Suburban Metros and S-Bahns

Liam O’Connell just wrote a deep dive into the history of PATH in the 1970s. I recommend people read it; as the unprofitable Hudson and Manhattan (H&M) system was transferred to Port Authority’s control, to be subsidized via the toll revenue from the Hudson bridges that had killed ridership starting in the 1930s, there were plans for expansion deep into suburbia, as far out as Plainfield. The expansion was a twofer: the H&M was unprofitable and needed change, and the same was true of mainline rail in the Northeast. Liam goes over the history of the proposal to expand service to Plainfield, and calls it an S-Bahn, comparing it to existing American examples of suburban metro like BART as well as to actual S-Bahn-type systems like the German ones bearing the name but also the Paris RER and the Tokyo subway.

In reality, there is a distinction between suburban metro service and S-Bahn service. Liam gets at one of the issues that derailed the Plainfield extension (it attempted to use high-cost capital expansion to paper over operational problems). But the distinction goes far deeper than that, and applies even to suburban metro services with a fraction of the operating costs of PATH, like BART. These are not S-Bahns, and understanding how they differ is critical.

The basic difference is that S-Bahns run on mainline rail tracks; suburban metros do not. This distinction has implications for capital planning, urban network shape, and urban growth planning. In reality it’s more complicated than that, but instead of drawing a sharp boundary, it’s better to begin by going over the core features of each of the two service types (in linguistics this is called prototypes).


The core feature of an S-Bahn is that it runs on mainline track and combines urban and suburban rail service. Every S-Bahn service I know of that bears that name or is otherwise associated with the core of the model shares track with other mainline services, but the busier ones (Berlin, Paris, Tokyo) do it only peripherally, because core lines are limited by track capacity.

The reason to use mainline track is that it’s already there, cutting construction costs. In most cases it also fits into a growth plan around existing town centers, such as the Finger Plan. Cities that build S-Bahn systems often have a surplus of industrial track serving declining manufacturing uses that can be redeveloped, for example the goods yards of historic rail terminals in European cities.

With a surplus of mainline track to use, S-Bahn systems employ extensive branching. There are more branches in the suburbs than urban trunk lines to feed them, so the system maximizes use of existing track this way. Conversely, the urban trunk lines need very high frequency to be usable as urban rail whereas the suburban branches can make do with a train every 10-20 minutes, so the branching structure generally matches frequency to both demand and passenger convenience.

Suburban metro

It is sometimes desirable to extend a metro system isolated from the mainline rail network into the suburbs. This is most commonly done when there are too few mainlines for adequate suburban service; China makes extensive use of suburban metro lines, and the commuter lines it does have are not run to S-Bahn standards (for example, the Beijing Suburban Railway is infrequent). Seoul, whose first subway line is an S-Bahn, employs greenfield suburban metros extensively as well, for example the Shin-Bundang Line.

Without an extensive system of existing lines to tap into, suburban metros necessarily cost more than S-Bahns. This means that there are fewer lines, so each line or branch has to be shorter, more frequent, and more intensively developed. Stockholm provides a ready-made example: it did not build an urban S-Bahn like the Copenhagen S-Tog, and instead built the three-line T-bana to a range of 10-20 km out of city center, with Million Program projects centered on T-bana stations.

In reality, it’s common for S-Bahn systems to also build greenfield suburban lines. For example, the RER A’s Marne-la-Vallée branch is greenfield, and does not look too different from the lines inherited from mainline rail; but it’s embedded in a mainline-compatible system, running through to legacy track on the other side of the city.

American postwar suburban rapid transit

American cities extending their urban rail networks into the suburbs ended up building suburban metros: they were never integrated with mainline rail. BART even runs on a different track gauge from the mainline network. Many of the other systems run alongside legacy lines instead of on them, at high cost. The high costs meant that there were fewer lines – the Washington Metro has complex interlining for a three-line metro, but by S-Bahn standards, it’s poor in branches.

Some of these systems had older metros to integrate with, including the Rockaways extension of the A in New York and the Green Line D Branch and the Red Line to Braintree in Boston; all three were taken over from disused commuter rail. The Braintree extension is notable in that the Old Colony Lines go much further than Braintree, but the conversion costs meant there would be no subway extension into suburbia past Braintree, and more recently the region awkwardly reopened the Old Colony Lines as low-frequency diesel commuter rail, with parts of the right-of-way encroached by the subway.

The PATH extension was to cost $402 million in 1975, or $2.2 billion today, about $80 million/km for an above-ground system that could run entirely on existing track. Newark-Elizabeth, on the Northeast Corridor, had plenty of spare capacity then and still does now – only after Gateway opens does the section need additional tracks, and parts of it are already six-track. Relative to what was required, the construction cost was extremely high. The projected two-way ridership was 28,200/day, or $78,000/rider, in an economy with less than half the average income of today.

The failure of postwar American rapid transit

Liam’s post mentions BART in the same sentence as the RER or the Tokyo subway system. This is a provocation, and Liam knows this. BART’s annual ridership before corona was not much higher than just the total number of boardings and alightings at Gare du Nord. The Bay Area’s modal split is comparable to that of provincial French metro areas like Marseille and Toulouse, with an urban light metro or light rail system and thoroughly auto-oriented character outside the historic core. So what gives?

This isn’t quite a shortcoming of the suburban metro model. Stockholm uses it, and so does all of China. Rather, it’s a combination of several problems.

