Connecting New Jersey to Lower Manhattan
In my regional rail series, I proposed a new tunnel connection from Hoboken to Lower Manhattan, allowing regional trains to use the line and serve Manhattan and continue to Brooklyn on new track. I would like to revisit this concept, in light of my more recent post about where the Lower Manhattan station could be located. Hoboken is just one of many former railroad terminals on the west side of the Hudson, and there are alternative locations in Jersey City.
The importance of the connection is threefold. First, it’s a potential relief line for the near-capacity North River Tunnels: not as important as quadrupling the tunnels, since Lower Manhattan is a smaller CBD than Midtown, but still useful. Second, it’s a direct connection from New Jersey to Lower Manhattan, making life simpler for travelers who don’t have access to the PATH transfer at Newark Penn. And third, it’s useful for serving jobs in Hudson County, which has sprouted a secondary CBD in Jersey City; currently trains can only reach inner Hudson County from the north and west, or from the Manhattan CBD, but not from Brooklyn or Queens. It is the third aspect that makes through-running valuable – none of the commute markets crossing Manhattan is large on its own, but all combined have about 200,000 people among them.
Since the secondary CBD is in Jersey City and not Hoboken, the Jersey City options should be explored in addition to Hoboken. Since my concept for how New York regional rail should look is something like the Paris RER, let me draw the following analogy. The Ligne de Vincennes, which became the eastern branch of the original RER A, had its Paris terminus at Bastille. However, due to SNCF pressure to veer off and serve a more southerly location at Gare de Lyon, it was cut off near the city line, and its route into the city follows a different course, while the original route was abandoned. The upshot is that existing train stations that are at inconvenient locations can be left off the mainline, or closed entirely.
Thus, we should not be wedded to keeping a regional station at Hoboken, whose primary advantages are merely the PATH and light rail connections and the large railyard. Although the railyard may seem important, it’s pointing in the wrong direction for trains from New York. For short-turning movements on the shoulder of rush hour, railyards should make it easy for outbound trains to veer off at a location just after where demand drops off – for example, east of Jamaica, and west of Newark or Secaucus or eventually Paterson.
Many of the original rights-of-ways used by the railroads to deposit travelers across the Hudson from Lower Manhattan still exist. Of course, given the cost of constructing a new station in Manhattan, the cost of Jersey City construction should be regarded as secondary, though non-negligible. But an entirely greenfield option using new tunnels would have sizable cost: at about 3-4 km from where today’s Erie trains turn due east to Hoboken to the Hudson; at normal-world construction cost this is $1 billion, compared with about $2.5 billion for ARC before the cavern’s costs exploded.
Based on the need to leverage existing rights-of-way, there are four options. For distance calculations, it matters very little where the Lower Manhattan station is, but it is notionally measured from the point where Erie trains turn east to the Hoboken station to Broadway and Fulton; the straight-line distance is exactly 6 km.
Description: trains follow the existing route to Hoboken, but start descending as they exit the tunnel to the east, and go under the existing station, next to the PATH station. They cross to Manhattan on one side of the PATH tubes, then turn south to Lower Manhattan.
Stations: new station under Hoboken Terminal, and a Greenwich/West Village station.
Length: 8.5 km.
Manhattan station preference: any, but with north-south preference.
Advantages: less construction in New Jersey, Hoboken has a spacious railyard for station construction, easy transfer to PATH and light rail, plenty of space in the Village valley to align along the proper streets for service to Lower Manhattan, possible transfer to the West 4th Street subway station.
Disadvantages: longest, too far from Jersey City jobs, passes very close to the uptown PATH tubes and must cross under them somewhere, more construction in Manhattan, the most useful Village station requires crossing the three-level IND West 4th station or one of its two-level approaches, most north-south alignments require crossing the Holland Tunnel or an east-west subway to Brooklyn (the more easterly alignments more than one).
Description: trains follow the Bergen Arches into Jersey City, go underground in the vicinity of Jersey Avenue, and enter Manhattan at Canal Street, turning south close to but without intersecting the Holland Tunnel.
Stations: Pavonia/Newport, possibly Chinatown, possibly a future infill station on the Bergen Arches.
Length: 7.5 km.
Manhattan station preference: any, but with north-south preference.
Advantages: space in the TriBeCa valley to align along proper streets, possible transfer to the Chinatown Canal Street subway station, the unused four-tracked Bergen Arches are
elevated in an open cut and thus have space for local infill stations if necessary later, not much construction in New Jersey.
Disadvantages: Newport is not Jersey City’s biggest cluster of towers, the ideal construction location of a Pavonia station is 300 meters from the PATH station and any less requires going under the mall or between tall buildings, a fair amount of construction in Manhattan.
Description: trains cross to Journal Square in a new tunnel, follow the former PRR mainline alongside the PATH tracks, branch off to the PRR Pavonia terminal, start descending at Marin, and cross to Manhattan at a street between Worth and Vesey.
Stations: Journal Square, Pavonia/Newport
Length: just under 7 km.
Manhattan station preference: any works, but east-west under Vesey is exceptionally easy and north-south under City Hall Park slightly less so, and east-west under Liberty is slightly harder than the rest.
Advantages: relatively little Manhattan construction, flexible about Manhattan station location, several easy Manhattan station options.
Disadvantages: the Newport-south tower cluster is smaller than the Newport-north cluster, requires new elevated structures through Jersey City for the kilometer between Newark Avenue and Marin as well as about a kilometer-long tunnel (neither terribly expensive).
Description: trains cross to Journal Square in a new tunnel, follow the former PRR mainline alongside the PATH tracks, go underground when PATH goes underground, and cross straight east to Manhattan on one side of the PATH tubes.
Stations: Journal Square, Exchange Place.
Length: 6.5 km.
Manhattan station preference: any east-west.
Advantages: serves the biggest office building cluster in Jersey City, easy transfers to PATH, all east-west Manhattan station options are relatively easy, least construction in Manhattan, shortest length.
