off-arterial destinations (see for example Sioux City).
            3. They are incredibly institutionally conservative, partly due to their perceived powerlessness (transit in small cities is unimportant so they are unimportant), partly due to the fact that ambitious people do more interesting things than run a network with 20 buses, partly due to political constraints. A lot of this gets laundered through “what if people yell at us at community meetings?”, because the chief skill of a potted plant manager is to deflect responsibility (likewise, “other people will say you’re a troublemaker” is about the worst threat your manager can make at you at the office).
            4. As an outgrowth of 3, they’re disconnected from broader society and knowledge. Random example: I called MVTA to confirm a story that I heard from an anonymous source within the agency about BEB trouble. I identified myself as writing for CityLab, and the person I was talking to to confirm did not know what CityLab was (or misheard me, maybe?). And that’s an agency serving middle-class suburban white flighters, not the poor.

          • adirondacker12800

            The planners never ride the buses, so they have no idea what works.

            Why would anybody in Sioux City who is capable of driving and owns a car take the bus? Why are you surprised? You do realize their idea of bad traffic is having three cars ahead of them at the light and there is parking everywhere? I’ve actually seen people with their mobility aids sitting on the bench just outside of the destination’s doors waiting patiently for the bus. In places like Sioux City the buses are for people who can’t drive and can’t walk across a parking lot. If they don’t bring their own electric scooter chair they use the ones the store provides.
            It seems that walking is more socially acceptable east of the Appalachians. In lovely Glens Falls, why would I wait for the bus when I can walk there in 10 or 15 minutes?

            middle-class suburban white flighters, not the poor.

            The Minnesota Valley Transit Authority? It’s Minnesota, for there to be white flighters there has be someone or something to be fleeing.
            Awww it’s too bad whoever you spoke to isn’t in your cultural orbit. Cross cultural sensitives work both ways ya know. Maybe you and whoever can have a sensitivity session where you bring food. Expect something involving Miracle Whip and Jello. Maybe you will both get lucky and find out ethnic Scandinavian could come from a New York City appetizer store.

            Why does it matter you are from CItyLab? It’s a public agency, anybody can ask for any public information.

          • Alon Levy

            My point is that in places like Sioux City, because the middle class, from which planners are drawn, drives 100% of the time, public transit planning is suboptimal.

            Re MVTA: they gave me the information I was looking for – my working for CityLab isn’t the issue, it’s that they didn’t know what it was. And there’s absolutely people of color for white Minneapolis suburbanites to be fleeing; the cops are doing their best to reduce the black population in Minneapolis (or increasingly suburbs like Brooklyn Center), but they’re not 100% efficient at that, and there’s absolutely negative suburban stereotyping of the city as a place with high crime (read: too many POC) and bad schools (read: too many POC).

          • adirondacker12800

            Poor people in Sioux City drive everywhere too. There is no traffic and there are plenty of places to park. The only way you are going to get them on the bus is to ban cars.
            Pesky Wikipedia. Minneapolis is majority white. So is St. Paul. I’m confused. Is that you are pissed off that poor people choose to live in inexpensive neighborhoods where they can live car-free or one-car or pissed off that they aren’t filled with white symbol manipulators? Or pissed off that they let people move to Minnesota changing the demographics from lily-white? To have Somalis move into a neighborhood someone has to move out. I suppose they could have just let the Somalis die in refugee camps.
            It’s a bad thing that middle class black people moved to Brooklyn Center? That works the same as with the Somali. For someone to move in someone has to move out or die. And Brooklyn Center had it’s suburban explosion when Minneapolis was still lily white. I’m confused.

          • Alon Levy

            Minneapolis public schools are only 34% non-Hispanic white. And the Brooklyn Center population is as I understand it black working-class, not black middle-class, rather like Ferguson, and as in Ferguson, the local political elite (e.g. police and police leadership) is white and funds services out of fines and fees and is hostile to the growing black population.

          • adirondacker12800

            I look at Census Data not memes. Yet again, it’s usually real easy to find the Quick Facts page for a place by asking your favorite search engine for the name of the incorporated place, state and the word census. Or the CDP if you happen to know that. All places in New England, the MidAtlantic and Michigan are in an incorporated place.


            Brooklyn Center is 71 percent not-black. The median household’s income is lower than the state’s but comfortable. It’s poverty rate is higher than the state’s but not outrageous either. I’m sure Minneapolis schools can be an interesting subject but public school students in Brooklyn Center don’t attend them.

          • Bobson Dugnutt

            @Alon Levy, there’s another aspect to the Sioux City-like institutional obtuseness of tiny transit systems. They are emaciated.

            A transit system is akin to a skin-and-bones child with really pronounced ribs when the system map is a daisy chain of one-way loops with all routes converging at a central transfer point.

            Even with service this awful, there is bound to be people using it. Usually it is the elderly population, because a bus system this bad is a step up from being a shut-in.

            The issue comes down to money. A community just doesn’t put in the money to make a service not-useless.

            There are also two features you tend to see with emaciated transit systems: 1) When a community doesn’t put money into a transit system, it can’t afford the domain knowledge of knowing how to plan and operate a transit system. Chances are, the transit “department” is a box-under-a-box in the city’s organization chart. It’ll be treated as a task of the Department of Transportation because, well, buses use roads and our traffic engineers can engineer a bus solution for us. They probably had to in college. We’ll assign it to the DoT department head.

            Or, it’ll be treated as a task of the Public Services department because the city parks the buses in the same yard as the garbage trucks and the utility vehicles. The head of Public Services has to draw up routes for the garbage collectors and schedule maintenance for the parks and streets, and buses need schedules and routes, so …

            2) The press release that uses the pseudo-empowerment word of “Partnership” with the private sector. In this case, local leaders announced they are “partnering” with an internationally famous private company with experience running buses in the biggest cities in France, the United Kingdom, some other prestigious European and North American city … and now they’ve come to help our humble little bus system!

            The reality: Transdev or First Transit won the tender on a contract we put out and for the next few years they’ll be the temp agency for our frontline workers. They will not do anything beyond the duties required of them in the contract … and they sometimes slow-walk the duties once they have the contract.

            Woe be to your city if your “partnership” calls on the contractor to supply the vehicles. Your fleet will probably be a school bus with the yellow painted over, a former airport shuttle bus with unusually wide doors, whatever can be trolled from a big-city transit agency’s expired-fleet auction, etc. Chances are if you see a city running a 1980s or 1990s vintage Gillig Phantom or RTS in regular service in 2021, it’s a “partnership.” In about another decade, you might even get an Orion or a NABI!

          • adirondacker12800

            Even with service this awful, there is bound to be people using it. Usually it is the elderly population, because a bus system this bad is a step up from being a shut-in.
            They want to go places that aren’t on a fantasy map you find aesthetically pleasing. The want to go places where they can do things like buy cheaper food to stretch their food stamp budget. To a bus stop right outside the door because they can’t walk across a parking lot.

          • Oreg

            Adirondacker points to an interesting problem. A good service for people with mobility impairments is fundamentally different from that for the able-bodied. The former need stations closely spaced and very close to their destinations. The latter are happy to accept longer walks for high efficiency with straight routes and fewer stops. These must be provided by separate services.

          • adirondacker12800

            *Bangs head on desk* The able bodied people already have a service. Their automobiles. *Bangs head on desk again* The buses already go to the destinations. There aren’t many or any that the buses you are imagining could go to. You don’t understand that it’s an hour to take a moderately brisk walk across the parts of the metro area that aren’t suburban petering out into woods and farmland. And you would pass many of the destinations the bus goes to because there aren’t many of them.

          • Henry Miller

            The current black statistics are not relavent to white flight. In 1950 there were close enough to zero black people in Minneapolis. People still left for the suburbs leaving the poor behind. Somehow since then the blacks have moved into North Minneapolis and Brooklyn Center, and the area is now considered by locals the dangerous parts of town (and police statistics back them up). Though I think most locals will agree that poverty is the cause of violence not race, and their action somewhat back that up (which isn’t to there isn’t a significant amount of racist people there, but they are a small %).

          • Bobson Dugnutt

            They want to go places that aren’t on a fantasy map you find aesthetically pleasing.


            My hot take: Time works the same way for everyone, everywhere. A person dependent on such an emaciated system is not a bus rider. They are a hostage.

            They are trapped on a system that is constrained by the budget and resources of the government body providing the service. These are some of the Basics issues Jarrett Walker covered in his “Human Transit” book. The basics are what every transit agency regardless of size, budget or urban form has to do.

            Frequency is freedom, but infrequency is captivity. Infrequent service is useless to most of the population because the start time of a trip and the end time at the destination must both work in order for ridership to occur. Why don’t not-elderly people use these services? Because their lives revolve around someone else’s schedules, like their employer’s, their school’s, or the start time of something enjoyable like a movie.

            Being on the way is impossible on a one-way loop. A bus network of one-way loops is drawn that way so that resources are spent equally across a political jurisdiction. *That* is the real aesthetically pleasing map, not the hypothetical one you wish I had drawn. The one way loops symbolize access, but the reality from behind the bus windshield is far more bleak. On a loop, if you are anywhere except the halfway point of the route, one of the trips will be longer than the other and force a ride out of the way.

            Short service spans hark back to what Alon pointed out of planners being out of touch on what works. True: Not only do planners of these systems not know what works, these planners also do not know what *does not* work. Emaciated transit systems keep planners hours for the same reason banks keep bankers hours (i.e., expanding transit system coverage for late-night hours or on weekends increases ridership but at a higher cost than weekday peak or midday service). They lean too heavily on sound practices taught to them in an academic or a public administration context to the detriment of the people the system is designed to serve.

          • adirondacker12800

            I live in a metro of 125k. I can drive from one side of it where there will never be a bus because there are too few people to the other side of it where there will never be a bus because there are too few people in 20 minutes during what passes for “bad” rush hour traffic. Why would I attempt to arrange myself around a bus that runs every 15? that leaves me at the far side of the parking lot where I can get a parking space closer to the door? No matter how hard you clap Wendy it isn’t going to happen unless you ban automobiles.

          • Phake Nick

            @Alon Levy No doubt planners who never ride bus, institutionally conservative, and disconnected from broader society and knowledge on planning of bus route would have negative impact on bus planning, however given the limited amount of funding and resources, which isn’t going to improve because there is an limit on how much passengers you can attract in such a small city, it make sense for the route to be planned this way. Like in the Sioux City example, yes you can break down the South Sioux bus route into two more frequent routes that run on two artillery with double frequency, but it would mean losing direct access to destinations that are away from artillery road, and even requiring additional interchanges for travelling between some different destinations in the South Sioux area. This is a trade off that would be deemed acceptable and even desirable in larger cities like New York due to the density often mean there are other alternative routes or modes nearby that would help connect or fill gaps in between radial lines or such, but in places like Sioux City it very likely would be deemed unfavorable by passengers who need to head to those places.

          • Alon Levy

            They already run an interchange-based system, with an hourly pulse in city center; what they miss is that hourly pulses at this range are terrible.

          • Henry Miller

            @Phake Nick true, but a subsided taxi for the disabled to live in the areas to be lost, would cost less than the several buses doing that route, so we take one bus away from poor parts, make the rest of the bus network sane for the places it serves, then you can actually get some people riding (remember there is several years delay between making a sane network and people responding), which gives us some fare recovery which in turn allows either more buses on the busy routes, or another sane bus route to get more people. Subsided taxi is generally refereed to as dial a ride, it will never pay for itself, but the point is for the cost of a bus we can cover 10 bad bus routes for the few people who will need it.

            Don’t forget that in a town of that size there are probably several WalMarts, and they all substitute for each other almost perfectly (not so for other stores – there might be 5 hobby stores in town, but odds are each specializes in something different). The bus network can choose the ones easiest to choose and not serve the others. The people dependent on transit will just go to the ones easiest to get to

          • adirondacker12800

            Having all six bus lines all arrive at the cargo cult transit center once a hour is better than three of them meet there at :00 and the other three meet there at :30.

