S-Bahn and RER Ridership is Urban
People in my comments and on social media are taking it for granted that investments into modernizing commuter rail predominantly benefit the suburbs. Against that, I’d like to point out how on the modern commuter rail systems I know best – the RER and the Berlin S-Bahn – ridership is predominantly urban. Whereas the typical American commuter rail use case is a suburban resident commuting to a central business district job at rush hour, the typical use case on the commuter trains here is an urban resident going to work or a social outing in or near city center. Suburban ridership is strong by American standards, benefiting from being able to piggyback on the high frequency and levels of physical investment produced by the urban ridership.
Here’s Berlin’s passenger traffic density on the U- and S-Bahn, as of 2016 (source, p. 6):
The busiest section of the S-Bahn is the Stadtbahn from Ostkreuz to Hauptbahnhof, with about 160,000 passengers per weekday through each interstation. The eastern sections of both the north and the south arms of the Ringbahn are close, with about 150,000 each, and the North-South Tunnel has 100,000. These traffic density levels extend into outer urban neighborhoods outside the ring – ridership on the Stadtbahn trunk remains high well into Lichtenberg – but by the time the trains cross city limits, ridership is rather low. All tails crossing city limits combined have 150,000 riders/day, so a little more than a quarter of the ridership density on the city center segments. Of those tails, the busiest, with a traffic density of 24,000/day, is to Potsdam, which is a suburb but is an independent job center rather than a pure commuter suburb like the rest of the towns in Brandenburg adjacent to Berlin.
I don’t have similar graphics for Paris, only a table of ridership on the SNCF-RER and Transilien by station and time of day and a separate table with annual ridership on the RATP-RER and Métro. But the results there are similar. Total boardings on the RATP-RER in 2019 was 399 million, of which 52 million originated in stations in the Grande Couronne, 186 million in the Petite Couronne, and 161 million in the city. If we double the Grande Couronne boardings, to account for the fact that just about all of those riders are going to the city or a Petite Couronne job center like La Défense, then we get just over a quarter of overall ridership, a similar result to the traffic density of Berlin. On the SNCF-RER, the share of the Grande Couronne is higher, around half.
The city stations include job centers and transfer points from mainline rail and the Métro – there aren’t 47 million people a year whose residential origin station is Gare du Nord – so it’s best to view the system as one used predominantly by Petite Couronne residents, with a handful using it as I did internally to the city and another handful commuting in from the Grand Couronne. This is technically suburban, but the Petite Couronne is best viewed as a ring of city neighborhoods that are not annexed to the city for sociopolitical reasons; the least dense of its three departments, Val-de-Marne, is denser than the densest German city, Munich.
The difference in this pattern with the United States is not hard to explain. Here and in Paris, commuter rail charges the same fares as the subway, runs every 5-10 minutes in urban neighborhoods (even less on the city center trunks), and makes stops at the rate of an express subway line. Of course urban residents use the trains, and we greatly outnumber suburbanites among people traveling to city center. It’s the United States that’s weird, with its suburb-only rail system stuck in the Mad Men era trying to stick with its market of Don Drapers and Pete Campbells.
Pedestrian Observations 
I’ve lived in both cities and can attest how useful both the S-Bahn and the RER can be for those kind of inner urban trips. Nonetheless, I think you would get a different ridership picture if you look at passenger km rather than raw passenger numbers (if this data is available).
I might use the RER for a quick trips of a couple of km from Gare de Nord to Luxembourg, or the S-Bahn from Alexanderplatz to Friedrichstraße, but many of the people around me would be doing 30-40km commutes from Poissy to Châtelet-Les Halles, or Strausberg to Zoologischer Garten, and I think it skews the picture to consider these trips as the equivalent of each other.
>> many of the people around me would be doing 30-40km commutes from Poissy to Châtelet-Les Halles, or Strausberg to Zoologischer Garten..
Yes, but the point is, not as many.
In case you want data, for Barcelona:
Click to access act%2020210309__PLA%20RODALIES%20DE%20CATALUNYA-CATAL%C3%80.pdf
Page 120 there’s list of stations and their daily ridership (in 2018). The first 7 stops in the ranking are in the city’s urban continuum, but this makes sense because people who go one way, then they have to come back the opposite way. I did some time ago a sheet with that data organized by line, in case you’re curious (I could clean it up better but don’t have time now):
https://docs.google.com/spreadsheets/d/1gZCz-vLTr5KcbkNGcDmU-e7fCZnGT-eH/edit?usp=drivesdk&ouid=106275498693178436401&rtpof=true&sd=true
In many of the corridors there’s some more or less clear boundary after which ridership drops. For example, in Vilanova (57 minutes away from city center), Martorell (45 min away), Terrassa (48 min away), Mataró (43 min away), and Granollers (34 min away). Most of those places are towns on their own, many with density and destinations for people in the area. Off-peak offer inside this circle goes between 4-6 tph, reaching 8 tph on morning rush. Another format for ridership data, broken up by time, is here:
https://datos.gob.es/ca/catalogo/ea0003337-volumen-de-viajeros-por-franja-horaria-barcelona1
For the FGC, the Catalan-run railway, the data from 2023 is here, in pages 125 and 126:
Click to access iig_fgc_2023.pdf
Note that in Barcelona – Vallès line, everything until Peu del Funicular can be considered city continuum, and in Llobregat – Anoia until Cornellà Riera or Sant Boi, and frequencies and stop spacing are metro-like. Their open data portal is here:
https://dadesobertes.fgc.cat/pages/frontpage/
Have you seen Lander’s proposal to fare integrate all within-city-limits commuter rail stations and make them $2.90? It’s very promising if they can actually implement that, as it’s that plus much better and more consistent city service that will allow this kind of RER/S-Bahn growth. Already the city ticket has helped city station ridership a lot (almost all recent ridership growth has been at those stations). But without higher ridership at them, LIRR/MNR won’t care about them enough, and the fare integration and NYCT fares would help jumpstart that. And better integration would hopefully also convince the MTA to treat commuter rail as the best reliever of all three of its busiest lines: Lex, the 7, and QBL.