Disadvantages: north-south Manhattan options are difficult, generally inflexible in Manhattan, passes very close to the downtown PATH tubes, needs about 1.5 km of tunnel in downtown Jersey City (for which there are some right-of-way options) and another km of tunnel north of Journal Square, partially duplicates PATH service.
The general theme here is that there’s a tradeoff between construction in New Jersey and in Manhattan, but most likely construction in New Jersey is cheaper. However, the cheaper options involving more construction in New Jersey are less flexible in Manhattan – the Hoboken and Pavonia-North option do not constrain the choice of Manhattan station locations as much as the two PRR options.
For the record, my guess is that the best option is Exchange Place if the Manhattan station has to be east-west, and Pavonia-North if it is north-south.
Alon, “a picture is worth a thousand words”
Would you consider make a quick map for these sort of posts? The one on station in Lower Manhattan could also have benefitted greatly from maps. You can use Google Maps to draw the lines or boxes – I imagine you are already looking at some sort of map to write these posts.
Here is a map of the options. It does not take Adirondacker’s comments into account – sorry, I created it before I saw them.
Do you have maps for any of the other recent, geography-heavy posts, such as the one on regional rail stations for Lower Manhattan?
I don’t have maps for my other posts yet. I’ll make them over the weekend. The Lower Manhattan one may take a while – a lot of options, boxes that intersect one another.
Agreed – I would like to see a map. I saw your series on The Transport Politic and thought it made sense, mostly – would like to visualize these changes.
plenty of space in the Village
Everybody wants to start their trip at an express stop and end their trip at an express stop. If everybody has a trip that starts and ends at an express stop no one has a express trip.
You have to cross the IRT too too unless you put the station west of Seventh Ave South.
The Port Authority, which can more or less do whatever it wants to do whenever it wants to do it, has been trying to add secondary egress to the Christopher Street and Ninth Street PATH stations for at least a decade. It’s not because the Port Authority thinks it would be a nice thing to do, it’s a fire/life/safety issue, the stations only have one exit/entrance. Yes there’s plenty of space. If the Villagers got out their torches and pitchforks over needed egress they are going to go ICBM ballistic over a four track station. And it sucks up time for people going to Wall Street making the trip less attractive. Remember Manhattan geology too. The bedrock dips between 29th and Chambers. Makes it much more difficult to carve out a station under historic buildings, most of the Village is a historic district. Or a tunnel for that matter.
You wanna go to the Village, change to the subway or PATH. You wanna go to Journal Square, Hoboken or Exchange Place, change to PATH at Newport. It’s an RER system not the neighborhood bus.
the unused four-tracked Bergen Arches are elevated
The are in a cut under the local streets. The bridges for the local streets over the cut are arch bridges which is how it got it’s name. It’s west of the Turnpike, parallel to Hoboken Ave. That green leafy thing in the satellite images. Lurking next to that ROW is the active freight tunnel that was the original Erie ROW and the “covered roadway” going to the Pulaski Skyway.
There were lots of ROWs to the waterfront. My chronology may be wrong. Predecessor to the Pennsylvania makes the first cut to their ferry terminal in the general vicinity of what we now call Exchange Place. They share it with the DL&W. The Erie tunnels to the general vicinity of Newport. The DL&W tunnels to Hoboken. The Erie tunnels don’t have enough capacity so they cut the Bergen Arches. Somewhere in there the PRR builds the viaduct parallel to 6th street for freight. The CNJ doesn’t dig tunnel or blast cuts. They come in from the south and build the terminal in what’s now Liberty State Park. The other railroads come in from the north clinging to the edges of the Palisades.
the ideal construction location of a Pavonia station is 300 meters from the PATH station
If you are launching the TBM west of the Turnpike Extension it doesn’t really matter much where you are tunneling to or that there’s an abandoned ROW 100 feet to the south. The ideal site to construct a station at Newport is the place where it connects well with PATH.
The problem with Newport back in the Erie days was that the Erie had planned on moving the terminal closer to the H&M station but never got around to it. Newport may be better placed than you think. There’s always the option of using the abandoned ROW on 6th to do cut and cover. Personally, to channel Cap’n Transit for a moment, put the suburban mall at Newport out of it’s misery and tear it down. Build something more urban in it’s place with the commuter rail station under it. Definitely redevelop the surface parking lots.
Whatever it is, it’s going to be deep. The H&M put flying junctions in somewhere between Hoboken, Newport and the uptown tunnels. Flying junctions between Grove, Newport and Exchange Place too.
You don’t need new tunnel until you are east of Journal Square. The ROW is at least 4 tracks wide. The PATH ROW is the same ROW that PRR used to get to the ferry terminal at Exchange Place. Alternatively all the trains from Newark, either Penn Station or Broad Street, could use the ROW to Hoboken to get to the tunnel to Wall Street. Nothing going through Journal Square that way. Plenty of space out in the Meadows for 6 or 8 tracks and flying junctions. Only one tunnel through Jersey City.
Ugh. Yes, you’re right, the Bergen Arches are in a cut. The photos I saw were unclear and for some reason I thought they were above ground. Not that it changes too much – an infill station is only slightly harder in a cut than on a viaduct.
As for the Journal Square tunnel, if you’re going from Secaucus, the alternative is to cross the Hackensack River twice, or maybe go around and through the western margin of Jersey City. Honestly, tunneling 1 kilometer is not a major cost by the standards of going under the Hudson.
The reason for the secondary stations is to make it easy to transfer away from just the main stations. (Also, with Journal Square, it avoids the reverse-direction transfer at Pavonia). The station spacing is quite wide, in the style of the RER, which makes multiple Paris stops, just at a spacing closer to that of an express New York subway line than to that of a local line.
Finally, my Google Earth tourism impression is that wherever on the waterfront the line hits will have to have the parking lots developed. The mall can even stay, if all the empty space for cars around it gets converted to new office towers. The same is true of Exchange Place, which on Google Earth looks horrendously auto-oriented for a place where nearly half the households don’t own cars.