          • adirondacker12800

            Then they don’t go to destinations and no one will ride them. You’ll have a nice map that runs buses half hourly, empty.

          • Phake Nick

            @Alon Levy Pulling Google Map of South Sioux, the logic of this bus route seems rather clear
            First, after departing from the city center interchange hub, it turn to the west side of those houses along the river, and then to the east side of them, then travel down the Dakota Avenue, and then up along G Street, with a detour to US-77 in the middle and then back to the city center.
            If you eliminate the loop near the river, then people from home on the far west and east side of that part of the city, or with destination in those area, could need to walk >1km before reaching their nearest bus stop. Mid-density apartments on the east, hotel, inn, and YMCA near the river, are all over 500m away from the nearest bus stop otherwise.
            For the segment of bus heading down the Dakota Avenue then going up along G Street, residents who are willing to walk longer distance can already directly walk to bus stop in the other direction to take the bus to get a shorter journey time, otherwise the bus provide direct access to houses and facilities further east from the main Dakota Avenue by travelling up the G Street which are otherwise six blocks away from the bus stops along the avenue.
            And in the middle of the G Street segment it diverted to serve Walmart and other facilities along the US-77, providing everyone living along the bus route in South Sioux direct access to those places, while still covering those houses in the middle of the detoured section through bus stops on the lateral streets
            By splitting the route into a US-77 route and a Dakota Avenue route, it would mean that anything along the river and anything further east of the main avenue would lost bus service, and even those still get direct services near the Dakota Avenue would not be able to access facilities along US-77 without interchange. Or to put it the other way, those facilities along US-77 aren’t particular unique but they serve the area for the convenience of local population. If an interchange is needed for the local population in South Sioux to access those facilities, they might as well just head to other shops that are near city center and have more convenient access options, and thus the US-77 route would have minimal ridership mainly use by staffs of those shops and occasional riders who decided they want to go to those shops even with an additional interchange, given that there are barely any residents along the US-77 for the bus to serve, and something like two-third of the South Sioux would feel they’re becoming further away from their nearest bus stop, with many now need ten or more minutes walk to their neatest bus stop from their home, cancelling much of the advantage of having a faster bus, in addition to the much narrower coverage area that make some of their destination also needing ten or more minutes walk away from nearest bus stop.

            @Henry Miller In such low density area, buses cannot just choose to cover where they want, they also have to let people from different part around the town go to different places for job, social support, and various other functions. This is not an option but is the need of them. An on-demand system for those in need might cost less, but it’s an additional hurdle for any residents who might suddenly arise the need to move across the city from any point to any destination for any reasons they didn’t anticipate ahead of time.
            To promote transit ridership, yes speed is important, but it shouldn’t come at the expense of sacrificing existing riders who use the system. Where resources permit, it would be a good idea to run an additional fast service along main corridor to attract transit user and development in such part of the city, but it shouldn’t come from existing passengers.
            Not to mention reducing coverage area this way could also have impact on ridership it can attract.
            As for attracting passenger to pay for bus system expansion through fare recovery, my understanding is that fare recovery rate of most of these system are in the range of 10+%, in other word, for a system with 10 buses, it need to double the entire system’s ridership by 100% without inducing additional expense in order to squeeze out enough revenue to add one single more bus into the system.

  14. AJ

    In your comment on HSR as a catalyst, is the vision to expand adjacent to an existing CBD or to create an entirely new node of density? This pairs with my comment on the last post on whether it is important to directly serve the city centers of smaller Midwestern cities if a brownfield site can provide a straighter & cheaper HSR alignment. Do you have a preference or does it always depend on the specifics?

    • Phake Nick

      Me myself think, given example of out-of-town high speed rail stations like Guangzhou South in China, Shin-Yokohama and Shin-Osaka in Japan, with plans to develop area around those stations into new central business districts but those location don’t reaaly seem to be living up the expectation, I think it is best to avoid such mode of development.

      • Andrew in Ezo

        For Shin-Yokohama and Shin-Osaka, the raison d’etre was to avoid excessive curves (HSR hates them) which would have been necessary to serve the central station (and *two* bridge crossings of the Yodo River in the case of Osaka Umeda). Development was likely a secondary consideration, and, Shin Yokohama at least seems to have been moderately successful- I am old enough to remember it back in the late 1970’s as a patch of marshy farmland with cheap and seedy hotels here and there, but now filled with office buildings. an arena and football stadium signifying a secondary cbd. Access will get certainly get better with the Sotetsu/Tokyu link with an underground three platform station scheduled to open end of FY2022.

    • Bobson Dugnutt

      @AJ, don’t know who you were addressing, but I’ll try to answer your question.

      The answer: Yes and yes! The decision to plan, build and run high-speed rail is inherently political. If you are paying for it, own a property affected by it, planning to ride it, or tasked with making a decision … people’s preferences are crucial, even essential in a democracy. These preferences must be weighed alongside, and sometimes against, economic and engineering decisions that might run counter to preferences.

      We often call a decision “political” when people’s considerations try to subvert, run counter to, or overcome economists’ and engineers’ brute facts and brute math. Deferring to experts is equally as political as deferring to political calculus over textbook calculus. A non-decision is a decision.

      There’s no right or wrong answer as to where to site a station. We have high-speed rail experience to learn from, and the French, Japanese, Germans, and Chinese have different engineering, planning and operations philosophies shaped by their lived experiences of politics, costs and engineering challenges. If we consult with one of these nations to build a high-speed rail network, we will likely import their philosophies into our thinking. Or worse, Americans will try to think things through themselves and just buy really fast locomotives to hook onto Amtrak trains. It’ll be like watching a sumo wrestler trying to sprint a 5-K lol.

      So if you hired the French, you’d probably get center city service only to the terminii but the cities in the middle will be served by outer-edge or beetfield stations. A Japanese approach would be ideal for a very linear corridor like in the Northeast, Florida or California. The French, German and Spanish approaches might be a better fit for the Midwest, Texas and Atlanta hubs, where you are dealing with a multidirectional constellation of cities and terminii.

      But America is so big and vast and maybe political consensus requires stitching together a transcontinental railroad. This is uncomfortable for Americans to bear, but we’d have to bring China to the table.

      China’s high-speed rail is philosophically very different from other examples in the world. It’s not trying to hit that sweet spot of faster than highway travel and moderately slower than plane travel but competitive at medium ranges. It’s literally building and running high-speed transcontinental rail to make thousands of kilometers of rail travel that used to take days into a day or less. China has overnighter service on lines that are at least 1,000 km long, and services that took several days can now be completed overnight. The Beijinger blog showed a sampling of some of the long-distance HSR routes in China: https://www.thebeijinger.com/blog/2018/08/01/just-about-everything-you-need-know-about-high-speed-train-travel-china.

      There’s a train from Beijing to Guangzhou, the prosperous port city on the way to Hong Kong and the gambling mecca of Macao. The end-to-end time is 8-9 hours, but the distance is … 1,400 miles!

      So we might want to call on China if we want or need to spend a workday along the Empire Builder, California Zephyr, Southwest Chief, Coast Starlight or dare I say the Southwest Chief. Just in case.

      For fun, I looked at the ridership statistics available on the Rail Passengers Association’s website. ( https://www.railpassengers.org/resources/ridership-statistics/ ) I specifically wanted to know if there’s any point of running “land cruises” of these picturesque but agonizingly long routes that take days to complete. I would think that in the jet and interstate highway age, Amtrak service relegated itself to mid-range regional travel. There’s no way someone would want to go between Chicago-Seattle, Chicago-Oakland, Chicago-Los Angeles, Seattle-Los Angeles or Los Angeles-New Orleans — which all happen to be the terminii of the land cruise routes.

      The city pair stats left me simultaneously delighted and horrified.

      • Phake Nick

        The Chinese overnight train service between Guangzhou and Beijing only run 4x weekly and even counting through trains to other destination they can only offer 6 trains each days. There are more daytine trains and more air flights between the two cities, with aircraft like A380 also deployed onto the route

          • Bobson Dugnutt

            @Ryland L, what I found jaw-dropping when I looked at the Rail Passengers ridership stats is how many Americans are riding long-distance trains. In 2019. With a mature aviation and interstate highway system. On Amtrak. Those multi-day land cruises have the end-to-end distance as the No. 1 or No. 2 city pair.

            The stats list popular city pairs but don’t list the tickets or sleepers sold for those pairs. You have to do some math by looking at the annual ridership and station activity.

  15. adirondacker12800

    And in NYC, it would mean dealing with the nightmare that is Penn Station — a thoroughly stressful and inconvenient experience.

    People who get paid to say things will come up with all sorts of stuff.

    Being stuck on the Helix for 45 minutes is a delight. It goes well with being stuck in one of the tunnels in Baltimore. Or the airport and TSA cavity searches. Fly into New York and take a train from EWR or JFK you end up in…………….Penn Station……… I suppose one could take PATH or the subway. The bus gets you to the salubrious experience of the Port Authority Bus Terminal. . …. ah the Connecticut Turnpike in the rain. it’s just fabulous….. It’s almost as good at the Merritt and the Hutchinson River Parkways!

    …. then there is the George Will screed about how Real Americans(tm) love Freedom and Liberty. And taking trains would impurify their precious bodily fluids turning them into Communists. Who was then seen soon after getting off a business class car of the Northeast Regional by Paul Krugman.

    When the Hudson crossings, in 2001, were closed or HOV-3 between 6 and 10 the teevee reporters went and interviewed people hoping to make it through before 6. They claimed it was “faster”. Than what? Kayaking down the Hackensack River?

    People who don’t want to rub elbows with people not their sort realize it’s impolite to say “Are you nutz? I might encounter someone! Who might be different!” and come up with all sorts of explanations or excuses. Especially if they are getting paid for it.

  16. adirondacker12800

    Upstate New York between intercity and regional trains.

    How are you going to ban automobiles first?

    There aren’t going to be regional trains between Amsterdam and Schenectady or Auburn and Syracuse or Rome and Utica. There aren’t enough people and they all own cars.

    It seems like the Vermont AOT realizes that there are only 600,000 people in all of Vermont and control their fantasies. The 7 or 8 trains a day that run once every two hours from Burlington to Saratoga Springs have to coordinate with a train to Boston and a train to New York. Perhaps there can be a once an hour train that stops at all the stations between Saratoga Springs and New York. That can leave a few minutes after the train to New York and the train to Boston leave. The people who want to get to Rhinecliff, both them a week, can gaze at the traffic on I-87 and the ramps to the parking lots for 15 minutes.

    And they aren’t going to be crossing platforms in Albany. There’s enough demand for Boston-Toronto and Boston-Montreal to fill whole trains. There can be different trains for New York-Montreal and New York-Toronto.

    • fjod

      In any other developed country, Amsterdam-Schenectady-Albany would have regional rail and (if you sorted the station locations/connections at both ends out) it would be very time-competitive. Given new onward connections in Albany with HSR, this would be a no-brainer. European countries have high car ownership too, so I doubt it’s *that* much of a problem – but with marginally better land-use policies (densification, parking minimisation, freeway removal) the car-use problem disappears.

      • adirondacker12800

        Bulldozing Amsterdam and the rest of Montgomery County to let it go back to nature isn’t very “green”. Make them all move to Albany and Schnectady it would bring Albany’s density back up to what it was in 1970. Not it’s peak which was in 1950. Household size was much larger back then. There may be more housing unit in Albany than there were back then,

  17. adirondacker12800

    Passengers can time themselves to arrive 10 minutes before the intercity train departs, even today.