fare integration is no longer the primary problem frequency is. Look at how few trains stop at Kew Gardens, Forest Hills, and Woodside off peak. It’s difficult to get someone interested in riding the train from these places when the LIRR continues to under provide frequency because the entire scheduling department parks at Babylon and takes a train to Jamaica. And of course one can’t even easily take a train from like forest hills to woodside!
Agreed. I don’t think frequency of urban service can be underestimated as important. The Chiltern mainline London stations have very weak ridership even though the train is very quick and the fares are zonal.
this is because unlike most London service it is only hourly
Why would I want to walk blocks and blocks to the Forest Hills LIRR station and wait and wait to go to Woodside when I can there faster on the subway? Why do I want to go to Woodside is better question. I’m much more likely to be going someplace other than the few LIRR stations scattered about.
I mean, why does the E train even bother stopping at 74th? It’s not as if there’s that much Forest Hills-Jackson Heights demand either.
I realize the view, east, across Sutton Place is almost as hazy as the one west across Ninth Ave.
I’m not the one who imagined all the people in Forest Hills live on Station Square and want to go to the Dunkin Donuts on Roosevelt Ave. in Woodside. Especially since there are frou-frou joints on Station Square and a Dunkin Donuts a few blocks away on Austin St. and Ascan Ave. Though the Dunkin on Continental Ave. between the LIRR station and Queens Blvd is closer. And on the way if they are taking the subway because it’s faster than taking the LIRR to Penn Station and changing to the E train to get to the World Trade Center. Or taking F train to Rockefeller Center because it’s faster than hiking from Grand Central. Someplace other than Woodside or Madison Square Garden or the Chrysler Building.
The express trains stop at 74th so that people on the local trains can change to the express. Or, this might cause confusion among the yokels from the hinterlands, the express train to a local. F train from Union Turnpike to the 7 train to get to work in Long Island City !! !! That is near a 7 train station not the LIRR.
And that can be turned on it’s head. Why haven’t you added yet another redundant commuter train station in Jackson Heights? I’m sure there are thundering herds of Long Islander too stupid to change to the E train in Jamaica to do that. Or the 7 train in Woodside. Ya know, trains that
…. though the 7 from Woodside to 74th is local..
Paris and Berlin have three-ish levels of service. New York is big enough to have four. It’s too too bad it offends your sensibilities. Or lack of imagination. Dallas barely has one and it is unlikely to ever have two.
I think they’re both quite important still. There are tons of bus to subway riders for whom taking LIRR would be nearly an hour faster. LIRR frequency is not nearly so bad that the bus + subway would be faster in those cases, and the buses are usually not nearly as frequent as the subway as well. So fare integration would shift a huge chunk of those radial bus to subway riders to the LIRR and save them hours each day. Those buses can then be repurposed into more frequent feeder buses instead of competing against the faster, higher capacity service.
Moreover, it’s quite hard to convince the commuter railroads to run RER/S Bahn-level service, but things like the City Ticket have already been successful, so I think it may be easier to first do fare integration, have ridership increase as a result, and use that to force LIRR to run more service. They’re often against infill stations, for example, because ridership under the current model is so bad and can’t steal any riders from the cheaper, more frequent, albeit way more crowded subway. Fare integration would jumpstart ridership to make infill stations and higher frequency more palatable to the railroads.
I find that extraordinarily hard to believe. People aren’t that frugal, masochistic, stupid or a combination.
People who can’t afford to pay for something with their money will pay for it with their time, and that it’s a major factor in cycles of poverty. Even when money is less of an immediate factor, humans tend to vastly underestimate the value of their time — it’s a widespread and commonly studied fallacy.
People who take long commutes aren’t doing it for shitty jobs with low pay. They can get shitty jobs with low pay closer to home.
The MTA, dastardly MTA, publishes schedules. It takes the LIRR around 20 minutes to get from Jamaica or Flushing – the end of the subway line – to Manhattan. It takes the subway 40. 40 minutes minus 20 minutes isn’t an hour. And, believe it or not, not all the jobs are in One Penn Plaza hovering over Penn Station. It takes time to get off the LIRR train and get on the subway to get to work. The 20 minute long hour is getting even shorter.
It’s hard to believe. Very hard to believe.
best not to respond to the troll.
The best reliever to the Flushing Line and Queens Blvd. Lines are mythical Northern Blvd, Queens Blvd., and Corona Ave./108th St. metro lines. The best reliever to the Lex are mythical 2nd and 3rd Ave. local subways. These mythical metro lines provide coverage to some of the densest rapid transit desert corridors in Queens/Manhattan, and don’t require running superfluous trains all the way to Port Washington, Trenton, Monmouth County, etc. In other words, the *per rider* marginal capital cost and marginal operating cost is still less using the metro (which has shorter trains going shorter distances in denser areas where the marginal person moving in is still much more likely to use transit) than overleveraging legacy rail which has longer trains, wider stop spacing, and goes longer distances. Passenger trains don’t exist to move; they exist to move *passengers*.