Jersey City isn’t the Upper West Side, it’s not even Queens Blvd.
an infill station is only slightly harder in a cut than on a viaduct.
It’s the RER not the local bus. Plenty of space to run a trolley line along it but people in Suffern don’t have an overwhelming desire to see the bright lights of Union City and people in Union City barely know where Suffern is much less have the urge to go there.
As for the Journal Square tunnel, if you’re going from Secaucus, the alternative is to cross the Hackensack River twice
Huh? Secaucus is on the same side of the Hackensack as Jersey City. There’s a wide straight flat currently being used for passenger service ROW within blocks of Journal Square or Newport.
Journal Square is two island platforms with four tracks for PATH now. It has freight and service tracks around that. It’s where you do the cross platform transfer from the World Trade Center-Newark trains to the Journal Square-33rd Street trains. You don’t need tunnels to go under the Journal Square Transportation Center.
You wanna haul a train from where all the lines to Hoboken come together to Journal Square use the ROW that runs parallel to West Side Ave. You can even see where the wye that used to be there has been taken over by parking lot just north of the PATH line. Plenty of space. Or use the one that’s west of the Turnpike extension. Stopping at Journal Square eats two minutes for 98 percent of the people on the train. Let ’em change at Newark or Newport for Journal Square. The only thing stopping in Journal Square does for you is avoiding the stop at Grove Street. Not worth much.
The reason for the secondary stations is to make it easy to transfer away from just the main stations.
Westbound through passengers can change in Jamaica or Brooklyn, Eastbound passengers can change in Jersey City, Newark or Secaucus. You go all RER-y the RER terminals are in places like Valley Stream and Rahway, the passengers can sort themselves out there. And can use an express between any of the three destinations in Manhattan and the RER terminal.
It would all be part of grand plan to give suburbanites three destinations in Manhattan wouldn’t it? The train that runs local from Hartford to New Haven can run express to Stamford because it met a train running local between New Haven and Stamford in New Haven. At Stamford it met a train running local to New Rochelle, at New Rochelle it meets the local RER train that shuttles back and forth between New Rochelle and Grand Central and expresses to Penn Station. At Penn Station it becomes the express to Rahway with local service to New Brunswick. Change at Newark for the local to Rahway. Change at Rahway for the local to South Amboy and the express to Bay Head. At New Rochelle it met an express to Grand Central that continues on to Wall Street. Change at Grand Central for destinations in Eastern Queens and Long Island or to the subway for western Queens. Change at Wall Street for not much of anything. Change in Brooklyn for subway service to Brooklyn local destinations. Express through Staten Island to Perth Amboy and change to the local to Long Branch and express to Bay Head but then why send the Harford-Rahway train to South Amboy. Send it to Plainfield. The SIRT ROW is wide enough for local and express service. You can almost spit at the Perth Amboy station from the Tottenville station.
You can play the game from Long Beach to Dover too. Local to Valley Stream, express to Wall Street not even stopping in Jamaica. Change at Valley Stream for the subway to Jamaica and the infill subway stations in eastern Queens. Change for the express to Grand Central that came in from the Far Rockaway branch and the express to Penn Station that came in on the Babylon Branch which also give you express service to Jamaica. It stops in Brooklyn, Newport and Newark where it meets the local to South Orange that’s coming in from Penn Station. Followed by the express to Gladstone. It gets to South Orange where it meets the local to Summit, Chatham or Madison and expresses to Morristown and then runs local. There’s about a gazillion ways this can be arranged. You don’t have to change trains in Manhattan except maybe in the dead of night when the trunk is running once an hour with shuttles to the branches….. Local between Philadelphia and Trenton and then express to Rahway, Newark and Penn Station. Change at Rahway for the express to Wall Street and Long Island points….
Journal Square isn’t the Upper West Side or Paris, but it might just be Vincennes or Saint-Denis. The point is that it can get 2 stops for regional transit. It’s not that big a deal; Newark gets two stops on the NEC (the airport station should be closed and replaced by an AirTrain extension to downtown, but that’s another matter; Newark could support an extra station elsewhere). At the speeds of the inner portions of regional rail, i.e. about 100 km/h, the cost of a stop is about a minute. On the Harlem Line, which is mostly 60-65 mph, the current time cost of 12 stops is 13-15 minutes judging by the difference between local and express runs. It’s not like the LIRR Main Line or the Jersey NEC, which can support much higher speeds.
The point of a decent regional service pattern is that it looks like a subway with wider stop spacing on the shared segment. In particular, if there are express trains, then they make predictable stopping patterns, without local-then-express-then-local schedule innovations. The ideal is to have four tracks and then the local/express division is just like on the subway – preferably more the IRT than the IND, but that’s another matter. It’s fine to have a train run local at the outer end and then express to the center, but it should make all the major stops in between, and this includes Jamaica and Flatbush; those are important job centers (Downtown Brooklyn looks like a bigger job center than the Place de la Nation), and they’re nifty as places where people can connect from the subway to suburban jobs. The importance is that this concentrates frequency: it ensures decent frequency of local service, and especially high frequency on the thickest markets. With short-turn trains and trains that skip a few stops (but generally not proper express or limited trains), most RER lines combine 4 or more different service patterns, which meet in the center to form a trunk line.
Plopping a regional rail station at Kennedy Blvd and Hoboken Ave isn’t putting a station at Jersey City’s equivalent of Columbus Circle, It’s plopping a regional rail station down on the Montauk line in Glendale or Richmond Hill on the Rockaway Cutoff.
I know, I know. The only reason to put a station there (or at any of the nearby north-south avenues) is if there’s a north-south urban rail link on the Hill to connect to. Otherwise, you’re right that there’s no point.
Stay on the bus for four or five more blocks and Journal Square where there’s some actual there there.
The Southernmost option could have a branch going through Bayonne on whatever surface ROW there is, to go to Staten Island. Maybe that would be 2-3 Km longer than the Brooklyn option, but could be much cheaper (almost no tunneling).
Regarding the ‘RER ain’t no local bus’ comment: consider that the RER Paris has station distances of like 1.5-2.5km in downtown. How far away from the city do you want to go?