  1. The suburban metro model requires extensive transit-oriented development to compensate for the narrower reach of the system. Stockholm built Vällingby and countless other suburbs on top of the T-bana. Washington built a handful of TOD centers like Arlington and Bethesda, and the other American examples built nothing, preferring parking lots and garages at stations.
  2. American construction costs were too high even then. The cost of the proposed PATH extension was $2.2 billion for 27 km on existing above-ground right-of-way. The actually-built Washington Metro cost $9.3 billion in current dollars by 2001, around $25 billion in today’s money, for a 166 km system of which 72 are underground. In contrast, the T-bana cost, in today’s PPP money, around $3.6 billion for 104 km of which 57 are underground, around one fifth the per-km cost of WMATA. As a result, not much was built, and in many cases what has been built follows freeway medians to economize, leading to further ridership shortfalls.
  3. BART specifically suffers from poor urban service. As pointed out more than 15 years ago by Christof Spieler, it has very little service in San Francisco outside city center; Oakland service is awkward too, with most residential areas on a separate branch from Downtown Oakland. The Washington Metro has done this better.
  4. The A train in New York has the opposite problem as BART: the Rockaways tail was tacked on so awkwardly, at the end of a line that runs express but is still not fast enough – Far Rockaway-Times Square takes 1:08-1:10 for a distance of 37 km. The Green Line D Branch takes 46 minutes peak, 40 off-peak to traverse 19 km from Riverside to Government Center. PATH to Plainfield would likely have had the same problem; the core system is not fast, and with no through-service beyond its Manhattan terminals, it would have had cumbersome transfers for onward travel.


There are two models for how to extend rapid transit into the suburbs: the commuter rail model of the S-Bahn systems, Tokyo, and the RER, and the suburban metro model of Stockholm and China; Seoul uses the S-Bahn model where legacy lines exist and the suburban metro model otherwise. The segregation of mainline rail from all other forms of mass transit forced postwar America to select the latter model.

But implementation fell short. Construction costs were far too high even in the 1970s. Transit-oriented development ranged from mediocre in Washington to nonexistent elsewhere; the systems were built to interact with cars, not buses or streetcars or subways or commuter rail. And most of the lines failed at the basic feature of providing good urban and suburban service on the same system – they either were too slow through the city or didn’t make enough city stops.

Moreover, much of this failure has to be viewed in light of the distinction between S-Bahns and suburban metro systems. S-Bahns had better turn their outlying stations into nodes with bus service (timed with the train unless frequency is very high) and local retail, but Berlin is full of park-and-rides and underdeveloped stations and suburban Zurich is low-density. In contrast, suburban metros have to have the TOD intensity of Stockholm or suburban Seoul – their construction costs are higher, so they must be designed around higher ridership to compensate. This should have been especially paramount in the high-cost American context. But it wasn’t, so ridership is low relative to cost, and expansion is slow.


  1. Sassy

    What is really the difference between a greenfield S-Bahn Suburban Section and a Suburban Metro? Considering the wide variety of things called “metro” and “mainline” saying that a Suburban Metro doesn’t look like mainline, is a bit vague/unsatisfying/subjective.

    Why can’t Korail build a line to their “mainline” standards and run trains into the Shinbundang Line? The trains are physically like the Korail S-Bahn service trains, the electrification is the same but trains can support both if mismatch, the signalling is different, but again the trains can support both.

    • Eric2

      “What is really the difference between a greenfield S-Bahn Suburban Section and a Suburban Metro?”

      It seems to me the key distinguishing factor in Alon’s S-Bahn vs metro distinction is development density.

      Suburban metros must be surrounded by very high density development to justify their high construction costs. Except in the largest cities, this means suburban metros must be short because the market can only support so much high density development.

      S-Bahns can be surrounded by lower density because their construction costs are lower. A greenfield S-Bahn branch can be financially justified if it piggybacks on a highly useful trunk line which was already built, and if it’s built on the surface rather than tunneled to keep costs down. Both of these factors are present with the RER A, which also has relatively dense development around it, more like a metro.

    • Alon Levy

      Good question!

      So, in Paris, the MLV branch of the RER A is very intensively developed, like a suburban metro, to pay back the higher construction costs. In very large cities like Paris, Seoul, and Tokyo, the distinction blurs, as they are so large that even suburban metros can have extensive coverage – Paris is even building an orbital suburban metro. In Paris there’s still a technological distinction – MLV is mainline-ish rail, GPE is automated suburban metro – but the development intensity is high either way.

      This means it’s better to look at the distinction in smaller metro areas, where a suburban metro is necessarily much smaller. I like Stockholm vs. Copenhagen and Stockholm vs. Zurich as case studies; neither is a perfectly clean comparison, Copenhagen because of cycling and Zurich because it has Swiss levels of TOD and not Nordic ones (cf. the Finger Plan), but I think both combined provide an interesting contrast to the T-bana.

  2. Phake Nick

    Tsukuba Express in Tokyo is more like suburban rail model.
    It terminating at Akihabara to the south is relatively limited in connectivity, how much can it improve if it is to be further extended south toward Tokyo Station and Tokyo bayside area as according to plan?
    Toward the northern end of the line it do not connect with anything either. There are recently suggestions that extend it to either connect to Joban line at the nearest station, or do a greenfield extension all the way to the preferential capital of Mito city and covering all the area that have no rail service now, which would be a more beneficial option?

    • Alon Levy

      Yes! This is an excellent example of the suburban metro model. To that I’ll also add some tails built specifically to extend some subway lines, like Toyo extending Tozai and the Saitama Railway extending Namboku.

      When I’ve tried to crayon Tokyo, I tried to find things west or southwest to connect Tsukuba with, via expensive tunnels serving Tokyo or Otemachi. Odakyu is probably the best choice, via a four-track tunnel between Shinjuku and Tokyo, two tracks for Tsukuba-Odakyu and two for Keiyo-Seibu-Shinjuku.

      • Phake Nick

        Keiyo line to Shinjuku already have reserved right of way from failed Narita Shinkansen construction proposal, and the current plan anticipate the line to further extend toward Mitaka then Tachikawa to alleviate Chuo line’s demand. But with the current situation of Tokyo urban rail, where even peak hours train have to cut frequency, and then the population forecast, there doesn’t appears to have a need for such project by year 2100.