Many railfans lose sight of that the reason the trains run is people.
People have tried to explain time and again to Alon and their diehards that the goal isn’t to find the lowest capital cost per km. [For that, you should just run short trains with wide stop spacing in a rural area where the cost of mitigating disruption is minimal.] or minimize the waste of the supply of track. Those are the constraints. The goal is to maximize ridership subject to a host of constraints, which include budget, the supply of existing tracks available, etc.
Also, per rider operating costs, which are eternal, are far, far more important than per km capital cost, which is a one-time cost. As we learned with H St. streetcar, rail routes are only permanent until voters are no longer willing to pay for the operating cost.
[Fwiw, I have a bunch of fantasy NYC Subway map configurations. Below is one of my more aggressive fantasy map permutations: https://www.google.com/maps/d/edit?hl=en&mid=19V9sQMoIAVfgSGl_dhdSV-zSlY4t2Ng&ll=40.7362364326179%2C-73.86827852119377&z=13
Would love to hear your opinion.]
Let’s compare operating vs capital costs for a hypothetical new Northern Blvd line from Flushing to Manhattan at current operating and capital costs.
It’d be about 10 miles, and at SAS2 costs per mile, plus probably $5 billion extra for an East River Tunnel (a 7 train tunnel to NJ was ~10 billion years ago and Gateway is $17 billion), it’d be $50 billion.
For the operating costs, let’s assume a 20 mph average speed, so 30 min runtime. If you run trains 30 tph during rush hour (7-10 am, 4-7 pm), 15 tph off-peak and weekends, and 3 tph overnight (6 hours), run 10-car trains, and have the same current NYCT operating cost per car per hour of $280.36 (which would likely be lower on a new automated line), that’ll be $180 million/year in operating costs.
At that rate, cumulative operating costs will only surpass capital costs after 280 years. So I don’t see why operating costs are far more important than capital costs. Saving $50 billion and investing that in operating costs instead could run so, so much service.
Before you splatter squiggles ask what services already or may exist? And define some goals. It has too much capacity and goes to the wrong places.
Khyber Sen — First off, my Northern Blvd. Line (which is in Alon’s maps as well), doesn’t require a new East River Tunnel: the Northern Blvd. Line uses the 63rd St. Tunnel. What the Northern Blvd. Line does do is provide some of the densest neighborhoods in Queens a legit alternative and faster commute to Midtown/Weil to the overcrowded Flushing and Queens Blvd. Express Lines. It also increases bus turnover, making getting rid of many express buses much, much easier (saving on operating costs). And why is per car hour (and not per rider) the appropriate scalar for operating costs?
Reality is the Flushing and QB Express Lines are still quite crowded. And the amount of labor hours necessary per passenger isn’t going down. The only way to increase capacity, reduce the amount of labor necessary per passenger, and provide rail coverage to dense rapid transit deserts such as Northern Blvd. is to expand the subway.
adirondacker12800 — My goals — to increase everywhere to everywhere access, add capacity, and reduce per rider operating costs.
You have to keep
“Passenger trains don’t exist to move; they exist to move passengers.”
in mind when you want to put a train on the tracks. Someday far far in the future the Second Ave subway will be completed. All of the Queens Blvd express train will be using up the capacity in the 63rd St. tunnel….. Add up the number of rush hour trains on the E and the F. And it is likely to be similar amounts in the future.
If the E train is going to Second Ave that leaves lots of capacity in the 53rd St. tunnel. To the local tracks. Which means the R train won’t be able to go to Queens Blvd any more. The W train runs during the day because people crowd onto the 7 and change to the N or the W. If the R and the W are running along Northern Blvd the people taking the Junction Blvd. bus to Roosevelt Ave. could just get on the R or the W. Rejiggering the services on Broadway will have to be done.
Especially if you send the Second Ave trains through the Montague St. tunnels. The R could terminate at City Hall and the W at Whitehall St. … Send the Second Av. train to the Fulton Street subway people at Lefferts Blvd could get more service and people who whine that the A train takes a long time to get to Times Square could complain that the E train doesn’t go to Grand Central.
I want to know what the new East River Tunnels are going to connect to. The Tenth Ave. subway? The Second Ave. express?
It depends on which way you want to cook the books. The books were cooked reallllllly reallly hard to come up with 280 years.
adirondacker — I have the same number of East River Tunnels as Alon does in their fantasy subway map. I have the following permutation for East River Tunnels:
The 10th Ave. ALM (has termini at 2nd Ave./86th St and City Hall), which is basically a stronger version of Alon’s 86th St. Tram.
Alon’s subway map has too much capacity. Especially when the suburban map has trains making almost as many stops as the express subway trains. The commuter map has too much capacity too. Combined they have far too much capacity and it’s in the wrong places doing the wrong things.
However you are slicing and dicing stuff, if the 63rd St. tunnel is full of Queens Blvd. trains, whatever they are, you can’t send Northern Blvd trains through it too. If the 53rd Street tunnel is full of Sixth Ave and Eighth Ave. trains Broadway trains can’t be on Queens Blvd too. Once the Second Ave subway is complete to … however it gets all the way downtown… they don’t have to build anything to use up all the capacity…… Northern Blvd trains have to go somewhere else. Because if “Passenger trains don’t exist to move; they exist to move passengers” a corollary is that once you have enough passengers to move as many trains as possible you need different tracks to move more trains.