Jersey City isn’t Paris.
Right, I forgot New York is exceptional.
Btw, London Crossrail has 6 urban transfer stations underground, with station distances of like 1.3-1.8 Km. So that’s going to delay through-routed trains by less than 5 minutes compared to a single downtown station, but quadruples the number of possible connections, and thus reduces the average travel time for passengers. Also, a single downtown station would have capacity issues both in terms of passenger flow, and in terms of dwell times to keep up 24tph – so would have to have multiple platforms per direction.
There’s a real problem with putting a lot of stations in the CBD when you go east-west. (North-south is something different, of course.) Lower Manhattan just isn’t that big – it’s 1.2 km from the southern end of City Hall Park to South Ferry. Midtown should get an East Side and a West Side station on the same line, for some trains, but that’s another debate.
That’s one reason why when I wrote the original pair of posts I tried to set up the preliminary line network to put as many transfers as possible at Secaucus, Sunnyside, and Jamaica – the idea is to minimize passenger flow at the CBD stations. This can also be done with good platform design, which means a reasonably wide platform with a lot of escalators coming out, Chuo Line-style, but it’s a good idea to avoid the platform crush loads of Tokyo.
Once an hour service from each of the branches that passes through Valley Stream means there’s four trains an hour from Valley Stream to Wall Street and four trains an hour to Penn Station and four trains an hour to Grand Central. Just a tad of schedule legerdemain and they can change to Wall Street from the Grand Central train or to Penn Station from either in Valley Stream Cross platform or same platform, No running up and down to the concourse at Jamaica. Wanna get really fancy Valley Stream has three tracks and two islands and all the transfers are cross platform. The M and R get extended to Van Wyck Blvd and the E and maybe the J go down the Locust Manor branch and the F goes down the St. Albans branch. You can change to the subway in Valley Stream. Means the infill stations in eastern Queens don’t slow down the people from Far Rockaway West Hempstead Long Beach and Babylon.
With enough traffic you are doing the Wall Street-Penn Staton-Grand Central fandango in Hicksville and the Penn Staton-Grand Central jitterbug in Stamford.
….once an hour to each destination so the people who want to drive everywhere all the time can’t use “but I have to transfer” as an excuse to drive.
It is not clear to me why another tunnel to Lower Manhattan from New Jersey is needed. As Jarrett Walker in his Human Transit blog frequently notes, connectivity is part of the success of a good urban transit network. Not every transit trip has to be a one seat ride.
Unlike midtown Manhattan which has a fair amount of off-peak and weekend travel from New Jersey and where the two tubes between New Jersey and Penn Station constrain growth on both Amtrak and NJ Transit, there is no capacity problem on the dowtown PATH service from either Hoboken or Newark off-peak and weekends and, except for a few peak trips, not from Hoboken during the peak. (Newark-World Trade Center does operate at peak over a long commuting span.)
All NJ Transit lines have service to either Newark Penn Station or Hoboken Station to connect to PATH. Travel time between Hoboken and World Trade Center is 10 minutes on a peak frequency of about eight minutes. And Hoboken also offers ferry service to two Lower Manhattan locations.
And, the Hudson Bergen Light Rail line and PATH provide distribution from Hoboken to the major develoments near Newport, Pavonia, Exchange Place, and Grove Street in Jersey City.
Yes, there are people traveling between New Jersey and Brooklyn, Queens, and further out on Long Island. But for those with a good subway ride into the vicinity of the World Trade Center, the high frequency of the subway combined with the high frequency of PATH will work much better than a less frequent commuter rail service. There are better places to spend transportation dollars.
I don’t think the problem is that people one one-seat rides. Rather, people are not going to do 4 or 5 seat rides – like if you went from somewhere in New Jersey to downtown Brooklyn. Downtown Manhattan stations should be set up to connect to other lines as much as possible, creating a network where many more possible trips are two trip rides.
Jarret Walker calls this The Connection Count Test.
There are better places to spend transportation dollars.
Having spent my share of time packed in like a sardine on PATH trains – at all hours of the day, the Port Authority is very very good at load management and appear to be aiming for 200% – and NJTransit commuter trains, what do you suggest as a better place to spend money?
PATH always crowded? On my visits to the New York area (fairly frequent), I am often on PATH. I have never found it crowded except during peak periods or service reductions due to track work. The secret is to know which end of the train is less likely to be crowded. Granted, the Newark line is crowded over an extended peak period unidirectionally. But most of Alon’s proposals would not alleviate overcrowding on this route.
Where to spend the money? I still believe more capacity from New Jersey to midtown Manhattan is a higher priority than more capacity to lower Manhattan. This can include both more rail capacity and improved bus access, including a badly needed bus garage/midday storage facilty adjacent to the Port Authority Bus Terminal.
And there are many projects in the New York region that make more sense to me. For example, completion of the 2nd Avenue Subway; extension of subways selectively to areas not served in Brooklyn and Queens, rail to LaGuardia.
You’re right, some of these projects are worthier. Second Avenue Subway is worth gold, and is okay even at present construction costs. The same is true of some other lines, some formally proposed and some not: 125th, Triboro, ARC Alt G, and Utica are high on my shortlist. But the Lower Manhattan rail link is still higher-priority than some of the other projects being bandied about. Rail to LaGuardia is a certain underperformer: airport-oriented transit doesn’t get as much ridership as the business class thinks, and even at the high mode share of Heathrow, we’re talking about 15,000 riders per day. If they can extend the subway east of 31st and Ditmars, they should just go east on Ditmars/Astoria, serving dense neighborhoods that lack subway service. Similarly, the 7 extension to Hudson Yards is a vanity project. And although ARC done right would be amazing, all of the proposals so far – ARC Alt P-Cavern, Gateway, and even the 7 to Secaucus – have major shortcomings, coming from inability to fix agency turf issues.