        As for Odakyu, they already through run to JR Joban line through Chiyoda line. Connecting Odakyu to Tsubasa Express through a tunnel from Shinjuku to Tokyo offer limited benefits as most key destination along such proposed through running service will already be relatively easily accessible by interchange from stations along Chiyoda and Joban line.

        • Alon Levy

          The Chuo Line is underfull going from the west – the capacity crunch is on the Sobu Line side.

          Odakyu is a four-track line with only two tracks’ worth of through-service, is the issue.

          • Phake Nick

            All JR lines in Tokyo are cutting capacity. The latest schedule adjustment in 2022 March have permanently dropped frequency on Sobu line rapid from 19 to 18 tph on morning rush hour, as crowdness down from 181% to 105% in FY2020, and also dropping the frequency on Sobu line local westbound from 26 to 25 tph on morning rush hour, as crowdness down from 194% to 111% in FY2020.
            Chuo line rapid’s busiest segment going east was from Nakano to Shinjuku, which saw its peak hour crowdness dropping from 184% to 116% in FY2020.
            Ridership probably recovered a bit in FY2021 but the data still haven’t been published, but is still not a full recovery and it is not expected to fully recover, hence the cut in frequency across all lines.

            Furthermore, upcoming Yurakucho line extension which bridge with Hanzomon line is expected to be able to divert some passengers from Kinshicho too.

            If Odakyu need more through service, then I would imagine going toward Ikebukuro and Saitama direction is probably more useful? Although I am not sure what it can connect to in that direction.

          • Eric2

            So that’s it? Japan will never invest in more urban transit, because their working population is in a long-term steady decline so commute volumes will never again reach the levels they were in the recent past?

            I guess the Chuo Shinkansen is still justified because old people go on long-distance trips (vacations and family visits) too?

          • Phake Nick

            There will still be localized improvement and finishing up missing links, like Yurakucho line extension that link up eastern part of Tokyo vertically, three different airport related rail through running or connection improvement project as air traffic is still projected to grow, a new subway branch to Shinagawa due to it being a business hub and transportation hub that is undergoing development, but that is about it inside Tokyo. Some new projects will be build in some other cities too but those are mostly of similar nature to this. Note that Tokyo city government is planning to fund the Shinagawa extension and Yurakucho line extension by selling half of their stake in the entire Tokyo Metro network, while Japanese national government also plan to sell half their stake to fund recovery and revitalization of Tohoku area after the earthquake and tsunami from last decade. Tokyo government have also established a fund around like 2018 to collect excess budget and hope that can be use to fund transit project into future, but so far that is only pocket amount.
            Nationwide, there are some other projects that might see investment, to further equality in development across the country, like the proposed Okinawa railway, where the Japanese government have special political responsibility over the island, despite they still cannot work out how to make the benefit of such project greater than investment even with societal benefit included. Possible Shinkansen route to Shikoku and along the Sea of Japan coast of Tohoku could also be conducted in similar way, although the Uetsu Main Line along Sea of Japan coast of Tohoku have such a low ridership now that it is unclear whether the line can sustain itself into the future at all not to mention Shinkansen construction, and Shikoku Shinkansen plan would involve large amount of investments on tunnels and/or bridges that are hard to justify, and there are proposal on single tracking instead of double tracking both lines they would actually build.

            As for Chuo Shinkansen was started some while ago, and is privately constructed, thus whether will it or how much benefits will it bring to the country isn’t really important, but the focus is how much can JR Central earn from it in the long term. The project is going to be a cash flow burden for JR Central in upcoming decades but probably they see something net positive from it longer terms.

            In reality, other than the Chuo Shinkansen, what is being left to do is probably Hokuriku Shinkansen extension to Osaka, West Kyushu Shinkansen connection to rest of the network, and minor speed up and operation stability improvement through new tunnel and gradual removal of level crossing of the two Mini Shinkansen lines. Something could possibly be done about Hokuriku Shinkansen connection to Nagoya too I guess. But I think that this is the end of it. Most regional Japanese cities are projected to see roughly two third of their current population gone by the year of 2060, that even the two area without Shinkansen service as I mentioned above will be hard to justify investment.

          • Alon Levy

            In Okinawa, shouldn’t the US fund this out of special obligation? It could plausibly displace the Battersea extension as the most cost-effective public transport projected funded by the US federal government…

          • Phake Nick

            It would probably make political sense for the US to help fund such project, especially to snoothen the relocation of the military base to Northern Okinawa both in term of local residents support and the logistic of US military themselves, but probably not in financial sense. First of all, Japan is now paying the US for the bases US have in the country, and the payment have increased in the past few years as the US president at the time demanded countries paying their fair share for the hosting of US military, thus getting US money back to Okinawa is just spending Japanese government money with an extra step and overhead of US administrative processes.
            As for cost effectiveness, Japan still cannot figuring out how to make the project have a benefit/cost ratio over 1…. Even after considering options like reduced loading gauge size, single tracking low ridership sections, simplifying station structure, the B/C for guideway transit option and Linimo-style maglev are still only 0.72/0.73, with conventional railway being 0.66, based on the requirement of able to climb 6% gradient and having max speed of 100+ km/h in order to connect northern part and southern part of the island within an hour. Tram train proposal have also been studied but its B/C is still less than 1 at 0.89, need more than 2 hours for whole trip, and would affect road users, with roads on Okinawa being nowhere near American roads’ width and are also limited in quantity due to many spaces taken up by military bases.