And you have to keep in mind that “Passenger trains don’t exist to move; they exist to move passenger” If the Queens Blvd locals are going through 63rd Street how do I get from express trains to the local stations between Queens Plaza and Jackson Heights. Expresses through 63rd means local passengers west of Jackson Heights can get to 8th Ave., 6th Ave., and can change to the Lex. or 2nd Ave. at 53rd and Lex.
…. Sixth Ave trains, on the lower level at West 4th St. can go to the World Trade Center. They have to go to the upper level. And Eighth Ave. trains can go to Houston Street they have to go from the upper level to the lower level. Much less chance of a problem occurring if the 8th Ave. trains stay on the upper level and the 6th Ave. trains stay on the lower level.
Trains on Third Ave. are half a block away from the trains on Lexington and a whole block away from the ones on Second. Six tracks is more than most places have. Perhaps putting trains where people don’t have trains would be a better choice. … trains moving passengers instead of finagling how to move trains…
….. there’s too much capacity and it’s in the wrong places.
adirondacker — You’re totally misreading what I’m doing.
For starters, both Alon and vanshnook have in their maps four additional 600 ft. subway tracks below 42nd St. [Alon’s is regional rail Harlem Line to Staten Island and Third Ave. to Utica.; vanshnook has a 4-track Second Ave. Subway below 42nd St.] My plan does the same: it has two tracks on 2nd Ave. (going across 125th St. uptown and to Coney Island via Nassau/Montague/Fourth Ave. Local/West End) and two tracks on Third Ave. (going to Queens and South Ferry).
The 2nd Ave. Local Subway
The Third Ave Local Subway
So both the proposed 2nd and 3rd Ave. lines do lots of different things.
There is no interlining in the Bronx, in Queens west of Briarwood, or in Brooklyn north of 36th St. — well, except for things like the 6, 6<> during peak hours. Please reread what I wrote about the routes in Queens, and look at the map again.
Honestly, given how much you castigate Alon for their dearth of improved crosstown travel, I thought you would enjoy my map, which has much greater emphasis on crosstown travel.
You and Alon have too much capacity and it’s in the wrong places. Take two deep breaths, stare off into the distance for a while and contemplate that in some alternate universe where the population of New York City is 10 million, that is an increase from 8 million to 10 million. 25 percent not 200 percent.
Increasing the population from 8 to 9 million is 12.5 percent not 200 percent. Though it appears the increase to 9 million is looking iffy. Second Ave subway parallel to the Lexington Ave Lines is increasing from 4 tracks to 6 tracks. 4 tracks to 6 tracks is a 50 percent increase. Likely more than enough though Second Ave being all local all the time is disappointing. But likely more than enough for a very long time.
Deinterlining is a railfan relentless urge to move more trains. It’s a passenger railroad, the goal is to move passengers not trains. You and Alon have too much capacity and it’s in the wrong places.
The clue seems to be sinking in!!
When there are 8 tracks of railroad within a block or two of Austin Street and Continental Ave, in Forest Hills Queens, some of the trains can express through because people have other ways to get there. There are three levels of service in Paris. There are four in New York. There aren’t going to be three levels of service in Dallas because they barely have one.
Call the M train Metro 42 because Sixth Avenue local trains are numbered 40 through 49, the R train Metro 33 because Broadway local trains are numbered 30 through 39. The F train becomes RER D1 when it’s originating at 179th Street and D2 when it originates at Jamaica Center. And D3 when it’s heading to 179th and D4 when it’s headed to Jamaica. It will confuse yokels from the hinterlands even more than calling all of them the F train but it’s gotta be designated that way because railfans are dazzled by Paris. Sixth Ave. expresses are RER D-something.
I fear your head will explode. Someday far in the future train from Jamaica to the World Trade Center will stop using the 53rd Street tunnel and use the 63rd Street tunnel to go down Second Ave. When it goes to Far Rockaway it can be RER F1. When the T trains comes down from the Bronx to go to Coney Island the RER F2. Because Second Ave. RER trains start with an F. The Q train will be Metro 147 because it wanders all over the place. Just think of the enormous list of trains available at Broadway junction because railfans want it to be Paris. Just think, people who complain about the A train from JFK can whine about the the former E train too!
Though calling the commuter trains the Sbahn makes it Berlin. I’m seeing four digit numbers so that when the train from Valley Stream leaves for Poughkeepsie it’s 4001, when it leaves for Croton it’s 4002, Spring Valley, 4003, Suffern via the Bergen line 4004, Suffern via the Main Line 4005, Port Jervis via the Main Line 4006, Dover via the Boonton Line 4007……. all the suburban stations have service to all the other suburban stations 2 or 3 times a day. Or it seems that’s what clueless railfans assume. 4001 through 4030-ish is for the trains through St. Albans to Penn Station and beyond. 4040 through 4070 or so can be for the ones through Locust Manor. 4100-30 for St. Albans through Brooklyn to all the fabulousness of through ruuuunnnnning means. 4140 through 4170 through Locust Manor. There are other ways to arrange that. Because railfans assume any suburban station they start out at will have one seat service to any other suburban station. The departure boards at Jamaica will be ginormous.
Whoop. De. Doo. In very round numbers Potsdam is as far from Berlin as Jamaica is from Manhattan. What’s the Long Island Railroad Ridership through Jamaica? What’s the ridership along the Queens Blvd. line? How ’bout between 125th Street and Grand Central along Park Ave. versus 125th Street and Grand Central along Lexington. Tis a pity it would be very difficult to tease out the difference between Lexington Ave Local traffic and Lexington Ave. express. Though for both the Queens Blvd. lineS and the Lexington Ave. lineS it would be possible to get numbers for the far flung stations that are served by express trains only. Where the suburban trains don’t stop.