    Not if you are on a branch that runs once an hour off peak, that arrives at :06. 6 minutes after Acela has left and a minute after the Regional. It will be less of a problem when it’s every half hour on most branches and somewhere around every 15 for wherever you are going. You do understand that the LIRR is very very clever and instead of having 20 tracks so that all the trains arrive at the same time they have the trains all use a much smaller set of tracks a few minutes apart?
    Hikari and Kodama stopping at Brooklyn, Jamaica and Farmingdale means they don’t have to change in Penn Station. Once an hour Kodama that also stops at East New York, Garden City-ish and Yaphank means they can schedule for that. Yaphankese and their neighbors seem to be interested in the concept. Yaphankians? Brookhavenites? Time it in the summer so the Hamptons express loiters around a bit for Bostonians and Philadelpians….. Maybe perhaps on summer Friday evenings and Sunday afternoons have the Hikari stop there too. There is likely a lot of to’ing and fro’ing from the National Labs. Hikari all the time might be good.

    commuter train arrives on track 19

    If I’m getting off the train I don’t care what track it’s on. Why are you printing tickets, in 2021? And they have had things like this for decades.

    • R. W. Rynerson

      Knowing arriving track numbers aids in making connections. It’s 1,000 feet between Tracks 1 and 12 in Denver Union Station, for example and the airport trains are on Track 1. There are plenty of stations in other parts of the world that have longer stretches.

      Also, in stations where open access to platforms is permitted, knowing that the ‘California Zephyr’ is arriving on Track 4 makes it easier to spot arriving friends.

      • adirondacker12800

        Having the track number printed on your friend’s ticket doesn’t do you much good.
        I hope other things arrive and depart from track four. Having a whole platform dedicated to the once a day land cruise seems a bit of overkill to me. So that you don’t have to glance as at in information board or check your smartphone.

        • yuuka

          Why would they? You know your friend is on the California Zephyr, if a Coast Starlight pulls in you know your friend isn’t on it. Come on…

          • adirondacker12800

            The Coast Starlight doesn’t go through Denver. That solves his problem. I don’t know how altered the schedules are right now. And assuming any of the long distance train are on time is iffy. Right now the southbound Coast Starlight is supposed to get to Emeryville at 8:30 AM on Tuesdays, Thursdays and Saturdays. The westbound California Zephyr is supposed to get to Emeryville at 4:10 PM on Mondays, Wednesdays and Fridays. I doubt there would be much confusion. The northbound and the eastbound arrive between 9 and 10 AM. On alternate days.

  18. James S

    Agree with everything you wrote. The argument Noah Smith makes is the second laziest. The laziest of course being “rail is old tech so its obsolete”

  19. yuuka

    On the other hand, this may have helped tank the Malaysia-Singapore HSR. Between decades of dirt cheap oil and cars, and the accompanying car-centric policies, or the Malay nationalists not being very happy at the outsized role Singapore plays in the HSR project, it’s not hard to see why. (name is clicky for more)

    But it may be different, since like in Japan and Taiwan the old meter-gauge lines are only limited to 140-160kph at most and any money going to HSR is money not going to meter gauge lines or other local transit systems. KL may also suffer from the same problems as LA, whereby not everyone wants to go to the area around the HSR terminal.

    They’re talking about cutting Singapore out of the picture, but the business case would sink even harder.

  20. Gok

    HSR in LA has another particular weakness: LA has 5 airports, so you can typically fly close to your real destination already.

    • Eric2

      I imagine that frequency is low to those airports. Also, the Bay Area has multiple airports. HSR supports a high frequency service between every LA stop and every Bay Area stop, finding a flight between your combination of airports at each end must be harder.

      • Car(e)-Free LA

        It isn’t. There are ~hourly flights on multiple airlines between all five LA Airports and all 3 Bay Area Airports. It’s an unbelievably well-developed air corridor. Still, LAUS is closer to more destinations than BUR or LAX and Transbay is closer to more destinations than SFO or OAK.

        • Bobson Dugnutt

          @Car(e)-Free LA, there’s also a wrinkle in L.A.-Bay Area market flights. While there is an abundant selection of airport combinations, you’d think that rationally people would just fly out of the closest airport to their homes.

          Turns out, many people will fly the LAX-SFO pair because it’s always the cheapest among the possible combinations even when they live closer to Burbank, Long Beach, Ontario or John Wayne. There’s something similar if flying out of Oakland or San Jose. The John Wayne flights tend to have the highest fares. San Diego and Sacramento are too far out of the LAX-Bay Area orbit. Santa Rosa and Palm Springs have flights too but their use is negligible.

        • Onux

          @Car(e)-Free LA: Technically you cannot get flights from Long Beach to SFO yet, although both United and Southwest are starting that route later this year.

          @BobsonDugutt: In 2019 an average of 9,655 people flew from LAX to one of the three Bay Area airports, but an average of 13,118 flew from one of the other 4 LA airports to the Bay Area. Most LAX-Bay passengers are to SFO and may have been for connections; they are #2 and #4 in the country for international flights. With the other four airports, traffic to OAK and SJC is larger than to SFO.

          If the other four airports carry significant intra-CA load, that isn’t great news for CAHSR. Burbank HSR and Burbank Airport will be next to each other, Burbank has almost as many OAK passengers as LAX, and CAHSR isn’t planned to serve OAK’s East Bay market. With access time the same, a shorter flight, and not debarking on the wrong side of the Bay, I don’t see CAHSR capturing those passengers.

          • Bobson Dugnutt

            @Onux wrote:

            If the other four airports carry significant intra-CA load, that isn’t great news for CAHSR. Burbank HSR and Burbank Airport will be next to each other, Burbank has almost as many OAK passengers as LAX, and CAHSR isn’t planned to serve OAK’s East Bay market. With access time the same, a shorter flight, and not debarking on the wrong side of the Bay, I don’t see CAHSR capturing those passengers.

            I don’t expect a war of modal dominance between the Burbank-Bay Area flights and HSR. This is where trains-planes would have a symbiotic relationship. You might have the same or fewer flight selections between BUR and the Bay Area, perhaps the fewest between BUR and SJC and fewer between BUR and SFO, where there would be some market overlap.

            It wouldn’t affect Southwest’s operations much. Southwest has this flight pattern in which a flight between SoCal and NorCal (it does this at Sacramento’s SMF too) is a stop along the way. So the BUR-OAK flight might originate in Burbank, stop in OAK for 30-60 minutes, then serve some other destination to the north or east. Sometimes the plane arrives at Burbank as the midpoint, from Las Vegas or Phoenix, and goes to the Bay Area as its last stop.

            I think the passenger market will sort itself out with the train and the planes getting the passengers they want at the right price, schedule and user experience. HSR will not win on cruising speed, but it will get enough passengers who are willing to so somewhat slower on the ground for a lower ticket price, an improved user experience (e.g. more leg- and shoulder room, business travelers gain back the 30 minutes of unproductive time during takeoff and landing), and the same or better ground connections by serving center cities (HSR will take you to San Francisco faster than getting out at SFO in the next county over, getting on the people mover to the parking lot, and hopping on BART or taking a taxi/Uber/Lyft/rental car to San Francisco — think about how much time it takes to get into and out of enormous SFO!). And this is an all-stops train serving the Central Valley cities.

            Burbank might get a bump from plane-train traffic. Think of HSR and Burbank airports as complements. With fast, frequent connections to HSR, the train effectively can take you anywhere there’s a flight out of Burbank. So, you kinda sorta have a bullet train from LA to Vegas if you think about it.

          • anonymouse observer

            Building Burbank Station at Burbank Airport does’t sound like a good idea for CAHSR. The station should’ve been near Downtown Burbank. The airport is farther away from the 101/134, and building the station at the airport would not give the time-saving advantage in station access against the commercial flights. The high-speed rail station in Downtown Burbank should work as park-and-ride station given proximity to freeway exits (3 within one mile or so) and plenty of land to build parking.

            Also, the airport location is hard to connect by extension of existing rapid transit system except for the existing heavy rail tracks, which diverges just south of airport. Even on the heavy rail, someone needs to figure out how to somehow connect two existing “Burbank Airport” Stations on two separate rail lines located 4,000+ feet apart.

          • Onux

            @Bobson Dugnutt

            I agree the passenger loads will sort themselves out, the problem is many people think CAHSR will revolutionize travel in CA and dominate the N-S market. If sorting out involves segments where flying wins (like BUR-OAK) then ridership will suffer, hurting the economic case for HSR.

            HSR will not complement Burbank airport. Anywhere you might take the train from (Orange County, Inland Empire, eventually) also has a close by airport with direct flights to Vegas, Phoenix, etc. People in Bakersfield might do this, but even Fresno has direct flights to Vegas, plus train to Burbank, schlep to terminal, go through security, then catch a flight is probably unattractive with even driving from Fresno.

            HSR complements air when it stops at an international airport offering long distance flights to places the train doesn’t go, not when it stops at airports serving the same destinations as the train, only faster. Eurostar has not captured a lot of London-Paris and London-Brussels travel because it makes people go to Heathrow to get on a train.

          • adirondacker12800

            There are other airports around London. a very brief check of Luton and Gatwick, they have service to Paris-Charles De Gaulle. On the flying bus companies, but they have service.

          • archie4oz

            There’s not a lot of room for growth in the air corridor though. NorCal airports are largely the gating factor here as LAX handles as much traffic as all 3 NorCal airports combined.

    • Bobson Dugnutt

      @Gok, HSR’s advantage is that the stations are already on the way and HSR fills in the gap in the middle by serving the Central Valley in the way airlines can’t or won’t. Fresno is roughly equidistant to LA and the Bay Area. On a train, a ticket would cost half as much as a full-length route. Fresno has flights to the Bay Area and L.A., but they’re hundreds of dollars on a tiny plane.

      Even on the ground level, traffic on any and all modes is substantial between Northern and Southern California. Car traffic between the two regions is heavier. For Greyhound, it’s one of the network’s busiest corridors in the U.S., and the market sustains ethnic intercity buses. There are a few companies catering to migrant farm workers bringing them to-from the Central Valley and Southern California, Las Vegas, Arizona or Mexico. There are hubs in L.A. for connections to New Mexico and Texas. There are a couple of bus companies catering to Korean, Chinese and Vietnamese passengers.

      Now rail? Amtrak’s disconnected California network does have the nation’s second, third and fifth or sixth busiest train lines. California would run a north-south train were it not for the difficult topography. The weakest of the three corridors, the San Joaquin, is technically the L.A. to San Francisco train but you need a bus at either end to get to them. The Bakersfield-to-LA trunk is the busiest of the Amtrak Thruway feeder bus network in the state, and more than one bus is needed to handle all the passengers.

      Going up the coast is even worse. Unlike the Central Valley cities, the Central Coast cities are smaller, farther apart and slower-growing. Only Santa Barbara has a commute shed extending into western Ventura County. Santa Barbara, San Luis Obispo and Monterey counties have not yet converged. Meanwhile, in the Central Valley, Merced County has now been pulled into the Bay Area’s consolidated statistical area orbit. Yikes! The Sacramento MSA (2.65 million) should be within the Bay Area’s orbit too, as eastern Solano County is sending commuters to Sacramento as well as the Bay Area. Stockton, the MSA south of Sacramento, sends residents to both.

      It’s all a matter of tying together Southern California’s 22 million residents, Northern California’s 12.25 million residents and the 2.5 million residents in the heartland with one train. With California’s populations and economy, we can afford another travel option.

      • Phake Nick

        The Central Valley is not going to be main source of passenger for California high speed rail. It doesn’t have that much people. And for such of regional connections high speed rail is also unnecessarily too costly.

        • adirondacker12800

          The train is going to be there anyway. It’s cheaper to let them use it than have a separate system.

          • Benjamin Turon

            Good point, Fresno and Bakersfield are pretty sizable and are only “small” because they’re in California, they are as big or bigger in population (urban and metro) then the largest cities in most states.