The approach of building a Lower Manhattan link is pretty much the opposite of that of building bus garages near Port Authority. It’s not so much about rail vs. buses as about what to do with expensive CBD space. The highest-value use of the land is high-rise office space; that’s how CBDs form. Railyards, bus garages, and other parking facilities are at the bottom of the list; one advantage of rapid transit is that it can serve the CBD, move on, and park at a place where land is not as expensive – when the organization is done right and the transit is rapid, i.e. not a bus or light rail, the CBD is transparent, allowing it to move across without suffering the same operational nightmares one would expect if buses had to terminate a few kilometers beyond the bus terminal.
Alon: A further comment on why I consider the bus garage (actually a Port Authority Bus Terminal annex) an essential project.
The PABT is, by most accounts, the second busiest bus terminal in the world. In the PM peak it barely functions; such that buses coming from New Jersey are often diverted onto city streets because there is no room in the terminal. This increases operating costs for the bus companies, adds congestion to the streets of New York, and is environmentally wasteful. If the buses could stage in a PABT annex, these problems could be addressed. And studies have shown that even if ARC were built, bus demand to midtown would remain high.
Secondary benefits include some midday bus parking (for buses used only during the peak period reducing some congestion through the Lincoln Tunnel) and parking for charter and tour buses while their passengers are at theaters or otherwise in the midtown area. The air rights over the garage could be developed as office or residential space.
And a note on rail to LaGuardia: I agree with you that dedicated rail makes little sense; but a subway extension from Astoria or Forest Hills makes a lot of sense given the high employment level at the Airport with shifts around the clock.
Ride PATH at 2am some Saturday morning. That’s more a problem of the Port Authority being cheap than a capacity problem but they are crowded. So crowded that people in the know take the first train in, doesn’t matter if it’s northbound or southbound, if it’s northbound they have a much better chance at getting a seat.
Get all of the people who change to PATH from suburban trains in Newark off PATH and you have alleviated some of the problem. You induce demand because you have taken four stops and transfer out of the trip. More demand means it’s worthwhile running one more train during each hour of the rush hours which induces demand because it’s every 15 minutes instead of every 20. Make it cheap as PATH and all of the people who take the bus to Penn Station so they can go to the World Trade Center stampede away from PATH. Enough space at rush hour that you could have bowling tournaments in the aisles. Maybe not because there’d still be people who want to get to Jersey City, Greenwich Village and Chelsea.
The solution to bus garage space in Manhattan is to put the bus terminal in New Jersey and have people take trains into Manhattan. Or just have them take trains from the place the bus originates. Lungs all over Midtown will thank you. There’s a problem with bus garage space in Wall Street too… put people on trains and you don’t need the buses.
Second Avenue subway doesn’t solve any problems for people in Long Island or New Jersey. Very very marginally useful for people in the Metro North territory because it relives overcrowding on the Lexington Ave Subway.
Rail to LaGuardia serves a few thousand people a day. If you don’t like taking the bus to LaGuardia take the train to JFK or Newark. One of the solutions proposed to relieve airport congestion in metro New York is to close LaGuardia. You can fly more planes into and out of Newark and JFK without contention for airspace from the planes at LaGuardia. Enough that it makes up for the capacity of LaGuardia. Poof you just lost your reason for the low use extension of the subway to LGA.
Those subway connections aren’t all that good. If you want to get from Brooklyn to Jersey City, you need to change at Fulton-WTC, and that’s a very long walk, and the subway’s speed is meant for people traveling to Manhattan, not across it. Add the fact that there are no commuter connection and what you get is that it’s a three-seat ride at best, often even more. Part of the beauty of the integrated timed transfers of the German-speaking world is that a large majority of trips requiring a transfer require just one transfer. There’s a real geometric problem with trying to make cross-Manhattan commutes work with the existing local networks.
Conversely, it’s not a fact of geometry that commuter rail frequency is low. In fact in many cities it’s as high as that of any local transit: Sydney, Munich, Tokyo, Paris, etc. The Bay Area did something similar, just much less usefully than outside the US because of the need for 100% greenfield track and the prevalence of suburban park-and-ride hell. Even in FRA-land, the concept of combining lines together to form a trunk line leads to a high frequency of trains running between Penn Station and Jamaica, making the LIRR a more useful way to get to JFK from the West Side than the subway.
By the way, if for some reason it’s completely impossible to run normal rapid transit on mainline track due to railroad or FRA intransigence, the advantage of the Lower Manhattan configuration is that most of the lines that can reasonably meet there can be severed from the national network completely. These include the Staten Island lines (cut west of Arlington), the LIRR Atlantic branches and maybe even Babylon with some flyovers, the Northern Branch, and possibly some of the existing Erie lines. The Erie Main Line serves dense cities up until Paterson, and with some concrete it may be feasible to turn it into a Ridgewood-to-Manhattan line separate from the mainline system and connect it to lines going to Long Island. Of course fixing the organization is preferable, but since the big concrete project is useful anyway, small chunks of concrete here and there can be used as Plan B.
As a long term event, most people tend to relocate their homes to locations where there commute is tolerable. To me, that means that few commuters (on a relative basis) commute from east of the East River to west of the Hudson River or vice versa. If you eliminate commutes from Hudson County to points east of the East River (at most, bus or light rail to PATH to a subway) or from east of the East River to Hudson County and Newark (in general, a subway or LIRR to PATH or NJ Transit to possibly a connecting bus) the numbers drop considerably.
As for specifics, does the Erie Main Line corridor have a lot of demand for service to points east of the East River? I have not checked journey to work data; but I suspect few along the route live in Jersey and work on Long Island (including Brooklyn and Queens). The character of the communities served also help define a commute. An NJ Transit employee once told me that the Bergen County line stations had higher Hoboken/New York ridership than the Main Line stations. Thus, the question of whether tot turn the Main Line into a RER type service depends on local demand and on funding for RER levels of service.
Yes, a large majority of the through-Manhattan market is actually people working in secondary CBDs in Brooklyn, Queens, Newark, and Hudson County. But what you innocuously call “at most, bus or light rail to PATH to a subway” is actually a very difficult commute. The PATH and subway stations are physically far apart, because they were never built as a combined station. In addition, this three-seat ride commute is also a three-fare ride, but that’s a matter of organization that can be fixed independently.