          • Alon Levy

            All of these numbers are unusually low – is that corona? Because pre-corona I saw numbers scratching 200%…

          • Phake Nick

            Yes. my earlier reply quoted pre-corona number to be 180-190% on all three of them. According to Japanese media, they describe the cause as people having more flexible commute hours. So while the number of riders on peak hour have reduced, trains before or after traditional peak hours are becoming about as packed as traditional peak hour.
            If what the media observed and described is true, then this would mean the transit ridership is still there just shifted to different time, but that would still mean no reason to construct anything if their goal is to transport more people in a single hour.
            This year (2021 April – 2022 March) report also put Nagoya as having more crowded train than Tokyo, with Tokyo overall crowdness at 108%, Osaka 104%, and Nagoya 110%. Maybe a reason behind this was that Tokyo mayor was the one who originally proposed more varied commute time as a response to coronavirus pandemic with the aim of reducing train crowdness? And I am not sure what she would feel now that her almost impossible election promise from last decade of “No more packed train in Tokyo” is now realized in this form.
            The most packed commuting train in Japan according to latest report is Nippori Toneri Liner, a 5-car guideway transit in Tokyo at 144% crowdness, then the second most crowded is Nishitetsu Kaizuka Line, a 2-car single-track conventional railway in Fukuoka, third is JR Musashino Line segment in Saitama at 137%, and fourth are JR Saikyo line at Tokyo and JR Kabe line at Hiroshima at 132%.

            By the way I forgot about one thing, that is it have been in the work for quite a while now that Chuo Line Rapid is now lengthening its platforms to accomodate two extra premium cars that will be in double deck, which was originally a way to increase capacity but also anticipated to be a way to increase revenue by making people who want a seat to pay. So there will be even more spare capacity on the line in the future during peak hour. But one thing is they cited “semiconductor shortage” as reason of the premium service EIS delaying from the originally anticipated 2022 to end of FY 2024, aka 2025 March. Maybe they aren’t in a rush to introduce more capacity now that there are spare capacity.

          • Richard Mlynarik

            Fake Nick: it’s always really informative to read your comments. So much information that it otherwise inaccessible to me. Thank you for all the effort you put into educating us here.

  3. Martin

    Taipei is also a very good example on a very subway centred system. The mainline tracks were/are run by the national rail company, and still serve urban transit needs very poorly (minor improvements the last few years). Not much TOD, just universal high density.

    Rest of Taiwan, is much like pre-1990s Taipei, where scooters basically substitute transit.

  4. IAN! Mitchell

    Would PATH to plainfield have pencilled out any better if the PATH-Lex connection was actually made?

    Obviously you’d have very few Pelham Bay Park-Plainfield riders.

    • adirondacker12800

      The IRT made a serious mistake and made the local platforms short. They lengthened them in the 50s. IRT trains are too long to run on the PATH system. It’s unclear, to me from little there is on the internet, if they were plotting a terminal at Astor Place or a connection. And the demand was and is for downtown, not uptown or the East Side.

      • Henry

        I think they are talking about the post-9/11 proposal for a WTC-City Hall connection.

        It gives PATH a direct connection to Midtown East (though one that is scraping the walls) though you still have the 10-car problem, although PATH is currently studying turning Newark-WTC into a ten-car line at least. The issue though would be then where would trains from Hoboken or 33 St go?

      • adirondacker12800

        Why wouldn’t the trains from 33rd Street continue to go to Journal Square or Hoboken?
        They went and built skyscrapers all over the place and any connection that may have been possible in 1908 has skyscrapers in the way. Including walls around the World Trade Center site that stops the tide from coming in twice a day. Along making the loop broader, deeper and moving it slightly to the west.

  5. adirondacker12800

    Liam O’Connell just wrote a deep dive into the history of PATH in the 1970s.

    Skimmed the surface.

    PATH to Plainfield would likely have had the same problem; the core system is not fast, and with no through-service beyond its Manhattan terminals, it would have had cumbersome transfers for onward travel.

    Yet you want to plop commuter rail stations all over Hudson County, Brooklyn and Queens and the Bronx.

    People from the Raritan Valley line use PATH today, they change trains at Penn Station Newark. Changing trains at Penn Station Newark is faster than it was to change trains in Elizabeth or wandering through Bayonne to get to the ferry in Jersey City.

    People changing from outbound/westbound PATH via the ramp from the PATH platform to the island platform between Tracks 3 and 4, in Newark.

    Repeat every three to five minutes during the afternoon rush hours.

    • Alon Levy

      Yet you want to plop commuter rail stations all over Hudson County, Brooklyn and Queens and the Bronx.

      Yes. Because the extra few stations are not enough to turn fast service into slow service. Contrast this with the A train to the Rockaways: the stops from Rockaway Boulevard to Manhattan go 88th, 80th, Grant, Euclid, Broadway Junction, Utica, Nostrand, Hoyt-Schermerhorn, Jay, High, Fulton. It’s not that awful – it’s 11 stops in 16 km – but it’s supposed to be express and isn’t, and the top speed is meh, and the curves are wide by subway standards and sharp by mainline ones (the IND was built to 90 m radius and I don’t think any of those is canted), and there are further slowdowns for turnouts and mergers. So the average speed is weak.

      Compare this with what I’ve crayoned for what I call Line 5, a.k.a. the Atlantic Branch. The line stays express for longer, so the stops from Jamaica go Lefferts, Woodhaven, Cypress Hills, East New York, Nostrand, Flatbush, Borough Hall, Fulton. It’s 8 stops in 19 km, or 9 stops if you add infill at Utica, and everything is built to mainline standards so that even the curvy bits let you do 80 km/h in Manhattan and 100 km/h farther out.

      • Matthew Hutton

        11 stops in 16km compares unfavourably with (say) Shanghai metro line 2 which has 10 stops from Hongqiao railway station to peoples square over 17.75km. And Shanghai metro line two takes 31 minutes to do that journey.

        • Alon Levy

          Shanghai is pretty slow, too, and railfans on SkyscraperCity compare it unfavorably with Tokyo, where the express JR East lines average 50-60 km/h to Tokyo Station.

      • adirondacker12800

        There are parallel subway lines, change trains.

        Newark>Harrison>Journal Square>Grove Street>Exchange Place>World Trade Center is six stops. Jamaica>Lefferts>Woodhaven>Cypress Hills>East New York/Broadway Junction>Utica>Nostrand>Flatbush>Downtown Brooklyn>World Trade Center seems to be ten stops to me.