          • adirondacker12800

            There is something going on where they want to connect five dots on the map. Relentlessly.
            Frenso, with a metro area bigger than El Paso’s and Bakersfield with a metro area bigger than El Paso are too small but El Paso more than 500 miles from San Antonio isn’t.

          • Phake Nick

            While it make sense for HSR to also serve local passengers, the amount of passenger in the middle would be so little that saying it is the advantage of HSR wouldn’t make the HSR worth constructing
            Drawing comparison from like El Paso doesn’t make sense either as most people aren’t going to believe it’s a good idea to draw a line into El Paso

          • adirondacker12800

            The tracks are going to be there. The extra expense is building a station

          • Richard Mlynarik

            “Urban” construction is billions more expensive than greenfield, and California is proof positive.

            “Existing Rights of Way” means billions of payola due to being “forced” to work around every-increasingly-extortionate freight railroad demands, while being “forced” to undertake extensive utility relocation ($$$) and being “forced” to construct extensive road and rail grade separations.

            “Existing Transportation Corridors” means billions for road and freeway projects out of the “high speed rail” slush fund — which is absolutely fine with the consultants and contractors who “design” and build this shit. The overwhelming majority of the photos you see out there of “high speed rail in California under construction bring good union jobs to marginalized communities” 100% unnecessary highway rebuilding, highway relocation, sprawlburb road grade separations, and out-of-control-dimensioned freight rail grade separations.

            “The tracks” are going “there” because of the “extra expense”. Except it turns out “the tracks” aren’t going there. Just the engineering consultant overhead, the agency overhead, the contractor overhead, the concrete and the rent-seeking profit extraction. The rails? Maybe, maybe not. Maybe single tracked. Maybe one train every two hours from nowhere to nowhere, maybe not. Maybe by 2040, probably not.

            I know, I know, I know, don’t feed the troll.

          • Benjamin Turon

            The only other way to serve Fresno and Bakersfield with HSR is branches like the TGV or the West Coast Mainline, leaving them out seems a bad idea both short-term politically and long-term social-economically. The problem with HSR in California is then incompetence of the CHSRA and their consultants.

          • Onux

            “ The only other way to serve Fresno and Bakersfield with HSR is branches like the TGV”

            The TGV is very successful with branches. The Paris-Marseille train does not stop in Lyon it goes around it, but both Paris-Marseille and Paris-Lyon are heavily trafficked markets. Alon has shown that this is not as good as through routing, but it isn’t bad.

            Even better, since the Central Valley cities already have tracks running through downtown used by the San Joaquin’s, instead of a branch you could keep the HSR tracks outside of the city with a crossover to the existing tracks on either side of the city. Electrify that portion and the trains serving Fresno/Bakersfield can still have the advantages of through routing and high speed service out of the city, with the trade off of slightly slower service, for much less cost.

            This setup also has the advantage of providing natural opportunities for times overtakes of regional trains by expresses, without making the slow trains sit at stations waiting to be passed.

          • michaelrjames


            “ The only other way to serve Fresno and Bakersfield with HSR is branches like the TGV”
            The TGV is very successful with branches. The Paris-Marseille train does not stop in Lyon it goes around it, but both Paris-Marseille and Paris-Lyon are heavily trafficked markets. Alon has shown that this is not as good as through routing, but it isn’t bad.

            Obviously I agree (as per my previous posts) and also I kinda supported (not really, merely by default) the early SNCF proposal of building on I5 land with branches into Fresno & Bakersfield. I previously gave the Paris-Bordeaux LGV as an example whereby they cut 40 minutes off the TGV journey to Paris by bypassing 5 towns (and this time saving is without those earlier TGVs stopping at those towns–it didn’t). However when looking at a map I find that Fresno is not 10km nor 20 km, not even 30km from I5. It’s a ridiculous 60 (sixty) km. That would mean a 120km round-trip diversion! Most of those TGV bypasses are barely out of the town/city, indeed Avignon-TGV is walking distance, and while Lyon-St-Exupery is 20km, it is next to the airport so has excellent connections but which is beside the point anyway because Lyon has two CBD TGV stations in addition to the bypass station …

            Now I suppose SNCF may have thought, if it was good enough for the highway planners to build the major north-south freeway in California to miss the biggest city in the Central Valley by sixty km, surely it’s good enough for the rail equivalent? Bakersfield is a bit closer but Palmdale is even worse, being an north-eastward diversion. This looks pretty weird but doubtless was driven by the same factor behind SNCF’s plan: I5 is taking the quickest route between SoCal and NoCal and those hicks in Fresno don’t want quick routes to LA or SF anyway, except for their annual Xmas shopping trip which today is done via Amazon. As usual the CaHSR is the victim of much earlier poor transport planning, notwithstanding all the other standard issues afflicting American infrastructure construction. No wonder both SNCF and the Japanese walked away from any involvement in the project. Impossible to resolve the mess without bottomless pits of money and endless often pointless political battles.

          • Benjamin Turon

            The Japanese with the Shinkansen does not seem to bypass large cities, the CaHSR system is very much more like the Shinkansen then TGV or other European HSR systems, I would expect to see a tier rail service, like the Shinkansen or Amtrak’s NEC. I think one good place to branch is at San Jose, with some trains going to San Francisco, but others going to Oakland.

          • adirondacker12800

            major north-south freeway in California to miss the biggest city in the Central Valley
            I-5 goes through Sacramento. It doesn’t go to San Francisco. It keeps heading towards Oregon from Sacramento. For another 120 miles or more depending on where it stops being Central Valley and turns into “foothills of the Cascades”.
            You have to see the extravaganza of highway they had planned in 1962 to figure out what they were expecting I-5 to do.

          • Bobson Dugnutt

            @adirondacker12800, a quibble: Fresno city is 525,000, Sacramento city is 500,000. Fresno metro is 1 million, Sacramento MSA is 2.3 million, Sacramento CSA is 2.6M and would be 3 million+ if Stockton and Solano County were attached to it and not the Bay Area.

            Fresno is more or less contiguous. Sacramento is comparatively tiny because it never annexed so many patches of unincorporated county neighborhoods around it before the contiguous sprawl to Citrus Heights and the county border with Roseville or the leviathan that became Elk Grove (a tenfold increase in population in 30 years).

          • michaelrjames

            @Bobson Dugnutt

            No need to quibble with Adirondacker. The point is that Sacramento happens to lie on the route of I5 which is not the case for Fresno or SF. I5 was built to connect the west coast states in the shortest, fastest, straightest route (which I think was his point, and was mine too).
            The difference between Fresno and SF w.r.t. I5 is that there are multiple freeway routes connecting the Bay Area to I5 but none to Fresno (at least on my ancient Rand McNally). These days perhaps the 99 route is not much slower than I5, but I doubt that was the case back in the 90s when SNCF made their proposal. It does beg the question of why similar savings in building the HSR would not be achieved by using the 99 adjacent land (ie. already owned by CA) and deploying the kind of bypasses in France we’re talking about (ie. ≈10 km from the centre).

          • adirondacker12800

            They let people from the suburbs use the train station. The bus station, the airport and the highways etc.

          • Phake Nick

            @adirondacker12800 your comment doesn’t make sense.
            The very first comment in this comment chain was that “HSR in LA has another particular weakness: LA has 5 airports, so you can typically fly close to your real destination already.”. Someone else responded “HSR’s advantage is that the stations are already on the way and HSR fills in the gap in the middle by serving the Central Valley in the way airlines can’t or won’t. Fresno is roughly equidistant to LA and the Bay Area.”, and then my comment was “The Central Valley is not going to be main source of passenger for California high speed rail. It doesn’t have that much people. And for such of regional connections high speed rail is also unnecessarily too costly.”.

            In other words, let say LA to SF is being served by a HSR, assuming there might be 2000 person travelling on the train per hour each way, and then maybe 20% of those, or about 400, would not take the train and instead take the flight because one of the five airports are more convenient to them than train stations location. Some argued that HSR have advantage in the sense that it can also serve Central Valley travellers, but my argument is that the number of travellers there would be less, like let say 100 or 200 per hour, and that cannot be used to support the HSR ridership if the flight took away most travellers (All numbers are made-up just for illustrative purpose)

            It have nothing to do with the cost of constructing the station or otherwise, and it also have nothing to do with whether there should be a separate system to serve them or not.

            @Onux Branch are bad and evil especially when they are unnecessary. Alon have also talked about it quite a few times from my memory. It reduce frequency unnecessarily.

          • adirondacker12800

            I want to establish you understand what a train is. You do know how trains work? It’s a very very bad thing if they leave the ground. I’m sure the rest of us hope they remain on the ground on the trip between Northern California and Southern California. Ideally they never lose contact with the rails which need to be attached to something firmly anchored. The undersea tunnel to have the trains go on the very straightest route would be very very expensive. So would tunneling through the mountains along the coast and viaducts between the tunnels. So they decided to sent the trains through the Central Valley where the terrain is much more amenable to keeping rails firmly attached to the ground and trains remaining in contact with them. They have to be there for people to get from Northern California to Southern Califonia or vice versa. Since they are going to be there the people along the way can use them too.

          • Phake Nick

            What you’re saying have nothing to do with the argument on the number of airport’s impact on the railway ridership

          • Onux

            @Michael James

            I was not proposing I-5 (see note below). I was suggesting HSR to Fresno, etc. but tracks stay outside the cities with access downtown through existing infrastructure. For instance, HSR tracks could leave the planned route at E American Ave, skirt west of Fresno via West Park/ Patton, and rejoin the planned route at W Herndon Ave. At both locations the CAHSR route is next to existing tracks into Fresno, so only two junctions are needed. The bypass is the same length as the planned route, plus electrifying 25km of existing track through Fresno. Fresno Station would be downtown; no 20-120km detour.

            Trip time LA-SF express (the critical market) is the same or better (if speed through Fresno/Bakersfield is less than 350kph), construction cost is much less (at grade with few separations), and Fresno can reach every other destination including SF/Sac/LA in 2:00 or less using HSR tracks everywhere but through CV cities. The disadvantage is that trips to/from the CV (as well as LA-Bay trips on services that stop in the CV) are slightly longer because of the slower speed through town.

            Ironically, this is how CAHSR is building tracks around Madera, Hanford, and Corcoran, with the disadvantage of out of town stations for the first two instead of junctions and existing tracks to get downtown.

            This is a combination of the German approach (build HSR tracks between cities, use existing tracks through them) and the French approach (build HSR tracks past cities, for faster service to destinations beyond). If it makes you feel better, you can say that Fresno and Bakersfield are on the mainline, while there is a high speed bypass around them for expresses.

            Note: The I-5/SNCF option is viable, with the shortest/fastest route SF-LA. If pursued, the solution would not have been branches to Fresno/Bakersfield. It would have been electrifying and upgrading to medium speed (110mph with junctions, 125mph with separations) a route down the east side of the valley (ideally the UP route along Hwy 99 directly through Turlock, Visalia, etc. for good local/regional service; not the current San Joaquin BNSF route with stops out of town). HSR trainsets would serve the CV, but at medium speed between junctions with the I-5 HSR line in Tejon and Stockton.

            While viable, this option is not good. The CV has 4.3M people, it’s not literally farm towns, and on I-5 SF-LA sees little gain but Fresno/Bakersfield large loss if SF/Sac/LA trips become 3+ hr. To save time SF-LA far better to pick Tejon over Tehachapi, or Altamont over Pacheco.

          • michaelrjames


            Really I wasn’t disagreeing but pointing out the nature of the original SNCF proposal and one reason for its failure to launch. We’re agreed that the “French” approach would have worked on both counts, ie. maintaining good timing while serving major cities but this is not what SNCF proposed. No way does the TGV in France ignore a city the size of Fresno or Bakersfield. I can only assume that back then there was little viable alternative because highway 99 was far less suitable (from what others are saying) than today with many more town bypasses and upgrading to expressway (Interstate Highway) standards. SNCF were correct in predicting that the politicians’ preferred solution would be horrendously expensive, and quite possibly anathema to the whole point of HSR–fast travel between the LA and Bay Area mega-connurbations. In retrospect it might have been clever to have proposed upgrading 99 in association with building HSR to share its ROW, something that would have got support of the road lobby. But it seems that the arguments fell into these extremes–a route that totally misses the main cities and towns of the CV and was a political non-starter, or one that builds right into their hearts.