The Port Authority makes it much easier to find information.
https://www.panynj.gov/path/en/about/stats.html
A bit more than the Journal Square station? Or twice as much as Journal Square’s Saturday ridership. I’m assuming you are yammering on about workday ridership. And PATH runs all local all the time except in the dead of night when it runs even localer. How much is the ridership of the U-bahn or the S-bahn at 3 in the morning on a weekday?
Potsdam-Friedrichstraße is 29 km. 29 km from Penn Station is New Hyde Park, not Jamaica.
The problem with talking about Queens Boulevard and the Lex is that there’s quite a lot of ridership past the endpoints in Queens, all of which goes on slow buses to Jamaica and Flushing rather than on the LIRR where it belongs.
What part of “very round” is confusing? I didn’t dig deep for which point Berliners consider the center of either, I let Google decide. Because I don’t do this for a living.
The endpoints of the Lexington Ave lines are in Brooklyn or the Bronx. With no transfer, if you pick the right train at the right station, service to the West Side in addition to the East Side. Or cross platform transfers if it’s not “wait for the next train”. People along the IRT have had “East Side Access” and “Penn Station Access” since the Dual Contract stations opened in 1917. Just awful. It offends railfans who are interested in moving trains instead of moving people. And Union Square access and Times Square access and Fulton Street access. It’s unfortunate that sometime in the far far future people using the S-bahn from far flung suburbs will have to make the decision between taking an RER train or a subway train but that happens when you have enough demand to have multiple choices. ….. like they do now at Penn Station and Grand Central. and choices at Jamaica, Newark or Hoboken. Like people in Philadelphia do.
The view across Sutton Place is very hazy. How do you get from College Point to the LIRR? I spent a few moments looking at schedules. I want to know how a 20 minute ride on the LIRR, from Flushing or Jamaica, versus a 40 minute ride on the subway, saves a hour. The savings assumes everybody works in One Penn Plaza hovering over Penn Station. Not Empire Blue Cross Blue Shield in Rego Park. Imagine that someone from suburban Briarwood commutes to a job in suburban Rego Park. Without going to Manhattan – preferably One Penn Plaza – or using the LIRR
….There is life east of Sutton Place.
Which areas of New York City would make up the bulk of commuter rail origins with a more modern system? If I had to guess, I would say that the majority of LIRR demand would be in the area from Flushing/Jamacia east to Great Neck/Mineola/Freeport, while the area for Metro North would be from Morris Heights/Tremont to Yonkers/Tuckahoe/New Rochelle. However, perhaps the boundaries of this petite coroune equivalent should be moved further out or closer in. I would describe most of this area as suburban, just relatively high-density suburbs, with satellite cities like Yonkers and Hempstead.
….Manhattan, in the afternoon anyway. Weekdays.
I think that’s about right. We could model this with STOPS, but the model gets iffier the farther it gets from urban rail, and even for urban rail it needs a fudge factor (in effect, it needs to be normalized to overall ridership – it can estimate ridership on a line given assumptions on the system but cannot estimate ridership on a system).
City line-ish. Reality intrudes. You have to define goals. Not going through Midtown is a high priority goal in Manhattan. Someone who is going Downtown directly isn’t in Midtown. Which means someone else can go to Midtown.
Squint at satellite images, because New York City property tax maps are slow, the LIRR between Jamaica and Floral Park is six or more tracks wide. There would be lots of screeching about the Great Wall of Queens but the embankment could be converted to something with retaining walls. And the underpasses updated to modern clearances. For six tracks, two for the local to Floral Park and four from the expresses from Nassau and Suffolk to Midtown. The local tracks could be for the Downtown trains that run through Brooklyn to New Jersey. Everyone can merrily change between Wall Street, Penn Station and Grand Central trains in Jamaica.
If you squint at satellite images the ROW of the branch through St. Albans it is four tracks wide. The local to Downtown could toddle along local tracks while the trains from Nassau and Suffolk express from Valley Stream to Jamaica. Where people on the branch through Locust Manor along with the St. Albans and Floral Park branch merrily change trains with the people on the Penn Station or Grand Central trains.
18 trains an hour gives each branch every ten minutes between the city line and Jamaica. Since people in Auburndale didn’t view the job market across Sutton Place they went and took jobs that aren’t in One Penn Plaza. It would cost a lot of money, one every ten minutes from the Port Washington branch along the Grand Central Parkway to Brooklyn would give them much faster service to Downtown. Without going through Midtown. Every ten minutes to downtown and every ten minutes to midtown means people who want to go from Little Neck to Jackson Heights can get a train to Flushing and change trains to get there. Fabulous inter-suburban-doesn’t-go-to-the-core trips that fascinate railfans everywhere.
The amount of stops on a local are somewhere between 5 and 10. Jamaica to the city line gives you that.
I feel like I should be sorry for asking but are these a fair comparison? The RER and Berlin S-bahn are some of the most urban “commuter rail” systems you can bring up. The RER definitely relatively unique but it’s often compared to subways because the a lot of the metro has the stop density of a tram, and the Berlin S-bahn is just the U-bahn with longer lines and more interlining.