        100 kph is too slow. All three commuter railroads have long stretches of 80mph/125kph a few places where it’s 90/145 and has been higher in the past and there are videos of the smartphone tracking the speed of the NJTransit train, from a passenger seat, on the local track, at over 100/160.

        • Alon Levy

          The current maximum speed on the LIRR in Brooklyn is 45 mph most of the way; it speeds up to 70 near the Queens County line and doesn’t hit 80 until past Jamaica.

          • adirondacker12800

            If the train is stopping as frequently as the express subway two blocks away it’s not going to be able to get to high speeds. There is a subway with local and express service two blocks away, change trains.

          • Martin

            What prevents higher speeds on LIRR in Brooklyn? Caltrain diesels coming out of SF quickly hit 79mph as soon as track geometry allows it. B

      • Sean Cunneen

        Why are you adding the Borough Hall Stop? Can’t people who work there just transfer to the 2/3 or 4/5? Is it for a connection to the A/C & F/M at the Jay St Metrotech stop?

          • adirondacker12800

            take the 2,3,4,5,B,D,N,Q or R.
            To go west from Flatbush Ave. and Atlantic Ave. the choices are to continue on Atlantic to the harbor or go up Flatbush. Anything else would be very complicated and very expensive.

          • adirondacker12800

            Excavating deep cavern stations isn’t cheap. Change trains

  6. Robert Campbell

    I think PATCO comes as close to an S-bahn as exists in the U.S. It was primarily built on the PRSL right of way east of downtown Camden, has recent headways, which formally had intentions to be a 3-legged system.

    • adirondacker12800

      It runs 24 hours a day. Most places in the world the railroad curls up and goes to sleep around midnight

        • Henry Miller

          If you don’t run 24×7 that means that at some point in their life everyone will be unable to get where they are going on your system. It is a rare thing, but at some point in their life everyone will need to be out in the wee hours of the day. red eye flight, bar closing (staff does clean up after this!), working late. If you can’t serve this market everyone needs to own a car “just in case”. Once someone has a car for even 1% of their trips it is cheap and convenient to use it for everything.

          It need not be a train of course. Half hour bus service with timed transfers is okay. It just has to work for those rare and unprofitable trips to really get good mode share.

          • Robert Campbell

            I think New York and Philadelphia run all night bus service on their standard routes. Seattle, Chicago, and San Francisco run Night Bus routes. Nothing in Washington DC or Boston.

          • Henry Miller

            Night bus might be international standard, but it isn’t US standard. We are lucky to get Sunday service, Saturday we get reduced service, and the bus service ends early in the evening even when it is running. Combine that with slow winding routes and it is surprising when you get even 2% mode share.

    • Onux

      The closest thing to an S-Bahn in the U.S. is indeed found in Philadelphia, but it’s not PATCO, it’s SEPTA Regional Rail. As designed by R. Damon Child’s and Vukan Vuchic, SEPTA service through the Center City Commuter Connector perfectly replicates S-Bahn typology, with thirteen branches combining in one trunk tunnel serving two underground stations downtown. Implementation is a bit messy – due to legacy right of way (Pennsylvania RR vs Reading RR) some lines from the west enter the tunnel from the east and vice versa – but the overall design is unmistakable.

      But after a quarter century of the kind of service Alon advocates for Boston with the North-South Link, or that New Yorkers dream about with the lost ARC Alt G (Philly had through running a decade before RER D in Paris!) SEPTA in 2010 rebranded from through lines (R1, R2, etc.) to thirteen named branches as if all service terminates in the CBD. Ironically (tragically?) the trains still through run but with no indication for passengers that you can get a one seat ride across the city.

      • Eric2

        A one-seat ride through the city isn’t very useful. All the branches go to the north and west (the Pennsylvania side of the river). So going from one branch to another branch generally means an unnecessary trip to the city center, when a radial bus trip is much shorter.

        I have this fantasy of merging PATCO and some other New Jersey ROWs into SEPTA regional rail as additional branches, so that the branches do cover the entire metro area and regional rail through the center is more useful, but this isn’t on anyone’s priority list…

        • Onux

          A one-seat ride through the city seems to be very useful in almost every other city that has it like Berlin.

          Conshohoken to Malvern might be faster by direct bus, but I can’t imagine the same is true for the airport to Willow Grove or Springfield to Langhorne. Those examples are pretty close to straight lines, since some branches go south as much as west or east as much as north.

          Is the key flaw of SEPTA the fact that it’s east-most line (Trenton) enters from the west? Trenton should logically be paired with Media or Thorndale, but without a reverse Swampoodle connection switching Trenton trains from the NEC to the old Reading main line, they meet at 30 St instead.

          • adirondacker12800

            North Philadelphia is over the street and North Broad is in a trench under the street. Perpendicular to each other. You would have to tear down wide swaths of North Philadelphia to build ramps to send Trenton Line trains to North Broad. You could send trains from Trenton to Thorndale today if you wanted to. They wouldn’t go to 30th Street but you could.

          • Eric2

            I think more to the point is that there are ~7 east branches and ~7 west branches. So if your origin is a random east branch and your destination a random west branch, 6/7 times you will have to transfer no matter what. As for the 1/7 cases, what’s at stake is a 7.5 minute average wait time (assuming 15 minute headways on the branches) plus a minimal time to transfer between platforms, none of which sounds like a killer.

            And regarding Trenton specifically, it appears that if the North Philadelphia/North Broad transfer is improved then the Trenton branch’s backwardness becomes an advantage rather than a disadvantage, as it allows anyone in the northern suburbs to transfer there rather than going in and out of downtown.

            By the way, thinking of your examples – airport and Trenton seem to be unique cases in that there is substantial demand from any random suburb to these stations. Whereas there is little demand from any random suburb to any random other suburb. I’m not sure if this has any particular implications for route design though.