            Incidentally, some of the other comments (by others) here on French versus other approaches are simplistic. The French don’t ignore smaller cities & towns, they just prioritize the faster direct route between major destinations. All of the TGV-bypassed towns retain their train stations and trains; indeed they often get TGVs just not as frequent as the mainline. In fact there have been attacks by neoliberal ideologues on the excessive number of TGV stations, of this ilk:

            France’s high-speed TGV trains ‘running out of steam’ – official report
            Audit Office blames pressure from local authorities for creation of ‘incoherent’ network that often failed to meet necessary criteria
            Anne Penketh, 27 October 2014

            The Cour des Comptes (Audit Office) blamed local authorities for pressuring the state to allow the TGV to pass through their towns, creating an “incoherent” network. As a result, there are now a total 230 TGV stations across France, many on lines which are loss-making for the state-run SNCF company.
            The report contains clear lessons for the UK, with chancellor George Osborne due to announce plans for HS3 between Leeds and Manchester on Monday. According to the French report, 33 years after the first TGV was launched, 40% of TGV trains still travel on conventional track rather than the specially built high-speed lines.

            This manages to get the reality completely backwards. Most of those stops are extensions–on standard not LGV track–which allows TGV service to extend to more parts of the network, and far from exacerbating ‘loss making’ actually improve the economics of a far-flung network. Closing them would not improve economic performance, unless closing them as destinations served by all trains, ie. a version of the Spinetta Report (destroy the network to save the network …). It doesn’t even apply to the betterave stations because, as discussed on this thread, Haute-Picardie, the prototypical betterave station in the fields of the Somme, wouldn’t be loss-making with its 400k pax p.a. It seems this kind of confusion continues to this day, some of it deliberate obfuscation, some simple misapprehension.

          • Onux

            @Benjamin Turon: The Shinkansen does not bypass cities, but Japan has a metro area with almost the population California, another larger than the LA Combined Statistical Area, another the size of the SF CSA, another larger than the entire San Joaquin Valley, two more the size of the Sacramento CSA, and another larger than Fresno or Bakersfield CSA, all on a line 25% longer than SF to San Diego. The three largest plus one as big as Sacramento are closer than LA and SF. Even then, the Tokaido Shinkansen bypasses the Osaka CBD by a few km and the Joetsu Shinkansen shares the Tohoku line to Tokyo station instead of getting its own tracks to Shinjuku Station – both because of the cost of building new HSR tracks through downtown.

            @Phake Nick: I would not call branches evil, but I do agree with you and Alon that they are generally bad railroad policy because they split frequency and remove the advantage of serving all stops on a corridor. I would not do French-style branches but prefer my bypass scenario above. But as you point out potential lost customers from SF-LA traffic probably exceeds total passengers from the Central Valley, so a cost-benefit analysis is needed. Fresno and Bakersfield are absolutely large enough to warrant service, but not large enough to drive the whole route or spend extra billions of dollars on.

          • adirondacker12800

            Why do people going from Northern California to Southern California or vice versa care where the station in Bakersfield is?

          • Onux

            “Why do people going from Northern California to Southern California or vice versa care where the station in Bakersfield is?”

            Generally, they don’t. However, they care a little bit if the station is in the center of the city and causes their trip to be slower because the train cannot go as fast through the city as it would elsewhere. If they are going from Norcal/SoCal TO Bakersfield they care because they very much want the station to be in the center of town, closest to where their destination is.

            Since Bakersfield is not big enough for the latter group to be very large, the best of both worlds would have been for the HSR tracks to go around the city, with a bypass on legacy infrastructure to access a station downtown.

          • adirondacker12800

            Then there are two ROWs to maintain and the people on the diversion get lousy frequency. Or the frequency is higher and the people on the train that diverts get a longer trip.

          • Bobson Dugnutt

            @Phake Nick, it will be difficult for any of us who support or oppose high-speed rail to guess what the ridership effects are.

            We keep going around in circles about how much or how little ridership between pairs are because of the availability heuristic https://en.wikipedia.org/wiki/Availability_heuristic . What we as supporters and opponents have available to us is a Pokemon battle of rail vs. plane or rail vs. car, and that there’s a single determinant that determines a winner and loser.

            The plane beats the train because speed is absolute. The car beats the train because access (you can come and go as you please) and no transfers are absolute. Ergo, the ridership will be zero.

            Except real-world scenarios are more complicated. In the absence of HSR, the car travel share of Southern California to Northern California dwarfs the air travel market. Does that mean air travel has failed to capture the car market? That logic is utterly fatuous.

            The travel patterns have sorted themselves out because of the availability of highways, airports and planes to serve the routes. California’s not at the point where it has to face a Sophie’s choice and choose one to survive.

            The reality is: Accurate, precise modeling is incredibly hard. And even experts who have the math and computing know-how to execute these models are ultimately spitting out paper versions of these Pokemon battles because the law requires it.

            The models need to be multidimensional, because real-world decisions aren’t single factoral and modal share is only zero when zero supply is available. Array a matrix of attributes, like Trip Price, Access, User Experience and Speed. Then display your choices: car, train, plane.

            In the absence of numbers: we can rank Car 1 or 2 for Trip Price, 1 for Access, 2 or 3 for User Experience, and 3 for Speed. (I use variables for trip price and user experience because of the great variation in car values, amenities and conditions. A brand new convertible sports car would offer a tremendous user experience at a high cost to purchase and fuel up. An everyday commuter car with some miles on it might have low costs to operate, but if you are like many Americans who cannot afford a major repair bill or a down payment on another car should it break down, the trip may be more expensive than necessary).

            Plane ranks 3 for Access (the airport and journey to and from are the most loathed parts of the trip), 2 or 3 for User Experience (you have to pay a lot for your flight to make a 3 a 2, and most of us will never be able to afford a 1 UX) and 1 for Speed.

            Where would HSR rank? Likely 2 for Trip Price, 1 or 2 for Access, 1 for User Experience and 2 for Speed. The 1’s are people who are likely to take the train anyway because they value access and the user experience and would set the baseline for ridership. The rest will come from the marginal scenarios in the car-train trip price overlap or the train-plane speed overlap. Some drivers might trade off driving for a more expensive trip in exchange for the better user experience or not risking an expensive repair bill over a long journey. This is actually a large pool of drivers, so even a small percentage of trips diverted would mean a much more productive train.

            The plane rider might be willing to go on a slower trip on the ground for a lower fare if HSR would be frequent. Flyers already have to plan their day around the airlines’ schedule; they might come out ahead by a slower ground travel time if the schedule is more convenient for their starting time or arrival time. HSR will likely have clockface schedules for all-stops service, and larger corridors might have high-frequency service during rush hours.

            When it comes to competitive train and air service within the same pair, there’s a cost disadvantage over short travel times or short distances that airlines can’t surmount. It’s the cost of an unsold seat. It is crucial for an airplane to fill every seat because its n(n-1) is just 2 for a nonstop flight. The cost of an unsold seat on a train is low because there’s a high chance that the seat will be sold somewhere else along the line, or a sold seat turns over and is filled again over the course of a journey. A five-stop train has 20 possible trip pairs to sell seats. A 10-stop train has 90! On trains, cars can be added or taken away based upon demand. Plane flights have to be canceled or, either added at great expense, or foregone and ticket prices raised instead because gates are not available.

          • Richard Mlynarik

            “Why do people going from Northern California to Southern California or vice versa care where the station in Bakersfield is?”

            They’d have been riding HSR between LA and SF for a decade now if rent-seeking cost maximization weren’t the sole priority of the consultants who 100% control and profit form the California High Speed Rail Authority, and do so, to the tunes of tens and hundreds of billions of dollars, especially when there are no tracks and no trains run and there are no passengers to care.

            And if, in 20 or 30 years, 30 or 40 years late, there is are marginally-better-than-today’s-Amtrak trains running on a botched route LA — “downtown” Palmdale — “downtown” Bakersfield — “downtown” Fresno” — Los Banos — “downtown” San Jose (excluding actual downtown urban San Francisco, actual downtown urban Oakland, Sacramento, etc, for at least another decade or two, because, you know, baby steps, America may still be learning but will soon be World Class Thought Leaders in High Speed Ground Transportation, fuck yeah!) those trips on those trains will too expensive (to pay off the fraudulently promoted route’s astronomical construction costs), too slow (to accomodate that route) and, well, 20 to 30 years too late to make any difference at all.

            Any difference at all except, of course, to WSP (née PB, aka PBQD), Tutor-Saliba, and their very very very very very very special friends. They’ve been doing great off train-free California HSR for nearly 30 years now, and they’ll continue doing great off train-free California HSR for as long as it takes.

            Once again, the problem with US “public” works is that there is no representation of the public interest, at any level, in any way, at any stage of the process. It is grift, rent-seeking, kickbacks and earmarking from start to finish.

            Nobody in the public is allowed to care “where the station is”, or anything else. That was determined by private contractors in order to maximize the public-private wealth transfer. “People going from A to B” get what they’re given … or in the case of California HSR, they get nothing, while paying many tens of billions.

            Freedom of choice! USA! USA! USA!

          • adirondacker12800

            It’s too bad they didn’t consult your Omniscience, isn’t it?

          • Onux

            “Then there are two ROWs to maintain”
            The city access ROW is inexpensive to maintain – Class 4/5 track and catenary. There would have only been several tens of km throughout the whole valley. If you spread billions saved by not putting Class 8 track through cities over 40-50 years, you easily pay for slightly increased maintenance.

            “the people on the diversion get lousy frequency.”
            They get the exact frequency as before. No one expects every CAHSR train to make every stop, no different than Nozomi/Hakari/Kodama schedules on the Shinkansen. There will be trains that go from LA-SF without stopping in Fresno, let alone Madera. If a train is scheduled to go to that city it takes the bypass, otherwise it stays on the HSR tracks.

            “the people on the train that diverts get a longer trip.”
            Yes, this is the tradeoff of a bypass option. On the other hand, if speeds are 250kph in the city and 350 kph outside of it, everyone on the train that doesn’t go to that city gets a FASTER trip. Since more people will be going SF-Anaheim or LA-Sac then from Fresno/Bakersfield to anywhere, this is a good tradeoff if it saves money.

            Of course CAHSR has already cast their route in concrete and spent that money.

          • adirondacker12800

            Apparently the Californians can’t imagine anything other than BART and imagine all of the train inexorably stopping at at all the stations all the time except for the single test train that won’t. Class 5, I scared the cat when I giggled that loud.

          • Phake Nick

            @Bobson Dugnutt
            Indeed we cannot tell what the effect of distribution will be, but as another commentator have pointed out this is what model and simulations are for. It is not simple ranking, but it is needed to figure out how many passengers each option can attract from each destination pair and compare the combined result. Serving more destination doesn’t always mean better ridership if it mean losing more passenger on city pairs to other modes that would otherwise attract much more passengers. It need full analysis to find out which way work the best.
            As for cost of unsold seats, airlines have spent decades optimizing their pricing formula and aircraft utilization to minimize the amount of unsold seats on any services that they do run and nowadays it’s common for them to have system-wide load factor of at least 80%+. Most trains cannot come close to this number, although one can argue there are little need for trains to match this number due to the relatively less variable cost for marginal capacity on trains.

          • Bobson Dugnutt

            @Phake Nick:

            As for cost of unsold seats, airlines have spent decades optimizing their pricing formula and aircraft utilization to minimize the amount of unsold seats on any services that they do run and nowadays it’s common for them to have system-wide load factor of at least 80%+. Most trains cannot come close to this number, although one can argue there are little need for trains to match this number due to the relatively less variable cost for marginal capacity on trains.