I just don’t see how you can reasonably compare these to any American commuter rail system except for SEPTA, especially when both Paris and Berlin also have the Transilien and a network of RB/RE trains respectively. American systems have longer distances and serve lower density akin to those, not to the RER/S-bahn. I would probably compare them to that, or to more “suburban” systems like Stockholm, Zurich, Vienna, or most other German S-bahns outside of Hamburg.
Not to say that American systems should have nonexistent to poor urban service and integration as they currently do (or even worse, be Metra with higher suburban than urban stop density), but I just don’t think you can pretend they are capable of being the RER or Berlin S-bahn in terms of servicing the city proper without virtually constructing an entirely new system out of each of them by trimming lines, separating tracks, and entirely redoing station spacing.
I think the comparison is fair. With 8.5 million New Yorkers over 778 square km = 10925/km2, NYC is arguably more urbane than Berlin or Paris. Once Staten Island is taken out of the equation because it lacks rail connection to the other boroughs, there are 8.2 million New Yorkers over 629 km2 = 13036/km2.
There are only 3.9 million Berliners over 891 km2 i.e. 4377/km2. The urban area Alon is discussing, Paris and the 3 Petite Couronne departments, has 6.7 million inhabitants over 762 km2 or 8793/km2. NYC has the population but lacks the fare integration and the new tunnels found on Berlin or Zurich’s S-Bahn, Paris RER, Milan’s Passante, Geneve’s LEX or London’s Crossrail.
I’d include Munich in this post if I had data for it.
In Paris, Transilien is more suburban than the RER, yes. Transilien H and J are predominantly (around 75-80%) Grande Couronne; but Transilien L is majority-urban, with Grande Couronne boardings only 19% of the total (so probably around 38% of ridership). And then Transilien itself is around a quarter of the combined ridership of Transilien and the RER.
I don’t know how Transilien maps to Britain, but only 63% of South East riders go to London and only 73% of East riders go to London.
(This is as per https://dataportal.orr.gov.uk/statistics/usage/regional-rail-usage/)
I wouldn’t be surprised if Transilien is more Paris centric – especially as I don’t believe France does through ticketing as much?
However even so if it was much more than 80-85% I would be surprised.
And the LIRR is more suburban than the Queens Blvd trains or the Flushing trains and locals on Queens Blvd, during the day, provide local service so other people can use the express trains. And the people on the express tracks through western queens are bright enough to change trains in Jamaica when they want to go to Rego Park.
You could make the case that the MNRR East-of-Hudson locals’ stopping patterns – e.g. New Haven making all stops between GCT and Stamford, are very similar to Transilien/RER.
Certainly in London there are all day regional express services.
They do in metro New York too. It’s just that some railfans didn’t pay attention when their mommy attempted to teach them how to read timetables and consulting the super computer welded to their hand to use a trip planner is too onerous.
https://www.mta.info/document/124231
….. twice an hour around a quarter past and quarter to… might be unintelligible because they don’t conceive of time in fractions of an hour. And “some time near :15 and :45” would mean they would have to use the super computer welded to their hand to consult the trip planner. Or, I know this is a very quaint idea, consult the paper timetable carried around. In a pocket or a briefcase or back pack. The young ones seem to be carrying around a weekend’s worth of stuff in their enormous backpacks and something that could be put in a letter sized envelope is too big.
I forget where I read it – it could have been the NE High speed rail report – but I believe that over the last two decades ridership into Manhattan overall has grown but ridership into Manhattan during “core commute” times – e.g. 8 AM – 9 AM – has fallen. One can conclude therefore that American rail passengers at least in the NY Metro area – who account for the majority of rail passengers nationwide – are already moving toward the European use case for a “regional” rail system, not a “commuter” rail system, with growth especially post-Covid on off-peak and weekend travel, reflecting changing work habits (even with return to office, flexible hours), increasing intolerance of congestion on roadways, recognition of driving expenses (time, parking, tolls), mental health, exposure to emissions. I think I saw some stats for Caltrain post-electrification which showed pretty dramatic growth for weekend travel as well.
In other words it’s a perfect opportunity to transition to the scheduling recommendations for NE HSR – reduce frequencies for rush hour or get to parity with off peak to suit the needs of the public which are already revealing themselves.
The only obstacle is the unimaginative, mostly incompetent leadership of our appointed transit agency heads. And man, is it a big one. I speak from experience, I lived in Fairfield County for a time. I vividly remember taking M2s on the New Haven Line with the southbound outer (platform-side) track closed for maintenance and having the rattly buggers be manually maneuvered into position against the temporary bridges over the outer track — the driver and conductor calling to each other how many feet were left by radio and PA. I was like “What year is this?!”
…. So this building new tunnels for more trains is silly and people should just take a train on Saturday? When the office is closed? I suppose they could do it after dinner. When the office is closed. Just what do you have in mind?
I didn’t say any of that.
If rush hour capacity is being underutilized, or being utilized because it’s mismatched with customer demand–i.e. if there are empty seats on peak service and latent demand off-peak, then it makes sense to equalize between peak and off-peak to get to a uniform clock-face schedule, all other things being equal like available staffing and equipment. The benefits of clock-face scheduling and coordinating takt between suburban/commuter/regional services and intercity are explained at length in the NE HSR report, and quite frankly, I trust the mathematician who writes this blog over, well, you.
Tunnel projects have a degree of independence from infrastructure improvements for scheduling purposes. New tunnel bores for Gateway are necessary because you can’t rehab the existing tunnels without curtailing service, especially with Amtrak’s construction efficiency and cost control. By contrast, there’s nothing wrong with Shell Interlocking – it didn’t get flooded by a hurricane – but no one can argue that grade-separating the junction would shave a huge amount of time off inbounds to GCT/NYP.