          • adirondacker12800

            There are a lot of state offices in Trenton. New Jerseyans can find them quite interesting. Pennsylvanians, not so much.

        • adirondacker12800

          Three of them go to the south. Elwyn, Newark and Airport. The schedule shows which trains run past the downtown terminal and the destination.

          • Eric2

            Those lines go to the west quadrant, not the south quadrant. You can draw a line through Philadelphia City Hall, and all but a handful of Regional Rail stops are on the northwest side of that line. Nearly 180 degrees worth of suburbs are unserved.

          • adirondacker12800

            To get quadrants you have to draw two perpendicular lines. If you only draw one line it’s halves.
            There is a big chunk of land southeast-ish of Philadelphia with 40 acre zoning. It’s always going to have 40 acre zoning because it’s sitting on top of trillions of liters of very clean fresh water. Real life has things like that, in it. Like it’s hilly on the Pennsylvania side. There’s a big estuary called the Delaware River in the middle. Pity that it’s not Sim City.

          • Eric2

            Maybe the aquifer you mention is the reason why there’s no SEPTA branches to the south and east. So be it. What matters for our purposes is simply that no such branches exist.

            Though, I do spot a bunch of rail ROWs on the New Jersey side that could be made into SEPTA branches. Maybe the aquifer means there will never be enough density to make such extensions financially viable. Maybe not.

          • adirondacker12800

            It’s the same branches they came up with in the 60s, 80s, 90s and again in this round. PATCO centric because the bridge doesn’t connect to the commuter system, it connects to the subway.

          • Eric2

            The bridge comes very close to the commuter system and connecting them would only require a tiny segment of new track.

          • adirondacker12800

            Through the Broad Street Spur and westbound, across two tracks of opposite direction traffic. Unless you want to take both out of service for a few years while they rearrange things.

          • Eric2

            I really don’t understand why the Broad Street Spur still exists, I would just cancel it. (Maybe build a Chinatown station on the tracks coming off the bridge to replace the current Chinatown station.)

            Anyway, yes it is possible to build an underpass under Regional Rail without shutting Regional Rail down for years. It’s expensive, but if that’s the only expensive thing in the whole project then it should be affordable.

            BTW OpenStreetMap has the Regional Rail on level “-2”, I’m not sure what’s on level -1 but maybe there is space to go over the current Regional Rail tracks rather than under them, which would make construction a lot easier. I would still recommend cancelling the Broad Street Spur though – no reverse branching allowed in my world.

        • Henry

          I mean, the low hanging fruit is to make a connection at North Broad and North Philadelphia that isn’t a aboveground pain in the ass.

          • Eric2

            I agree. Move the North Broad station platforms north so that their north end is adjacent to the North Philadelphia station and tracks.

          • Onux

            Absolutely this. Both sets of platforms should have direct access to the N. Phila. station on the Broad St. Metro line as well, even though it would involve some digging.

      • Onux

        Important question for Alon: why didn’t SEPTA Commuter service take off after CCCC opened in 1984? It’s a center city tunnel linking two fully electrified networks, Philly is one of the big six transit cities in the US (big by US standards that is), and the legacy networks are highly branched and serve existing, pre-automotive town centers. This checks off all of the points you say would make similar service so successful in your Boston and New York plans; why didn’t SEPTA outperform in ridership (more accurately riders per mile, since larger metros like NY and Chicago might still have greater total ridership).

        • Henry

          There were supposed to be frequency improvements as well, but that largely didn’t pan out, and SEPTA spent much of that time needlessly fighting unions. (Not only did they want to scrap conductors, which is painful but necessary, but to add insult to injury they also wanted to *lower* everybody’s pay, which certainly did not help labor relations)

        • adirondacker12800

          Because there isn’t a whole in some distant suburb that I can’t already get in the one I’m in?

        • Alon Levy

          To add to what Henry said: SEPTA didn’t do any of the operational treatments that S-Bahns have. The frequency is low, and there’s no integration whatsoever with urban transit, so the in-city lines have weak ridership while parallel bus lines like the 23 are the busiest in the system. SEPTA Regional Rail is isomorphic mimicry of an S-Bahn without the actual service, which Vukan Vuchic periodically still criticizes in the media.

          • Onux

            This then raises a different question Alon: what do you think is the US system most like an S-Bahn?

            Also, if service quality is so important, you should carefully consider how you frame or advocate plans such as NSL. In this very post you state: “The core feature of an S-Bahn is that it runs on mainline track and combines urban and suburban rail service.” and “S-Bahn systems employ extensive branching”. These are very definite physical features that SEPTA doesn’t just mimic but meets. If that isn’t actually the definition of an S-Bahn you need to say so: service quality or fare integration can’t be seen, but tunnel and tracks can, and it would be very easy to convince people to oppose NSL because it looks exactly like CCCC but SEPTA didn’t get good results.

          • Alon Levy

            I don’t think there really is one. SEPTA went the furthest in deliberate isomorphic mimicry of the concept – as you note, it meets the narrow definition – but nothing else there works as in an S-Bahn. This is why I define S-Bahns by prototypes, not by boundary-based definitions: normally nobody coming up with such definitions thinks about frequency, urban infill, or fare integration, because they’re obviously part of any urban rail system, but they’re absent on SEPTA RR.

      • Onux

        Quick add, SEPTA also meets another of Alon’s criteria for S-Bahns, use of mainline track. Some of the branches share track/stations with Amtrak NEC and Keystone services, in addition to commuter only branches.

      • adirondacker12800

        Alt-G had fatal flaws. And was going to be a third terminal at Grand Central that didn’t connect to anything else.

    • Luke

      I would think somewhere closer to an intercity line. Although local interests keep wanting to add stops in the suburbs (kind of a reverse S-bahn, if they do), there are only about as many stops within Seoul as there are in the various cities outside which will be served by the different GTX lines. Given the planned accelerative capabilities of the trains, the small difference in stop spacing compared to the average on, e.g., GTX A (between 4-11km) means an expected average speed of 100km/h, whereas the <3km stop spacing that the Bundang Line paralleling it south of the Han means that the trains are not even designed to much exceed 100km/h.