The speed really is a problem as well. All of those lines are run to compete with the speed of driving to Manhattan and the difficulty of finding parking there, but not of driving through or around Manhattan. The point of this through-running exercise is to reverse the current situation in which Manhattan slows down train riders more than it does drivers. The same is true for lines oriented around other hubs: for example, the HBLR runs in streetcar mode through Exchange Place, and this ensures that even if it’s extended to Staten Island, it’s not going to be very useful for Staten Island-Jersey City commutes.
You’re right that the Main Line has less ridership than the Bergen County Line (see here, and bear in mind that the lines are combined from Ridgewood north). The reason I think that if only one line can be RER-ified it should be the Main Line is that it passes through dense, low-income areas – namely, Passaic and Paterson – and those could have depressed ridership because commuter rail is expensive and doesn’t work for anything except peak commutes to the CBD. Right now, getting from those areas to Manhattan usually involves a bus to Port Authority, a dreary station that always makes me like Penn Station, or across the bridge to a long subway transfer. Their ridership is low today by the standards of the LIRR Main Line or the NEC, but they look like low-hanging fruit. For comparison, the Ligne de Vincennes was only electrified in 1969, when the first new tunnel opened to Nation; it had been shortlisted for modernization before, but they never got around to modernizing it until just a few years before the RER opened.
The PATH and subway stations are physically far apart, because they were never built as a combined station.
14th and 23rd are intimately tied together. 33rd Street isn’t all that bad.
When PATH was built it tied beautifully into the 9th Avenue El and the 6th Ave El. Short ride from Hudson Terminal to South Ferry and interchange with the 2nd Ave El and the 3rd Ave El.
PATH is going to be at the western end of Fulton Transit center. No matter where you put whatever in Wall Street, unless it’s under the Fulton Transit Center, it’s not going to connect well to something. Life is a series of compromises. If everyone gets an express ride to their destination nobody gets an express ride.
Okay, let me clarify: PATH is located far from the subway at Lower Manhattan, where the most useful transfer from Brooklyn is. Queens travelers can take the F or M and transfer, but a) to Newark it’s still a three-seat ride, and b) the F and M are a lot slower than regional rail would be.
And you’re right that some of the transfers at Fulton are going to be nasty no matter what. My repeated comparisons to Châtelet-Les Halles include this possibility: RER-RER transfers are easy, but RER-Métro transfers take a good five minutes of walking from platform to platform. The difficulty of transferring there is so legendary that family members who’ve visited Paris and taken trains there responded “Oh, God” when I told them my conference commute involved changing from the RER A to Métro Line 7. The advantage of this system is that travelers from Brooklyn or Queens to New Jersey would both have faster trips and save a hard transfer.
across the bridge to a long subway transfer.
They aren’t that stupid, if they want to go to Midtown they get on the bus going to Midtown. Most of the buses going across the GW are for people who work at Columbia Presbyterian and all the support around it or to local things that you can get to by walking or the Fordham Bus or the Tremont Bus. … East Paterson is positively absolutely suburban to someone from Washington Heights…. Grass, trees, a place to park the car and hang the wash and a place to put the washing machine without the landlord whining about how the lease bans washing machines!
Queens travelers can take the F or M and transfer, but a) to Newark it’s still a three-seat ride
You forget that Penn Station Newark was “my” station for decades. Transfer to the bus to the suburbs there even when the suburb was Vailsburg.
If I was someplace in western Queens and I wanted to get to Newark I’d work my way to the E and change to NJTransit. Used to be 15 minutes before Shirley DeLibero made the trains run on time by extending the schedules. Cowboy engineer on the Trenton Express would do it in 12 minutes from the time the doors closed in Penn Station to the time they opened in Newark, Wasn’t unusual for it to be 13 on other rush hour trains.
Only very very frugal masochists use PATH to get from Herald Square to Newark. Or you missed the last train of the night and the hour long ride via Hoboken is a better option than sitting in Penn Station for two, three hours. If you still want to be very very frugal but less of a masochist, E train to the World Trade Center. No change in Journal Square and it’s faster.
Sometime PATH is faster, sometime NJTransit to Penn Station is faster sometimes a bus to the Port Authority is faster, Options I had at the end of my block for decades because the bus to Penn Station Newark and the bus to Port Authority stopped there. ..
reverse the current situation in which Manhattan slows down train riders more than it does drivers.
Making them stop at every local stop between Far Rockway and Woodside including Woodside just as effectively slows them down.
No need for every train to make every stop – just major stops. It’s fine for trains to run express from Valley Stream to Lower Manhattan, i.e. make intermediate stops only at Jamaica, Flatbush, and Borough Hall if there’s an infill station there. It lets trains use the route west of Valley Stream as a 130 km/h racetrack. Even an all-stop FLIRT from Far Rockaway to Fulton would have a technical running time of about 36 minutes; running limited-stops between Valley Stream and Flatbush would cut it to 29-30. It’s fast for the stop spacing, which is a little more than 2 km for an all-stops train, but with a few extra percent of schedule pad it wouldn’t be unrealistically fast by the standards of the RER or the Moscow Metro or the Singapore MRT; 130 km/h running when possible and regional trains with high power-to-weight ratio, even the M7, do a lot to speed things up. (The M7 would be maybe 4-5 minutes slower for the local option and 3 minutes slower for the express option.)
Boro Hall, meh. It’s going to be hard to place. While Downtown Brooklyn isn’t Midtown or Wall Street it’s got what seems like a subway stop on every corner and skyscrapers. So you are carving an ESA like cavern under downtown Brooklyn. And then threading entrances and exits through the subway tunnels and wedging egress into the bottoms of skyscrapers. Or putting it under Cadman Plaza which is on the west side of Downtown. Make the connection between the subway and Atlantic/Pacific easier and you can take the subway to someplace closer to work.