            The principle you describe is yield management. The advent of cheap computing has made yield management a commodity in not only aviation, but also for trains (Amtrak uses it for long-distance trains and sleepers, but on state-supported services like in California* still uses an old-fashioned high/low season regime due to intercity trains doubling as commuter service), intercity buses (it’s how Megabus/BoltBus tries to bait you with $1 fares), hotel rooms, rental cars and ticketed sports/live arts and entertainment on both the first-run and resale markets.

            Yield management doesn’t give any entity using it a competitive edge because it’s so ubiquitous. The industry is probably a commodity market, where the lowest-price seller tries to dominate market share. If the market is saturated enough, sellers and producers try to segment the market and move up the value chain and sell higher-price and higher-value goods and services. (Think hotels and automobiles; trains and planes are class-segmented but must share a common vehicle.)

            Back to trains and planes. Trains don’t need to match 80% loads like planes do, for one because the load factor is invisible since seats turn over constantly throughout the trip and it’s normal for a line to have peaks and valleys among station pairs. On a train, you normally want the heaviest populations at the end of the line as the tent poles. It’s less efficient when the most dominant station is in the center because the loads tend to be biased toward the center. Conversely, it makes much more sense for airplanes to operate as a hub-and-spoke model for their network effects. (Southwest and low-fare carriers have a different route structure that’s a hybrid of rail and hub-and-spoke, because their revenue structure involves keeping the planes on the ground for as short of a time as possible.)

            Aviation has a peculiar problem: traveling in the third dimension (in the sky) is and always will be the most expensive form of transportation due to the energy required to defy gravity. There’s a physical limit as to how low costs can go. Make planes smaller, and the fixed costs of a flight grow. Make planes larger, and you lock out much of your flight market because of the physical infrastructure needed to support a Boeing 747 or Airbus 380 class jet is only viable at the heaviest-trafficked airports. Airlines cannot afford to get it wrong.

            Trains, on the other hand, have to get the market right only once: before the first track is laid. Unlike when railroads were the tech sector of their day, today the trains are trying to go where we know the people are today rather than trying to bring people to a settlement and grow a market. It’s less of a question than it was in the days of manifest destiny.

            *Amtrak California is kinda odd. It’s a business arrangement in which the service is branded as an Amtrak service and Amtrak acts as the contractor, but it fulfills the state’s business plans. The Pacific Surfliner, for instance, has a role of supplementing commuter service when Metrolink and Coaster don’t run, but it must also handle its dominant market of Los Angeles-to-San Diego intercity traffic. So the fares are fixed, like a commuter service, but one-way and roundtrip ticket prices rise slightly from Memorial Day to Labor Day and from Thanksgiving to New Year’s Day. There is one summer surge around the weekend of the San Diego Comic-Con, where a one-way trip between L.A. and San Diego would be say $33 or $39 depending on the season, jumps all the way to $55 or more. There are ticket books of 10 one-way pairs at a steep discount over the low-season rate (they are closer to the Metrolink and Coaster one-way fares), and generally the only blackout period is Comic-Con weekend, Thanksgiving weekend and Christmas-to-New Year week.

          • Alon Levy

            It matters that the highest-modal-split intercity rail networks – Japan and Switzerland’s – do not use any yield management. Not everything France’s airline-brain railroad execs do is good, in fact the TGV ridership trends since the late 2000s scream “hit the reset button on SNCF, without any French managerial elite in the loop.”

          • adirondacker12800

            Amtrak when they still ran unreserved trains and before that, when any coach ticket was good in any coach I seem to remember the announcement was “Your ticket is good for carriage, it does NOT guarantee you a seat” Lovely.

          • Phake Nick

            @Alon Levy In Japan, train companies are increasingly using things like earlybird tickets and internet ticket reservation system to push people getting their tickets ahead of time instead of buying the ticket as they travel, but it doesn’t seems like many are using them
            @Bobson Dugnutt Yield management isn’t just dynamic pricing, it also involve deploying vehicles when and where the capacity is needed. Airlines are much more flexible than train companies in this aspect, just as you can see during this pandemic when they just add an drop services every month to correspond to the rapidly changing market demand. That trains only need to make the decision on network topology level is a weakness for the trains as it cannot dynamically adjust the infrastructure according to how demand move, and thus it wouldn’t make sense to do what airlines do, like building and operating a service to a beach resort only during the summer.
            This sort of measures, together with many other measures sheer focused on cutting cost, compares to railroad where infrastructure maintenance must be committed, mean the cost of flights can also get to a lower level than high speed rail, when the travel distance is long enough. LCC lowest fare can almost always beat high speed rail, and even more expensive fare brackets can still beat high speed trains that need four or more hours travel time, and even legacy carrier regular fare can beat high speed rail service that need more than six hours travel time.
            And no, load factors on trains are not invisible, yes you might not be able to tell the load factor of each single trains when you offer unreserved seating, but if you are running 100 trains a day and only see 1000 people entering and exiting all the relevant stations in a single day, just as during some time in the pandemic, every single person would know the load factor is extremely low and need service adjustment to lower the cost to match the lowered demand.
            In regular time, it’s unlikely to have such drastic situation, but that doesn’t mean you should not monitor when and where the passengers are going with how many of them on which train, to run your trains according to demand. Even when energy are cheaper on the trains, vehicles and staffs still cost money, and if let say you have a line which is sketched far out into exurban area where only very few people left into a big city with 30 million population, it doesn’t mean it make sense to run a train every 2 minutes into that exurban terminal, it would reasonably save you vehicles and cost of energy to terminate most of your trains somewhere midway.
            Again, it can argue that it might be better for trains to not get the optimization to that extreme level airlines are using due to the difference in nature between airlines and trains and it might be a good idea to have some empty spaces on a train, but it doesn’t mean it’s free for train companies to do this.

          • Alon Levy

            Dropping trains in and out is how SNCF timetables trains, which is a complete disaster for O&D travel that isn’t between Paris and a major provincial city. There’s a reason why France with its fast trains and high modal splits vs. air has a lot lower modal split overall than Switzerland and less than Austria and the Netherlands, all of which rely on national takt planning.

            Moreover, this isn’t an accident, the same way we can pick at a positive feature of an otherwise low-performing system. No – the airline-oriented managers came in at the TGV’s peak in the 2000s and have directly led to the system’s stagnation. The introduction of OuiGo did not raise ridership but did tank revenue, because it relied on providing worse service at worse frequency to keep the high end on mainline TGVs.

            It’s not even a “trains shouldn’t do optimization” argument. It’s an argument that trains should optimize, and that optimization on a rail network doesn’t look anything like on an airline, and bringing in business analysts whose experience is in other industries has been a disaster for anything SNCF touches. Optimization on a rail network is having trains run as fast as necessary and connect to other trains on a clockface timetable. It means not looking at load factors, because most rail costs are driven by peak service and not base service, and this includes crew (American split shifts on commuter rail are hideously costly). It means not acting like an airline that only flies planes but like a state that buys and runs trains, builds infrastructure, and makes decisions on development, and integrating all of these together. France is bad at all of this, and because it’s bad at this, it is stagnating while countries where the main trunk train routes average half the speed of Paris-Lyon or Paris-Marseille keep growing.

          • michaelrjames


            I’m sure there’s some truth in your complaints but still you seem to exaggerate.
            First, I second what Onux wrote about comparisons with Switzerland and Japan. It’s also no accident that the two other countries you compare are The Netherlands and Austria, another two quite small nations; most of the Dutch population lives in the Randstat (of about 8 million people) which is 3/4 the size of Ile de France.
            Second, while I can’t track down systematic data since 2010, it does seem to have plateaued. I don’t quite understand the Ouigo argument; it could impact revenue but why pax numbers? Remember that Ouigo and the like are a response forced by the EU directive on Open Access. Also when LGV L’Océane (Paris-Bordeaux) opened in 2017 it increased patronage on the line by 70%. This might point to the need to complete the network as long planned. Funding issues (and the debt they carry and pay interest on, unlike most other national networks) constrain them. At least it seems Macron realises the network needs action and his policy on replacing domestic flights should bring a fillip.
            Third, on another measure France runs the most successful HSR network:
            Passenger traffic on selected global high-speed rail networks in 2019, by national rail operator (in billion passenger kilometers)

            Nation…..paxkm (bn)…..paxkm/cap

            In absolute patronage, Germany is about one half that of France, despite having a 29% higher population; on a per cap basis (last column above) it is about one third. Oh, and that was in 2019 when SNCF claim they lost 8% patronage due to the strikes over the pension issue.
            No doubt there is room to improve but saying they should just do whatever Japan, the Swiss or Dutch do is a bit simplistic. What can one realistically learn from comparing travel Rotterdam-Amsterdam, the two main cities of The Netherlands and as it happens at the extreme ends of the Randstat when they are 59km apart and average train journey is 46min; versus Paris-Lyon which takes 2h but to traverse a distance about 7 times longer?

          • Alon Levy

            1. The open access mandate wasn’t exogenous to SNCF. Its own management treated trains like airlines long before then and acted as a private operator with no integration with anything else abroad, including on Eurostar and Thalys. The imitation of low-cost carriers through OuiGo was an entirely domestic process with no reference to open access. If anything, SNCF’s managerial decisions have to be understood as a cause of open access rather than a reaction to it.

            2. Looking at all railways combined shows a different picture. Switzerland is by far #1. This is despite the fact that Switzerland is small, so for the same intercity rail usage one would expect lower overall p-km/capita, and ditto Austria and the Netherlands. This includes regional rail, but the numbers for the larger countries are driven by intercity rail – INSEE has a breakdown of p-km by category and the majority of French rail p-km are TGV.

            3. What France can learn from the Netherlands and Switzerland is not to ever think in terms of “Amsterdam to Rotterdam” or “Zurich to Basel.” It’s about an everywhere-to-everywhere system. St. Gallen-Biel is a viable trip with a more or less half-hourly takt thanks to timed connections in Zurich; Amiens-Grenoble isn’t, and somehow Haute-Picardie-Grenoble manages to be even worse. Even Paris-Grenoble makes you wait 20 minutes to an hour at Part-Dieu for your connecting train, where a more competently run railway network would have all trains connect within a 10-minute window.

          • michaelrjames


            1. That is purely interpretative, ie. opinion. The EU directive began life in 1991, and the current version became law in 2012 which means it was cooking at least 5 years (or 21y) before that. I don’t know its history but it is just the thing to be created by neoliberal beancounters, probably during the ascendency of the likes of Wolfgang Schauble, and given privatisation in the UK, by the Thatcherite rat-pack. Its own description uses the term “liberalisation”. Very hard to believe it got French support. Until I read convincing evidence to the contrary I reckon Ouigo was (a) a response to continued pressure by politicians to improve financial performance and (b) pre-emption of the EU directive. Admittedly it does sound like something a LCC operator would dream up for trains … (I still haven’t heard a convincing argument as to why Ouigo would reduce overall patronage. It seems unlikely and plateauing patronage seems almost certainly to be coincidental due to other factors. But I am open to explanations even if it seems hard to prove causation.)

            2. Changing the goalposts during the game! The title of your piece is HSR ….

            3. Agreed, more or less. However, “an everywhere-to-everywhere system” is assuredly vastly easier with a small geographical area. In fact Paris-Ile de France does as well as the Dutch via RER (and Transilien) where all the line lengths are more than double the longest line length in Randstad.

          • Alon Levy

            1. You can choose not to believe it, but the Spinetta report was written by an airline exec, and the Air France-to-SNCF exec pipeline predates him. Moreover, you can see for yourself how Eurostar and Thalys are operated separately from the TGV, to the point that their tickets do not substitute for each other – if your London-to-Paris Eurostar is delayed, you can’t on the same ticket get a London-to-Lille Eurostar and change to a domestic TGV even if it’s faster than waiting for the next direct train, because Eurostar does not cross-ticket with the TGV.