      Additionally, especially along the GTX A, there is technically a lot of track sharing with "mainline" (high-speed) rail through the Yulhyeon tunnel, and AFAIK, there is a lot of planned track sharing outside of Seoul proper for lines A and C, as well.

      I guess where the GTX fits in this binary typology depends on whether you consider service and performance or infrastructure and correspondant development the more significant measurement.

  7. Lee

    American systems like BART never really attempted to compete with cars nor did American politicians ever consider anti-car policies. At best they were built to make car use only partial rather than total but for most people driving all the way to home and your destination in a car made sense rather than having to do a two or three stage trip of car to rapid transit and maybe bus afterwards.

    • Henry Miller

      It is very hard for transit to compete with a car you own.

      Cars are just to convenient. They are always sitting there waiting for you to be ready to go. No checking an app to see when the next bus is, or sitting at a stop/station for “a long time”, just go start your car and go. You never miss your car like you can miss the bus. The car is also point to point, no transfers (which your bus or might or might not be late to), no sitting around for “a long time” at some station you don’t want to be at. (Note that I keep saying in comments speed and frequency is important)

      Most of the cost of a car are fixed, you still have make payments, pay license fees, and insurance even if the car sits in the driveway while you use transit. Gas is for most people only a small part of the cost of a car, even at today’s high prices the fixed costs are the majority of your cost. (this is also why most people have a large SUV – just a small car could do 99% of their trips, but having a second car just for that 1% doesn’t make sense as they don’t drive nearly enough for the variable costs to overtake the fixed cost)

      competing with cars requires either expensive parking (you just threw out the suburbs which are important) or bad traffic (not in suburbs, and people vote for more roads despite the high cost) so long as you don’t have a system that is good enough to get rid of the car. Once you have system useful enough to get rid of a car completely it is a lot more compelling, that means more riders, and more less traffic. Note that I said a car not all cars – this gets rid of a lot of objections people have about moving a couch or vacation, since family tend to own several cars.

      • adirondacker12800

        People buy big vehicles so they can brag about how much money they have without moving their lips.

      • Matthew Hutton

        Cars aren’t a magic wand.

        If you drive more than ~10k miles a year cost per marginal mile is very high, something like US$0.50 a mile. And especially with second cars there are extremely high up front costs of ownership – it’s difficult to spend less than a few thousand dollars a year on car ownership.

        Cars also provide a significantly worse experience over distances of longer than 300 miles/500km or so compared to even a decent classic express train (although flying can also be faster/cheaper over those distances as well).

        • Henry Miller

          > If you drive more than ~10k miles a year cost per marginal mile is very high, something like US$0.50 a mile

          You have it backward, The more you drive the lower your costs per mile. If you drive 25k miles per year your costs drop closer to about $.30 (at today’s high gas prices). Not that it matters, cost per mile is a stupid measure. It is only useful if you are charging someone for use of your car, which doesn’t to most people.

          >Cars also provide a significantly worse experience over distances of longer than 300 miles/500km or so compared to even a decent classic express train

          No, cars are a much better experience. Sure if your destination isn’t car dependent a train is nice, but if you need a car once you get there you may as well drive and have your own car when you get there. This is the reality for most Americans – we don’t have good options when we get there so a car is much better. Even if the distance is 1200 miles (well outside of train range) a car is competitive with flying because you avoid the airport and expensive rental cars.

          • Eric2

            “if you need a car once you get there”

            That’s what it boils down to. If you don’t need a car when you get there, rail is much preferable to driving. (Unless you are moving house or something and the car is carrying your goods)

          • adirondacker12800

            People who think cost per mile is a stupid measure are cooking the books.

          • Matthew Hutton

            If a car costs $30k dollars on the road and lasts 175k miles then the depreciation alone is $0.17/mile. And beyond that on top of fuel there’s insurance, vehicle taxes, tyres, brakes, servicing, plus any repairs – all of which get done on a per-mile basis once your mileage gets above a certain point.

            I mean look I don’t live in a big city – so I own a car of course. But on a rational basis as a family one would need to think long and hard about owning a second car.

          • adirondacker12800

            Depreciation isn’t lineal. Signing the paperwork for a new car costs thousands in depreciation. If you plan on owning it for 175k you have to spread out that depreciation over years. Most people won’t do that. Someone else eats the cost if you buy used instead of new.
            Adding up fuel costs, state requirements, insurance, regular maintenance and repairs is hard and most people don’t do it. I drive from here to Washington D.C. it’s two percent of a new set of tires and a quarter of an oil change. The kind of thing most people don’t account for. ….. People who think cost per mile is stupid are cooking the books.

          • Matthew Hutton

            I guess cars depreciate more at the beginning, but they also need fewer repairs, so I guess overall it all roughly balances out ☺️.

          • Henry Miller

            People who account for cost per mile are cooking the books a lot more than those who don’t. If you lease a vehicle for 30,000 miles at $500/month (very common terms, though my numbers are rounder than anything real) that is already $1.60/mile before we account for insurance and license fees.

            Yes older cars need more repairs, but as those who actually drive older cars know, repairs are not nearly what a payment is. As such an older car is far cheaper. Well assuming you don’t restore it to like new. Almost everyone with an older car accepts seat and paint wear, and other general things that make it less desirable than a new car: so long as it gets them down the road.

            That is why cost per mile is stupid: it is far too variable to have any meaningful conversations about. The people who really care about cost have a much lower cost per mile figure than anyone is quoting here. If you want transit to compete on cost you need to figure closer to $.20/mile as that is realistic for people who care about cost – note that is door to door miles, so if your transit system doesn’t go door to door add in cost of uber for the last mile (or whatever).