“As a long term event, most people tend to relocate their homes to locations where there commute is tolerable.”
What if you work on one side of the NY area and your spouse on the other side? Or if you work on one side and want the option of taking a job on the other side? Long-term relocation trends do not address these problems. The number of people who currently choose these commutes is low – but the number whose life would be substantially improved by being able to choose them is high.
Eric, I think this is usually an issue overlooked by many people on the idealistic side of urban planning and transit: assuming a paradigm that is most applicable to people who are single, young and live in small homes without much stuff – thus making moving in-and-out a relatively easy process.
Whenever you get, as today’s overwhelming majority of 2-adults households, man and woman working or expected to be working, and you throw kids in the middle, merely moving out to live close to your new job (or considering move to get a job out of your easy commute way) is a far more difficult proposition. After all, kids can take a higher toll moving in and out new schools and neighborhood friends often, and there is always a partner in the picture that is not changing her/his job at the same time.
I am sure that many people, if they lived alone without kids and engaged only with work (not with a local church, some other group activities etc), would move far more often to reduce commute stress and time, but that is simply not a viable idea for the majority of population. And, to this point, I didn’t even throw the housing price, school quality arguments in the mix.
Even if this is a bit off topic, I felt like writing because I found consideration for the realities of 2 adults (+ kid(s)) households largely ignored on this and related blogs like Infrastructurist, Human Transit, The Transport Politic, Cap’n Transit etc.
Outside the blogosphere, I occasionally stumble upon some interview or discussion in which planners try to shift the blame for congestion and long commutes (by car or transit, doesn’t matter) because people “insist living in communities that don’t have any jobs for them nearby”.
There is also a quasi-hatred, at least here in Northern Europe where I’m living right now, of families “insisting in living in single houses with private gardens despite a multitude of public parks and spaces offered” etc. However, even if neighborhoods became all mixed zoning, that doesn’t mean the paradigm of the 2-income household would be solved, as long as both worked away from home.
I honestly wished bloggers, specialists and other people in the planning and transit communities explored these issues with honesty. Though not fancy or hyped, realities of living with a working partner and kids are heavy upon location decisions.
Once I read here Alon, when discussing the Tappan Zee Bridge replacement, pontificating that ferries could to the job and people should just not expect to live in one side of the Hudson and work in the other. I’ve found similar positions from other people in regard of projects in Seattle, Bay Area etc.
I think there are three different issues here. Starting from the end, the Tappan Zee is a separate issue: it’s too expensive for the benefit provided, hence my comment about ferries (which was mildly facetious anyway – the old bridge is structurally sound, just higher-maintenance than the state would like). I know damn well ferries are a lower-order form of transportation, and would inconvenience travelers. I just judge the $8 billion construction cost a bigger inconvenience to the state’s taxpayers.
The two main issues here are directions of sprawl versus individual choice of where to live, and how planners approach things. Planners love to berate ordinary people for not behaving the right way; they weren’t any different in 1900 when they wanted sprawl. It’s in the nature of people in authority to look for ways to feel superior to the common people. So right now they consciously talk of environmental issues as a matter of virtuous living versus the vices of consumption; they could easily talk in terms of making it easier to not own a car and taxing or regulating the problems coming from driving, but they leave that to Streetsblog.
The part about directions of sprawl is why there’s a difference between exurbanization and two-body problems. It may be different in truly multipolar regions, as Randstad is, but in regions with a clear core, two-body problems cause people to live in inconvenient places that are not at the edge of the region. Maybe they live in the center and pay very high rent so that one partner can work in Long Island and the other in North Jersey. Maybe they live at a place that’s more convenient for one person and then the other takes the hit and commutes 2 hours each way. Put another way, the social causes that make people live in suburbs like Kearny or Tenafly are not the same as those that make people live in Pike County, Pennsylvania. The former are a necessity of both multipolarity and outright job sprawl, and should be served (which is why I’m gung ho about these regional rail systems); the latter reflect a set of economic incentives and zoning policies that compel people to live 2 hours from Manhattan and an hour away from the nearest edge city.
Thanks for this educational exchange. I rode out to Boissy-St. Leger on the last rush hour from the Gare de la Bastille in December 1969 and in March 1970 repeated the trip from Nation. I think that the most important element of that original RER line was that relocating it provided it with much better Metró connections and an alignment that fit with the express routing through central Paris (BART came to my mind as a comparison when I first saw this plan).
For those who are interested in seeing what SNCF/RER started off with, there are several YouTube videos of the old line and now the new line. To me, this was a wonderful inspiration in the age of BART when we were being told by consultants that we had to start everything over, to see how an old operation could be made useful again. Your proposals make good use of lines that are in place now.
Here’s the address for my video (with vintage Sony casettecorder sound). Other related videos will come up with it: http://youtu.be/YtY8UvQYKgE
Thought experiment: Wall Street to Staten Island via Brooklyn has it’s charms. Gets the few Metro North riders who want Brooklyn into Brooklyn and since Staten Islanders tend to have lots of ties to Brooklyn, Staten Islanders to Brooklyn. Apparently the Fourth Ave Subway line has capacity problems. Park Slope is the kind of neighborhood that will use a subway station or two if it is built…. Four tracks from Flatbush and Atlantic and move the D to the new tracks? There’d be some reworking to be done at 9th Ave but they have the space. Down to 65th Street where the Triboro will be terminating. Interchange with the Triboro there?
Adirondacker12800, As you may already understand, I think that the problem on the 4th St. corridor, there are currently two lines on the express tracks (D & N), and any new line from Staten Island would presumably need to be an express also in order to attract riders from SI. There is only the R line on the local tracks, so the D or N could be switched to the local tracks (current passengers would no doubt disapprove). There might be switching issues to get the SI express back on the local tracks to go through the Montague tunnel (assuming you believe the SI people want to go close to where the ferry has been leaving them) and to get the other (D or N) line back where it belongs.
Your teaser about dropping a new set of tracks from Atlantic/Flatbush of course begs the question of exactly where and to what you would attach it up there. I take it that you would still be using the 4th St. corridor for the SI line.