I’m new to this blog but it sure sounds like you’re a prickly fellow, and I cannot for the life of me figure out what you’re in favor of.
J.G., just don’t feed the incoherent reply-guy troll.
He’s been at it for decades now.
Ha! Thank you for the warning.
What a depressing existence theirs must be.
Silly me I thought the subject was the real world. As in
Where new tunnels are being built. Because the existing ones are at capacity. Not railfan’s speculation about something somewhere maybe perhaps sometime happening.
Oh man, if only the Northeast US’s passenger rail services could follow your trolling takt time!
They can’t because there are capacity problems. Which I’m sure the report you supposedly read droned on and on and on about.
Bad troll! Bad! Back under your bridge!
It was in the HSR report, cribbed from the ETA commuter rail report. But it wasn’t just the last two decades but also the last six – the comparison is with the Hub Bound report of 1961. It’s a long-term change, which corona greatly accelerated.
Thank you for your response! I recently signed up for ETANY updates too and I’m slowly reading through their reports!
Do you foresee this trend continuing in the face of return to work moves by major corporations and population migration patterns in the New York metro area?
If so, that would certainly make the case stronger for policymakers to adopt these recommendations.
If we lived in a logical world, of course.
In both the ETA and the Marron report, the figures only go up to 2019, precisely because I didn’t want to make assumptions on the impact of working from home. I expect that return to office will be done in a way that accelerates the prior trend, as much of the return involves hybrid work or more flexible working hours, rather than rigid 9-to-5 hours (which were in decline in the office sector anyway).
That makes total sense. I’m sorry, I have one more question if I haven’t exhausted my welcome.
Do the ridership models you use at any of the organizations you do work, take into account projected population changes by the time the proposed capital projects and operational changes (e.g. scheduling) might be completed?
I greatly appreciate your work, your writing, and your willingness to engage with a new follower. I have learned a lot.
They don’t. STOPS explicitly only considers current development and current ridership – we need to estimate overall system ridership before we can estimate the ridership of a new line.
Arithmetic keeps rearing it’s ugly head. A slim majority of New York City households are carfree. I’m going to go out on a limb here and speculate that they don’t use automobiles much. Every last one of the people increasing the population from 8 million to 9 million might, assuming car owners never use mass transit and drive everywhere, increase mass transit demand 25 percent. It doesn’t need three times as much railroad as exists today.
I’m still wondering how a 20 minute ride on the LIRR saves an hour over a 40 minute subway ride. ….Pesky arithmetic.
An increase in population from 8 to 9 million in New York would raise transit ridership by close to 1 million users (>2 million daily trips), not just 25% of current ridership, because in practice, if someone’s moving to New York, it’s to work in or near Manhattan, not to commute from Corona to a retail job in Flushing.
Your upper middle class symbol manipulator biases are showing again. Badly. You have to coordinate and collate your myths, legends and fables.
The common stereotype is that the immigrant works a low paying job, the kind “Americans don’t want” something like….I dunno, working retail in Flushing while living in Corona…. Maybe someday is able to claw their way up to move to the suburban parts of Queens. To take the bus to Flushing or Jamaica. And take a 40 minute subway ride to Penn Station that takes an hour longer than the 20 minute LIRR ride. For some odd and peculiar reason because they got a better job in Rego Park.
Is it that the immigrant bussing your table doesn’t count as a subway rider or is it that he or she is invisible? Or the immigrants that make your hotel function?…. or the immigrants staffing mass transit. or the ones staffing hospitals. or…. I suppose it could be both, that they don’t count because they are invisible. Because they don’t have college degrees allowing them to type for a living or just that they aren’t typing and education doesn’t enter into the invisibility?
Census Bureau counts everyone regardless of academic achievement. Or country of birth. Or employment status. Someone from Kansas with a business degree counts as one person. As does the recent immigrant’s newborn.
The recent college graduate that moves to New York, like other New Yorkers, will be able to work at home. Or walk to work. Walking to work is as common among rich people as it is among poor people. Or bicycle. Or, I know this is for peons in the outer boroughs, take the bus. I do hope that wasn’t too disturbing, the concept of someone of your class taking the bus. Even if all of the million are people you empathize with because they type for a living, all work, it’s only going to be 1.5 million subway trips. Three quarters of them using the subway. The other quarter will be doing something else. Like:
Deciding to use the magical mystical through running S-bahn trains that make almost as many stops as the RER trains. To take one of the fabulous jobs people in New Jersey are just dying to take, in Jamaica. I’m still unclear how running NJTransit trains through to Metro North makes it easier to get to Flushing. Or Jamaica. Unless it’s one of the truly magical trains that start out in Trenton and runs to New Haven. While also going to Babylon.
Arithmetic rears it’s ugly head. Again. Though I’m still trying to figure out how a 40 minute ride on the subway is hour and twenty minutes long. Pre pandemic recent peak ridership was 6 million. 2 million more than 6 million is 33.33-ish percent. Not 50, Post pandemic peak was 4 million. Add 2 million to 4 million would be 50 percent. Its also 6 million. Like the pre pandemic peak. But then they aren’t all going to be using the subway.. Hmm. It’s definitely not 100 percent or 200 or 300.
Keep all of the tales in mind. Along with reasonable expectations. The box of 12 crayons would probably be enough. … though a whiteboard and dry erase markers might be a better idea. And a case of paper towels for all the changes you will have to make.