            2. Sure, but HSR works as part of an intercity rail system.

            3. No? Absolute size has very little to do with this. Switzerland is just implementing what Germany did in the 1970s and 80s except a lot more systematically. Germany and France could do the same – in fact Germany is moving in that direction, slowly. Italy is likewise accreting regional rail networks on the German model, since it’s poor enough it understands it needs to Germanize instead of associate Germany with austerity (in fact, in 2020 Germany was the least austerian big European country), whereas France still has delusions of grandeur.

          • michaelrjames


            1. You are doing your trick of bait & switch again. I wasn’t denying that SNCF have had airline execs running the show but you’ve extrapolated that to conclusions about Ouigo. I was saying that it was your opinion. Perhaps it is logical but I’m not sure it is true.

            Again I am not sure either of us, or perhaps anyone, really knows the reasons for that inflexibility with Eurostar. It’s a tiny bit reminiscent of the BART, Caltrain, Amtrak, San Mateo Co and Santa Clara Co. mess at the birth of BART, all with their different agendas causing a very compromised outcome. It’s taken a long time but with the British government selling out their share of Eurostar, the now two main shareholders (SNCF & SNCB) of both Eurostar and Thalys are planning a merger to be called Green Speed. One of the aims:

            • to provide seamless travel for passengers with one ticket covering any journey across the network and better connections between trains

            Now, I suppose SNCF could have had a change of heart from what you presume, or this was their intention all along …

          • Onux

            “It matters that the highest-modal-split intercity rail networks – Japan and Switzerland’s”

            I have to point out that Japan and Switzerland also have excellent geography for rail – mountainous countries where generally everything lines up in valleys, both maximizing the utility of a mode where stopping at every point on a line hits all destinations, while minimizing the advantage of cars since they have to follow the same valley instead of taking a more direct route.

            In much the same way, most of the countries you have praised for Covid response are all islands, where it is very easy to isolate and prevent foreign spread (Taiwan, Japan, NZ, and yes S Korea since traffic over the DMZ is effectively zero).

            I am not suggesting adopting NY exceptionalism and not looking to places that do well, however “Japan has high rail ridership” and “Japan does not use yield management” even if both true do not mean that “yield management lowers rail ridership.” The high ridership could be due to a different variable, it is also possible that ridership could increase in Japan and Switzerland with yield management if it meant people could get cheap tickets by planning ahead.

          • Bobson Dugnutt

            @Phake Nick wrote:
            That trains only need to make the decision on network topology level is a weakness for the trains as it cannot dynamically adjust the infrastructure according to how demand move, and thus it wouldn’t make sense to do what airlines do, like building and operating a service to a beach resort only during the summer.
            This sort of measures … mean the cost of flights can also get to a lower level than high speed rail, when the travel distance is long enough. LCC lowest fare can almost always beat high speed rail, and even more expensive fare brackets can still beat high speed trains that need four or more hours travel time, and even legacy carrier regular fare can beat high speed rail service that need more than six hours travel time.

            I’ve boldfaced terms like “weakness” and “beat.” Transport professionals understand these terms as constraints but recognize that an underlying service has to be provided and maintained.

            Weakness and beat are terms you use when you’re playing Modal Pokemon. Cost, speed and UX would be like the hit points, defense and health of each of the Pokemon. You could describe your customer use case as the type of Pokemon, like Fire, Water, Bug, etc. Air and rail would then meet in the battle arena until one’s team is completely depleted.

            It’s a way of understanding and engaging with the world, and it’s the availability heuristic rearing its ugly head again. Only a few people are needed to have enough experience and domain knowledge to keep an airline or a railroad running. Most people don’t know or care about these details and just want to know if a vehicle is available, when and for how much. And life goes on.

            Most people, however, have watched a sport or played in some kind of game where the outcome is zero-sum (a winning and losing side, like team sports or Pokemon) or tournament (one participant prevails after outcompeting the many, like a lot of board games). It’s familiar, and often it offers participants a satisfying narrative.

            Except that for transportation, because it involves both infrastructure and burdened assets, Modal Pokemon is folly. If an airline drives a train out of the market, or vice versa, who wins? I mean, the impact on the loser is very clear. But what does the winner really win? Could it absorb the loser’s customer base without a costly service expansion? (No, because either mode tends to economize its physical property, plant and equipment and not be burdened by assets that aren’t being used to maximize revenue.)

            What is the winner supposed to do with the loser’s assets? Melt it all down and build more vehicles or buildings? (I’m being facetious.) Auction it off? (Very few people are capable of taking custody of a secondhand aircraft or locomotive and passenger cars)? Take over the remnants of the losing company and still offer service? (Alon argues that France is suffering poor rail service because its managers are trying to shoehorn aviation practices into a different mode in a different context where the rules are unworkable).

            All modes in a given corridor will have some nonzero level of use based upon the available infrastructure to support it and a path to the destination. While it would be intriguing to be able to drive or take a train from the continental U.S. to Hawaii — and Phake Nick, I’ll assume you’d know why we don’t and won’t — there are obvious physical and geometric constraints that preclude us from doing so.

            From the standpoint of Hawaiian islanders, do ships have a weakness because they cannot dynamically adjust to running outside of water? Or, necessity being the mother of invention and all, why haven’t Hawaiian islanders been able to produce innovations in aviation or aeronautics? Like, say, inventing a plane with the size and payload of a barge or building a freighter that can approach Mach 1 and maintain internal stability and/or not create tidal waves? Neither have Southern Californians or Houstonians, who also have the benefit of deep-harbor seaports and military largesse, but they’ve had more modest ambitions and yielded more practical innovations in flight.

          • anonymouse observer

            “It matters that the highest-modal-split intercity rail networks – Japan and Switzerland’s – do not use any yield management.”

            There would be another reasons why these countries succeed without yield management besides geography and turning as many seats as possible throughout the trip on the train as mentioned by Bobson Dugnutt and Onux. To stay away from the yield management practice, there needs to be good infrastructure, service design, and operating practice which could adjust the service supply to match demand (increase/reduce service or number of seats offered). Based on what Amtrak does on Acela and Pacific Surfliner, what yield management or even bucket pricing does is to control the demand so that it fits within the supply because on these services, they have next to no flexibility in terms of running extra trains or adding extra cars. For instance, you cannot add extra cars to Acela train. Number of the Surfliner trains they can run are also fixed because of track access agreements with Class Is and other track owners like LA Metro, OCTA, and NCTD. If they have flexibility, meeting the demand by offering additional cars or service would be relatively easy as railroad operations has very high fix cost.

            I am not sure the situation in Switzerland very well (do they add extra cars on busy trains? – please tell me how they do it?), but JR companies do this by building extra slots in the timetable and turn on or off these extra slots depending on the demand as well as coupling an extra trainset or cars. In order to do this, both infrastructure and timetable needs to be built/designed correctly. The way JR Central schedule trains and adjust number of trains actually operating (schedule 12 Nozomi trains but runs just enough Nozomi trains to meet the demand) is a good example of this:

            Click to access CR08050FU1.pdf

            Looking at the intercity and high-speed rail corridor network in South Korea, and compare with Japanese, I believe the network/service design is also a key. To have such flexibility, the network structure needs to be:
            – More like hub-and-spoke or something which could avoid branching and make passengers from/to secondary destinations to transfer;
            – In high demand or traffic density corridors, the intercity trains needs to operate on the dedicated track all the way to big city terminals unless there is ample track capacity to do so while sharing with others;
            – A fare structure which would not discourage or financially penalize transfers;

            Staying away from yield management scheme would not work out by itself. The ticketing would need to be more flexible in terms of changing and cancelling the reservations. Within JR service, you can change the reservations once (for paper tickets) or as many time as possible (EX-IC service) for free as long as it is regular reservation. Also, fees for second (or nth) changes or cancellations are extremely low (220 yen for each base fare ticket and surcharges for express/Shinkansen/reservations/classes/etc.) except you cancel the reservation ticket within 24 hours of departure time (30 percent of the reservation surcharge).

            The combination of customer-friendly change/cancellation policy and fixed pricing (or not adopting yield management) gives an advantage to intercity rail service over other modes in Japan. Because customers can stay flexible and know the rail operators could add capacity (trains, cars or both) when demand is high, it makes customers typically show up at the station, buy tickets, and ride the train when they want as if they are using a subway. I think this boost ridership even more by lowering the barrier (and attract more occasional rider).

          • adirondacker12800

            For instance, you cannot add extra cars to Acela train.
            They could if they had more cars but they don’t. The new trains they are testing are longer. They rent commuter trains from NJTransit or MARC during the Thanksgiving peak and very rarely other times. But they are able to do that because the commuter trains aren’t running on weekday peak schedules. There can’t be a 5:30 Acela because there is a different train using the 5:30 “slot”. And probably not a 4:30 or a 6:15 either. They need more tunnel. And eventually another set under the Hudson River along with another set under the East River.

          • Richard Mlynarik

            I am not sure the situation in Switzerland very well (do they add extra cars on busy trains? – please tell me how they do it?), but JR companies do this by building extra slots in the timetable and turn on or off these extra slots depending on the demand as well as coupling an extra trainset or cars.

            Same in Switzerland.
            “Extra slots” are mostly for intra-regional commute hours local trains.
            “Add extra cars” happens both for selected heavily-used inter-regional trains and some peak commuters. 400m long trains aren’t just for HSR.

          • Onux

            @anonymouse observer and @Richard Mlynarik

            Excellent information on demand management in Switzerland and Japan using extra cars or extra slots.

          • Phake Nick

            @Alon Levy I think it is a matter of how you schedule your trains, not fault of adjusting the schedule according to demand. Even if in low demand time period, you drop the schedule down to 1 tph, you can still have each train depart at a specific minute of every hour, instead of just randomly sending the trains out randomly. Then when demand is higher you can add in a departure 30 minutes from the set departure or 15 minutes from said departure, and thus it will still be able to maintain a pattern schedule

            @Bobson Dugnutt The word “beat” in my sentence mean competitive advantage in one single aspect (which is, fare). And “weakness” mean competitive disadvantage. Of course, it isn’t a winner take all scenario. But there are still comparative advantages and disadvantages of each modes that need to be recognized. My wording might be imprecise but I hope I can get the meaning across

        • adirondacker12800

          North Korea has a much lower automobile share than South Korea. Almost no one, in North Korea, has a weight problem either. Odd how very similar countries can be so much different. It doubt the train fare schemes have much to do with.

  21. Benjamin Turon

    Bakersfield and Fresno are I have read the largest cities in America not served directly by a Interstate Highway.,, not even a three digit connection like you see for Rochester and Schenectady NY. There is the “Golden State Highway” California State Route 99 which Caltrans has rebuilt into a freeway and are now working on upgrading to Interstate standards.

    • Onux

      Not only that, but Modesto, also in CA-99, is #3. There is always talk of resigning CA-99 into I-7 or I-9, but it is now basically at interstate standards regardless.

      Interestingly, most of the rest of the top 15 cities without an interstate are also in California along US 101, either in the south (Oxnard, Thousand Oaks, Ventura, Santa Barbara) or near the Bay Area (Salinas, Santa Rosa). Large parts of 101 are full freeway, but unlike 99 it also still has portions with intersections.

  22. anonymouse observer

    Based on what happened in Japan, I strongly believe high-speed rail itself can also work as a commuter rail for sure and potentially as a regional transit even if it runs on the dedicated track (with no track sharing with regional rail) as long as it is designed, operated, and priced to passengers right.