Perhaps you have heard of another plan along those lines – to spin off a new branch from the IND at Prospect Park or Fort Hamilton Parkway and run a new line south and west through to SI (presumably crossing/connecting with the various other lines on the way). There are excess tracks on the Culver line, but it’d be heading for the Rutgers tunnel (and an already booked up 6th Ave. corridor – would probably need a connection to something else (the 2nd Ave Subway or the Montague St. tunnel).
All in all, I think I’m in favor of some sort of subway connection to Brooklyn from SI, as a first step to a future direct commuter line to a future Fulton Street nexus. The overall price tag is bound to be a lot easier to swallow – tunnel between SI and Brooklyn with connections to existing track/stations at both ends versus tunnel between SI and a co$tly share of a station in lower Manhattan. It would be much easier to justify the direct tunnel after the Manhattan station linking tunnels from Brooklyn, New Jersey and Grand Central had already been built and paid for. Meanwhile, a one seat ride on the SIRT continuing on to lower Manhattan on the same train would probably be a winner compared to a transfer to the ferry and some extra walking/waiting, even if the trip didn’t save much or even any time.
Re-analyze everything with reference to the seawall which is going to have to be built to protect Lower Manhattan from sea level rise (10 feet or more at latest estimates)
And with reference to the seawall which probably won’t be built to protect Hoboken. 😛
A good design will be flood-resistant, which partly determines the choice of station location.
Build the seawall from the Atlantic Highlands in New Jersey to whereever in Brooklyn. Then one from Throgs Neck in the Bronx to Whitestone in Queens. The harbor becomes freshwater… the tide reaches all the way up to Albany so you get Sandy Hook to Albany out of it and Sandy Hook to Passaic. Only need two Panamax locks between Sandy Hook and Coney Island. I’m reminded of the plans in the 60s to make Long Island Sound freshwater. Seawall from Montauk to Westerly and you get all of Connecticut and the North Shore of Long lsland. And a much shorter trip from Montauk to Westerly.
If they really need a seawall, then because the US will have to accept a couple tens of millions of climate refugees, its cities could expect much higher transit use, making cost-effectiveness calculations less important…
Are the non-Hoboken plans supposed to entail the closure of the existing Hoboken station?
Which lines would run to Lower Manhattan? Your maps on The Transport Politic show five branches from it (Main, Bergen, Pascack, West Shore, Northern Branch) which seems like a lot for a two-track tunnel if all are to be reasonably frequent.
I was not thinking about the first question. Most likely, some overflow trains would run to Hoboken. For example, intercity trains to Scranton, initially (the line should be electrified, but it’s more important to electrify the local commuter lines first). It might be left as a stub like Hunterspoint.
The intention is for all five lines to use the tunnel, running at frequencies of 10-15 minutes. If one line has to be dropped my guess is it should be Bergen County because it splits trains from Paterson and Passaic, but you could make an argument for any except the Main. (The West Shore Line may face too much opposition from CSX, the Northern Branch may be given away to the HBLR, the Pascack Valley Line has very low ridership.) 10 minutes should be fine for these branches – that’s the same as the base all-day frequency on the branches of the RER A and B. Off-peak, if the central segment is overserved, they can divert Pascack Valley and/or Bergen County to Hoboken with a transfer to the Main at Secaucus.
If use of Hoboken Terminal were to drop to the level of LIC/Hunterspoint then it seems like a rather poor use of a large chunk of land right on top of a PATH station, and selling the railyard off for development would seem pretty attractive. (Actually, why hasn’t this happened to LIC Terminal?) Maybe it has a part to play in the regional system, but sending some lines there only some of the time seems to go against the consistency and legibility you have stressed elsewhere.
Unrelatedly: “the airport station should be closed and replaced by an AirTrain extension to downtown” Really? 4km of vendor-locked construction (potentially disrupting the NEC and requiring significant renovations at Newark Penn) so that cramped, slow monorail trains can run to downtown Newark? Presumably with no intermediate stops, given the unsuitability of the vehicles and the unattractiveness to commuters of transferring at Newark Penn, so all for a typically underperforming airport link.
Yes, probably in a shrinkage case parts of the station should be closed and replaced with new development. Even if the station is closed completely, it’s still connected to Manhattan and Newark by PATH and to Jersey City by PATH and the HBLR.
Pulling very rough numbers about EWR and JFK AirTrain ridership levels, and arbitrarily setting the benefits of a Downtown Newark connection at equal to raising EWR ridership to JFK levels (a difference of about 10,000 per weekday), the worth of such an extension is around $200 million. It can be done elevated next to the NEC with small property impacts; the vendor lock may raise costs above that. Of course nothing of the sort should be entertained while the region builds busways and mostly-on-existing-track light rail systems for this cost.
Not sure why such an extension would be expected to particularly increase ridership. Anyway, extending PATH seems better as it could use the existing tail tracks and serve an intermediate station at South St Newark.
The main (only?) advantage of an AirTrain extension, as opposed to a PATH extension, is that it’s friendlier to intercity travelers. The question then is whether integrated air/rail tickets are worth this hassle.
The reason it should increase ridership is that either a PATH extension or an AirTrain extension would massively increase frequency for travelers coming from New York.
Would it really be friendlier? It would require them to backtrack if coming from the south. Anyway, I’ve never seen air/HSR ticket integration at sane prices, and it seems especially unlikely to make sense in New York where the airports will have enough trouble just handling local origin/destination demand. Also, air passengers expect things like through-checked luggage that are not really compatible with typical HSR operations.
Wouldn’t the regional rail plan involve the frequency of the NEC itself increasing quite a bit? You’d have four tracks all the way from New York Penn to EWR. If the regional rail headways to EWR are more than 10 minutes off-peak you’re squandering the capacity of the Penn Station tunnels. The marginal benefit of PATH’s frequency on top of that does not seem so great (and PATH should be running in sync with the NEC in Newark for convenient transfers to downtown, not offset from it).