Thank you again!
“Potsdam, which is a suburb” – really!?
Like Berlin, Potsdam is one of Germany’s 16 state capitals. In addition to being the seat of Brandenburg’s government, Potsdam is over TWO CENTURIES OLDER than Berlin. What “suburb” has such stature and predates its “core city”?
Plymouth is older than Boston, Port Moody is older than Vancouver, Petah Tikva is older than Tel Aviv, Brandenburg an der Havel is older than both Berlin and Potsdam, Songjiang is older than Shanghai (which subsequently annexed it, but Songjiang remains an outlying area).
Plymouth may have stature in American history lore but is an outlying town rather than a “suburb” – its development is not contiguous with Boston and the cities surrounding it
Port Moody is no provincial capital
Petah Tikva is clearly subsumed into the urban sprawl around Tel Aviv
Not sure how Brandenburg an der Havel gets dragged into this, but it’s arguably more of a suburb of Potsdam itself than Potsdam is of Berlin
Songjiang is an even more extreme example than Port Moody or Petah Tikva – not only subsumed into the continuous urban development but actually part of the “core city”
None of these red herrings changes the position that Potsdam has too much history AND stature to be relegated as merely a “suburb of Berlin”. Suburbs are “suburbs” because they are undifferentiated and blend into their core cities – lakes and forest separate Potsdam from Berlin’s southwest (itself originally composed of independent towns and villages such as Wannsee and Cedelendorp). Indeed, the divide between Berlin and Potsdam was once so stark that it was an insurmountable international border cleaving the Free World from the Eastern Bloc.
I’d offer that the distinguishing feature which drives passenger traffic in one direction or another or both, and what should be discussed in relation to this post, rather than arguing over nomenclature, age, or political status, is whether it’s an independent job center or a residential area. And it
This may be a little idiosyncratic of me, but ever since reading the sci-fi novel series The Expanse, in which the authors depict an alien hive mind consisting of untold trillions of organisms signaling to one another via light, the limiting factor on growth becomes not communication (as the hive mind is at the theoretical maximum for that) but mass transfer, I have thought of effective mass transit as efficient mass transfer, and measures of efficiency could be calculated based on the rate of mass transfer over distance (e.g. passenger-km/h) divided by a normalizing parameter such as operating cost, or operating cost plus amortized capital spend.
In this paradigm the equation does not care about what us silly humans call the places we live. For a commuter community there is a transfer “push” in the morning and a transfer “pull” in the evening; in an independent job center it is balanced, because it is home to both residences and places of work; and in a central business district the flow is reversed relative to the commuter community.
Then again, I’m not a mathematician but an engineer (not in this field though) and I probably have no idea what I’m talking about. Cheers!
P.S. read The Expanse.
I think the arc in books 7-9 isn’t as strong as the ones in books 1-6, though. The show, wisely, increases the focus on the strongest elements of the book series, namely depicting the Belters in space, and also fronting Avasarala’s role.
Well, as someone who enjoys fantasy & sci-fi worldbuilding while recognizing that not everybody does, nor does worldbuilding alone make a good novel, I did absolutely love that aspect of 1-3. I think you’re right in part because the focus in 7-9 broadened to The Existential War, to the detriment of our squad of heroes.
For a very brief moment in season 1 of the show, it depicted a map of the Ceres Station tube network. I was all, “No! Wait! Show more! What rolling stock does it use?!”
Riders of urban and suburban transit are segregated for socioeconomic reasons. Suburban riders going into the city do not want to be slowed down by extra stops. Moreover, they don’t want to be mixed with disruptive lower income urban passengers, especially in the Black neighborhoods near Jamaica and the Brooklyn LIRR stations. The same logic applies to Metro North in the Bronx.
It’s not 1950 anymore. They let people who aren’t White Anglo Saxon Protestants move to the suburbs.
https://www.census.gov/quickfacts/fact/table/nassaucountynewyork/PST045224
Even though people have been changing…. modes… at the rivers since before the Revolution, railfans think suburbanites won’t change. Even though the suburbs started to develop when the trip to Manhattan involved changing to a ferry. Suburbanites, unaware of the railfan myths and fables, have been changing trains in Jamaica, Newark, Hoboken, Woodside, Stamford….for generations. And very likely can cope with changing to the subway when their trip isn’t to Manhattan. Because the vast majority of them can understand maps, timetables, using a website to look at either, finding the toll free number to call for advice….
The higher fare and lack of integration discourages urban residents of the poorest and blackest neighborhoods from getting on the commuter rail. Expect some of the homelessness and erratic behavior to spill over to the commuter rail if they’re integrated.
What makes you think the people with erratic behavior paid a fare?
Yeah, it’s partly racism like this, but not just. Service to rather middle-class neighborhoods on the LIRR (e.g. Bayside) and Metro-North (e.g. Riverdale) is poor as well. In Chicago, Metra barely serves the North Side and actually makes more South Side stops on suburban lines. Suburban racism compounds the problem, but it boils down to siloed planning and suburb-vs.-city idpol even when there aren’t big income or demographic differences.
People can lead rich rewarding lives in the suburbs without going to the city. Even if the suburb is the quasi gated community of Forest Hills Gardens. Where, I know this gives you the heebie-jeebies, they could afford rent on 108th and Amsterdam but choose Forest Hills. As could the ones in Bayside or Riverdale. Who chose those instead of 108th or Forest Hills. And apparently this it difficult for you to understand, they looked at a schedule before making that decision. And decided they could survive being that far from Zabar’s.