    For instance, people in Japan do use Shinkansen for short-distance trips which could be more suitable for regional rail like Tokyo-(Shin-)Yokohama, Shizuoka-Hamamatsu, Nagoya-Toyohashi, Kyoto-(Shin-)Osaka, Fukushima-Sendai, and Kokura-Hakata. I heard that Tokyo-(Shin-)Yokohama is so popular for commuters that JR Central set up a special rule to allow commuters boarding from Shin-Yokohama to take any unsold reserved seats on the regular class cars on any eastbound trains during the morning rush hours.

    Someone might argue that catering to the short-hoppers would eats up the track capacity which could be used for long-haul runs, but during the first couple of hours of revenue service, those long-haul trains does not show up in the segment near the other end of the line. If the end-to-end run is 3-hour long, and the first train departs from the origin at 6:00 AM, the first long-haul train arrives at the other end at 9:00 AM. On outbound direction, short-haul trains can fill the service gap in late night hours because the long-haul trains needs to depart from one end around 9:00 PM in order to arrive at the other end of the line before midnight (assuming 3-hour end-to-end trip time). You see a lot of short-haul trains like these in Japanese Shinkansen timetable:

    Click to access 210428timetable.pdf

    Because the demand for return trip are distributed in terms of timing (everyone does not necessarily go home at the same time) in both long-distance passengers and short-distance passengers, it could be absorbed if the operators can increase the capacity by running a few extra trains or coupling another trainsets during the afternoon peak hours.

    I am somewhat surprised by the fact that earlier service vision for CAHSR called for such service to use the underutilized track capacity and reduce the size of yards near big cities (potential capital cost reduction):

    • archie4oz

      I don’t think you could run much more trains, they’re already running as tight as 4 min headways on the Tokaido, (one nearly needs to look at the progression of how much operation service has shifted towards Nozomi runs over the past 20 years), and you can’t really add coupled trains since all trains are currently 16-car sets (which is pretty much what you can fit in stops).

      • Phake Nick

        On the other hand I think there are potential for more stops to be added along e.g. Hokkaido Shinkansen for some sort of local service
        Potential for more local service on Tokaido Shinkansen can be realized by the opening of Chuo Shinkansen which provide capacity relieve for the Tokaido Shinkansen

      • anonymouse observer

        I am not saying JR Central should do it. I just mention it as an example…

        Also, where did you hear they cannot add more trains? Have you checked the latest Tokaido Shinkansen timetable? They do run a few extra Kodama trains from Tokyo to Mishima during the afternoon commute peak hours while they still maintain hourly slots for 12 Nozomi trains, 2 Hikari trains, and 2 full-length Kodama trains in each direction during the afternoon commute peak hours. Looking at the timetable carefully, there appears to be a room for a couple of more short run Kodamas on top of current 14 train per hour hourly pattern if JR Central wants to do so.

        And lastly, have you seen Kodama trains with seats fully occupied for extended periods of time in the segment west of Shin-Osaka before? I have not seen it. Running Kodama trains in a way they do (16-cars with mostly non-reserved seats, once every 30 minutes throughout the day) provides the same benefit they could achieve by adding cars to the train (even if they could technically) because they do not need 16-car train for such low demand most of the time.

        • Richard Mlynarik

          Just because I find diagrams these easier to understand than reading a printed timetable, do know if there are “string diagrams” (distance/time) of any of the Shinkensen lines?

          Way back in 2007 I scraped PDF timetables and generated this 200710-Tokaido-Sanyo-Shinkansen-graph.pdf (using some code that’s bit-rotted or lost) and I found the traffic density amazing even back then.

        • Phake Nick

          According to my understanding, while it is possible to changing the number of train car for each Shinkansen trains, it is not a type of work being done commonly, and is usually done only after a train need to change the role or line it serve and is expected to continue such form of service for some years, hence you didn’t see any such change in car length even during this coronavirus pandemic.
          The upcoming N700S Shinkansen train is supposed to increase flexibility of changing number of train cars, last time I heard about it
          And then Tokaido Shinkansen also fixed the arrangement, number of train car and number of seat for all trains serving the line for the purpose of ease of management, so they aren’t going to change that according to demand either.
          West of Shin-Osaka would be San’yo

    • archie4oz

      “Moreover, the destinations within Los Angeles are not centered on Downtown; for example, high-end hotels are the most likely to be found on the Westside.”

      This really isn’t the case anymore. Yes the West side is still desirable, but DTLA is more of a destination now than it used to be, but from there you can still get to a lot of places.

      “However, there are two saving graces for trains to Los Angeles. The first is that Los Angeles’s transit ridership is so low because the city’s job geography is so decentralized that the network is bad at connecting local origins with local destinations. If it is guaranteed that one of the two points connected is Union Station, the city’s network is still bad for its size, but becomes usable.”

      More usable than people give credit for. It’s one of the reason’s I chose to live in DTLA when they re-zoned it. You had access to Downtown Long Beach, Hollywood, Pasadena, Ktown, and Santa Monica when the Expo Line opened. My aunts liked it was well since you could get there from the IE via Metrolink (you could also get to Anaheim Stadium/Disneyland area via Metrolink from there as well). You’re right about the local to local though.

    • Phake Nick

      Japan is a big counterexample to if you price the train right people would still use it as commuter rail as the high speed train cost almost double what regular train cost
      Yes there are still people who commute by Shinkansen, but it’s now used by less people than say in the 1980s when there was a big economic bubble in Japan making everything seems cheap. My understanding is that, at least for commuting into major cities like Tokyo, it is now mainly used by people whose time worth a lot of money, like executives or professionals, although in rural area where public transit are much worse there are more regular use of Shinkansen in daily life.

      As for trains that depart before first long-distance train arrive, while there are also some commuter traffic, take Kodama train 800-820 (Eastbound) as example, their main function is to carry intercity traffic from Shizuoka into Tokyo before the first train from Nagoya or beyond enter the line section. They are still high speed intercity traffic, but due to smaller size of cities in Shizuoka, they need to make all the stops along the way to gather enough demand to justify the operation of those trains

      • Matthew Hutton

        This is one of the reasons I’m less sure about TOD further out. I don’t think many people who lives 30-60 minutes by regional train from London will be poor. So you’re probably better off building good parking and executive homes near those sorts of stations than building TOD to house the poor who will work elsewhere.

        • Phake Nick

          No one say TOD must be for poor instead of rich people. But the problem is how do you market it given the image of mass transit and dense housing in America.

          • Autolycus

            You do developments like Alpharetta, GA’s Avalon. It’s not TOD, but the developer planned for a heavy rail transit stop on the edge of the development and even offered to pay for the station. Unfortunately local politicians killed the line extension in favor of “BRT” in some toll-managed lanes and inline stations that’s stalled as well.

            Avalon has a luxury hotel, a couple office buildings, a high-end outdoor mall with movie theater, fancy shops and restaurants, a luxury apartment/condo building and over 100 luxury townhouses. It, of course, also has several large parking decks. It’s not an ideal urban development as far as density is concerned, but it’s certainly been a profitable development for everyone involved, and a rail station wouldn’t have reduced the value one bit. Indeed, I think a rail station would have increased the profit on the denser parts of the development, which are also next to the potential station location.

          • Matthew Hutton

            You build luxury 2 bed apartments of maybe 70-80sqm, as well as 3-5 bed, 2-3 bath homes of 105-130sqm with gardens roughly 10-15m long and off road parking for 1-2 cars at the front. And around the station maybe have some local shops and a nice restaurant.

            And obviously you should have some denser social housing as well – but perhaps put that away from the station, perhaps next to some outdoor parkland.

          • adirondacker12800

            You put the social housing near the station and retail because they are the people who are poor because many of them have mobility problems along with their other medical problems. And more likely to walk or take the train if they are able bodied.

          • Matthew Hutton

            No one poor lives 45-60 minutes away from a major city and works there.

            The people who are poor who live in Didcot (the most successful parkway station in the UK that isn’t an airport) work in Didcot itself or perhaps in Oxford. Now if we wanted to add TOD on the Didcot to Oxford line at a “Didcot north” station, now that could be a good idea.

          • adirondacker12800

            The through running boyz think that there are people just dying to take a hour long commute to a low paid job on the other side of metro New York. When they can get a similar job nearer home. And even though there are multiple branches on either side of the through running everybody is going to have a one seat ride!

          • Bobson Dugnutt

            @autolycus, San Diego has quite a few of those developments like you are describing, and enough of them that the transit system has in-house experts to make true transit-oriented development possible.

            The most interesting case studies are: America Plaza and the City College Smart Corner in downtown, where the skyscraper and the apartment complex, respectively, are built with the trolley tracks as architectural elements. At the outer end, an auto-oriented shopping center in Santee serves as a Trolley terminal, with tracks cutting through the middle of the parking lot. From the satellite view, it looks like train passengers have a path that avoids them from having to walk through parking.


            There are two residential developments along the Green Line that built Trolley platforms to serve the complexes: Rio Vista and Fenton Parkway. The caveat is that this is in Mission Valley, a low-density linear suburb that is oriented around Interstate 8. To the Green Line’s advantage, it hits all of the important developments parallel to the freeway: San Diego State University (the Trolley’s only subway station), the former Chargers stadium, the trifecta of shopping malls and Old Town.

            As far as ridership, despite the developments, the Green Line is in the middle. The old stalwart Blue Line, the one that goes to the border, remains the most dominant line despite relatively hostile development — it has water to the west and is quite a long walk from the parallel arterial roads through National City and Chula Vista. The bus ridership parallel to the Trolley remains very high, so the train and buses serve an intensely busy corridor. The Orange Line has persistently low ridership. The suburb of Lemon Grove has TOD around its stations, but despite the relatively high poverty around the southeast San Diego stations (this is also the industrial part of San Diego), ridership is very low.

      • anonymouse observer

        Passenger kilometers among Shinkansen monthly pass was increasing before this pandemic:

        I think there are several factors supporting this gradual increase in Shinkansen monthly pass usage:
        – Subsidies by local governments where the Shinkansen station is located to retain or increase population;
        – College kids who would like to live with parents (just as preference or financially) or whose parents not allowing their kids to live in a big cities by themselves (again, just as preference, safety/security concern, or financially);
        – Commuters who want to avoid overcrowding commuter trains during peak periods;
        – Shinkansen service more frequent than parallel conventional rail service;

  23. bruce hain

    I tend to differ with producing all these high-speed rail maps of countrywide systems. It may even be counterproductive. Of course Alon has considerable experience at conceptualizing passenger lines, but if you look at that map shown in Noah’s article that gained such a big following, it’s certain for instance, that you can’t do a steady 220mph through the Colorado Rockies west of Denver in a train. For high speed rail I think I’d go around, within some reasonable directional approximation. In laying these things out there are myriad practical considerations of physical reality to be dealt with in every hundred feet of track. Get it wrong and it’s etched in stone and steel till the end of time effectively.

    Otherwise I think one central tenet of high speed rail is that it should SERVE THE POPULATIONS ALONG THE LINE. This business of planning things that tear through as remote geography as they’re able to find to in fact AVOID areas of population cannot hope to bring half the benefits that high speed passenger rail has to offer. (and they won’t get all that many passengers either) If it can’t be done with an express stop then it must be done with a convenient connecting service on the same line, or the system of transportation isn’t doing it’s job. What’s wrong with a few extra minutes with people get on and off, to savor the atmosphere of a few places if nothing more. Or be served some of the local cuisine in the diner. Railroads used to be like that. AND ALL THE TOWNS ALONG THE LINE BENEFIT. That’s my dream: to revive these sad railroad towns abandoned by their creator the railroad long ago, and really get the real estate markets there cooking again. (That kind of thing seems to be part of the theorizing in China, and it seems to work that way in Japan, and Germany.) It is (or was, and might be again) an American way of life, meeting the train at the station, and now with the advantage of high speed likely to increase a little bit more, it offers a vastly better kind of life for those lucky enough thereby to experiece the best of both worlds, by riding the train back and forth the the big city.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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.