          • adirondacker12800

            $4.00 a gallon at 40 mpg is ten cents a mile. The insurance is five cents a mile. Consumables are are two cents a mile. Registration and inspection is penny a mile. If you think your car cost twenty cents a mile you are fooling yourself.

          • adirondacker12800

            Yes it is. Which increases your cost per mile. 30 mpg at $3.00 a gallon is still ten cents a mile. Twenty cents a mile is very optimistic.

          • Henry Miller

            > $4.00 a gallon at 40 mpg is ten cents a mile


            > The insurance is five cents a mile.

            Insurance is not charged by the mile it is charged by the month. It doesn’t matter if you drive the car 24×7 (stopping only to get gas and swap drivers), or it sit in the driveway never moved at all, you pay the same. (If you know in advance you won’t drive it you can get a cheaper insurance rate)

            >Consumables are are two cents a mile

            about right.

            > Registration and inspection is penny a mile.

            Once again this is charged by time not distance. I pay registration once per year. It doesn’t matter how much or little you drive, you pay the same fee.

            > If you think your car cost twenty cents a mile you are fooling yourself.

            I kept careful records of my last car, bought it new, kept it for 13 years (sold in 2019), it costs me just under $.15/mile, or 280,000 miles. That is real numbers for a real car. (I transferred to a city where there was good transit: I almost never drive now so there is no way I could get numbers like that again)

          • Matthew Hutton

            Right so if one ran a ~100mph express train where the tickets are $0.15/mile that would be decent for the train company and cheaper than driving for at least some people. And if there was some public transport, taxis, car hire and parking at both ends that would help it work for people.

          • adirondacker12800

            Insurance is charged by a lot of things and the insurance company wants to know how many miles you drive in a year. Your insurance premiums were higher for 20.000 miles a year or you lied to the insurance company. I want to see your books.

          • Henry Miller

            I’ve only once had insurance company ask about how far I drive, and that was a car that I insured for a lower rate because it was rarely used. In general I get unlimited miles for my insurance as does everyone else I know.

            I threw the books away when I sold the car so you will have to take my word for it. Though my lifetime average was about 47mpg on that car (diesel that mostly used on very long trips), which brings the costs don’t a lot.

            > Right so if one ran a ~100mph express train where the tickets are $0.15/mile that would be decent for the train company and cheaper than driving for at least some people

            Per ticket or per family group? You are competing with everyone in my car. Most people I know going on longer trips are going with friends or family, and so you are not competing with 1.2 people in a car (IIRC that is average during rush hour), but with 4 people in a car. That is why I think local transit systems should use number of riders on an unlimited rides monthly family pass as a metric: it forces them to compete in ways that will make transit better for everyone. (though for long distance trips the concerns should be different)

          • adirondacker12800

            Either the insurance companies don’t ask you how many miles you drive in a year or they do have differing rates depending on how many miles you drive. Drive your diesel cat into the ground, 300,000 miles, a $25,000 car cost 8.3 cents a mile just to buy. If travel was Ozzie and Harriet and their two kids Ricky and David the airlines wouldn’t have business class. And Acela would be empty. You are cooking the books.

          • Sascha Claus

            Most people I know going on longer trips are going with friends or family, and so you are not competing with 1.2 people in a car (IIRC that is average during rush hour), but with 4 people in a car.

            On a tangent, that is why in many German transit systems, you can take an additional adult and up to three or four children with your monthly pass on weekends and weekdays after 17:00 or 18:00 or 19:00 or so—to compete with the car and its ability take the whole family after work.

    • Alon Levy

      The postwar European systems didn’t come in a package with anti-car policy either. To the contrary, the cities that built them usually also built motorways at the same time; to this day, German Greens think that subways and S-Bahns are all about getting transit out of cars’ way and the only moral public transport is trams. Anti-car policies like bike lanes and pedestrianization happened much later – in Paris only starting under Bertrand Delanoë in the 2000s and more recently Anne Hidalgo.

      • David S.

        > German Greens think that subways and S-Bahns are all about getting transit out of cars’ way

        You’re probably aware of this, but for subways and the Stadtbahn systems this was at least partially the Intention at first (for example the original proposal for a Stadtbahn in Stuttgart by W. Lambert pretty much says so directly).

        Looking back, this was a mixed bag: The new (underground) infrastructure is usually much faster, higher capacity & more reliable (and thus much more, but has fewer options for going through the city center. In Stuttgart, this leads to a few bus routes that were formerly above-ground tram routes through the city center and are at 5 min headways, which should be trams and only 2 trunks through the city, limiting capacity (compared to the former tram network, which was much slower but more extensive).

        I think new projects like the Karlsruhe Kombilösung are much better approaches than the 80s Stadtbanproposals (putting high capacity trunks underground, but also leaving flexible routing on the surface elsewhere and even adding to it).

        • Alon Levy

          The Kombilösung unfortunately also includes a road tunnel :(. Same way how here, the only party that’s really interested in U- and S-Bahn expansion, SPD, dumped the Greens from the coalition and brought in CDU for a term in order to build the 16th phase of A100 and in theory also supports the 17th.

          • Sascha Claus

            Karlsruhe had multiple attempts at getting a tram tunnel through a public vote, and only after bundling it with a road tunnel, the city mayor finally got his wish of “getting trams out of his view from his office window”, as some local put it.

  8. liguangming1981

    Most of Shanghai’s suburbs are served by suburban metro lines, though in most cases (except Lines 5, 16, and 17) these lines do go all the way into the city centre (unlike Beijing, where many suburban metro lines terminate at suburban interchange stations with other lines that do go to the city centre). However, Shanghai at least has also embarked on a major program to build proper high capacity, high frequency, high speed suburban rail, with two lines (Airport Link and Jiamin Lines) currently under construction and several more planned for construction in the next decade.

  9. Pingback: Vancouver, Stockholm, and the Suburban Metro Model | Pedestrian Observations

Leave a Reply

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

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

Twitter picture

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

Facebook photo

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

Connecting to %s

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