In the previous article in this series, we examined plans for through-running between different railroads serving Penn Station, but not a different type of through-running: trains running through Midtown Manhattan between Penn Station and Grand Central Terminal (GCT), the home of Metro-North and its lines. They serve Westchester and counties north of there in New York State, as well as Connecticut, as far as New Haven and with branches to New Canaan, Danbury and Waterbury. Historically, the lines in New York State (Hudson and Harlem Lines) were part of the New York Central Railroad, while those which serve Connecticut and the part of Westchester County on the way to Connecticut, were part of the New Haven Railroad.
In this article we will look at efforts to link the two busy Midtown stations together, so trains could stop at both. The idea, put most simply, would allow riders heading toward the City to have a choice of going to the East Side or the West Side, whether they came from areas served by New Jersey Transit (NJT), the Long Island Rail Road (LIRR), or Metro-North.
No Easy Trip
Midtown Manhattan is well-defined; both by its boundaries and its central dividing line. Its northern boundary is 60th Street; a dividing line now monitored by cameras and E-Z Pass readers that collect the newly implemented Congestion Pricing toll. The southern boundary is not defined quite as specifically, but there seems to be some understanding that it’s 34th Street, with a slight southerly dip along Seventh to Ninth Avenues to include Penn Station. Some people say that it extends several blocks further south. While there has been some recent effort to build office buildings on the “Far West Side” west of Penn Station, most of the office space in Midtown is located northeast of Penn Station, and there is more office space that surrounds GCT than is located within a short walk from Penn Station. The east-west dividing line everywhere north of Washington Square in the Village (New Yorkers never call it “Greenwich Village”) is Fifth Avenue. Numbers for addresses extend eastward and westward from there, and the East and West Sides at any latitude in the City are generally considered different places.
In his 1960 novel The Night They Raided Minsky’s, Rowland Barber noted that much commerce and other activity in Manhattan runs on the north-south axis along the avenues, while much less runs crosstown. That remains accurate today, as demonstrated by the subway system. Four trunk lines have crosstown components only when heading to one of the “outer boroughs.” Only the historic BMT line such a northwesterly component in Manhattan. Only the L Train has full crosstown service on 14th Street, and there is also some in Midtown, as several lines wend their way to Queens. No subway line goes directly between Penn Station and GCT to deliver a one-seat ride between the two stations.
So how do riders get between them? It’s not easy, and planning to catch the second railroad segment of a trip less than 30 minutes after the first segment arrives at the “other” station is foolhardy. The stations are more than one mile apart and walking between them in 25 minutes is almost an athletic feat. The only “one-seat ride” on the subway between Penn Station and the East Side office towers is on the E Train to 53d Street and Lexington Avenue. There is the Q32 bus from outside Penn Station on 32d St, which runs north on Madison Avenue and stops one block from GCT, but that route does not work as well going toward Penn Station. So, for many riders, the alternative to walking or springing for a taxi is a two-seat subway ride: from Penn Station north to 42d Street, and then east to Grand Central.
Amtrak gave its own riders the exact same advice, as difficult as it might have been to take. When “America’s Railroad” moved the Empire Service trains from Penn Station into GCT for a summer, Amtrak issued a flyer that advised riders going upstate to walk, take a taxi, or a two-seat ride. The flyer included a street map with no indication of any transit, and there was no shuttle bus. Empire Service trains used GCT regularly until 1991.
So, it makes sense to connect the two stations if that can be done. For riders going between the LIRR or NJT and Metro-North’s catchment area north and northeast of the City, or vice-versa, their core benefit would be that the train they catch at their point of origin could take them to either Penn Station or GCT. They could then catch their connecting train in the same place where they got off, without having to make the trip across town, and having to traverse ten blocks on the north-south axis. Penn Station riders on NJT and the LIRR would have easy access to the East Side of Midtown. Similarly, Metro-North riders would have easy access to the West Side. Within the City, they would have direct transit connections to their destinations, or a short walk if they are going elsewhere in Midtown.
Early Proposal
If there is one advocate and planner who has earned the longevity prize for advocating for service to both sides of Midtown, it’s George Haikalis. Now 89, he is still active, as he continues to promote transit connectivity through his organizations: the Institute for Rational Urban Mobility (IRUM), Vision 42 (promoting light rail on 42d Street), the Village Crosstown Trolley Coalition, and the Regional Rail Working Group, which consistently advocated for a unified approach to rail service in the region, rather than individually-centered approaches to service on NJT, the LIRR, and Metro-North.
Many of the studies that Haikalis conducted over the years, as well as other documents pertaining to rail service in the region can be found on the IRUM website. One of his earliest studies (download below) was promulgated in June 1970, when he was a Research Engineer with the Transit Evaluation & Development Division at the now-defunct Tri-State Transportation Commission. The study, titled Improved Access to Manhattan’s East Side – Suburban Railroad Passengers, called for relatively modest infrastructure changes (by today’s “megaproject” standards) that would allow trains to run between GCT and Long Island, as well as between GCT and Penn Station.
In the Introduction, Haikalis wrote: “Each weekday many thousands of Long Island Rail Road passengers bound for the East Side of the Manhattan CBD must transfer to subways or buses for the mile-long trip from Penn Station. Considerable travel time and wear-and-tear on passengers could be saved if Long Island trains were routed directly to a new East Side terminal” (at 1). He then mentioned an MTA proposal to build a terminal on Third Avenue at 48th to 52d Streets, using the 63d St. Tunnel in Queens. The “new terminal” that was later built deep below Madison Avenue, uses the 63d Street Tunnel, and opened for service in January 2023. Haikalis continued: “A preliminary analysis of their extraordinarily expensive scheme indicates that the benefits fall far short of the costs. A modification of the MTA plan is outlined herein which will increase the benefits to Long Island passengers bound for the East Side, provide benefits to suburban passengers from other sectors, and cost considerably less.”
Haikalis summarized his plan this way: “East Side access from Long Island can be obtained by connectiang the 33rd Street L.I.R.R. Tunnel to Grand Central Terminal, as shown in Fig. 1 (below). The 32nd Street Tunnel would continue to serve Long Island trains bound for Penn Station. The portion of the 33rd Street Tunnel from Park Avenue to Penn Station would not be used for L.I.R.R. Trains and could also be connected to Grand Central Terminal as shown in Fig. 1, providing a direct Penn Station – Grand Central route” (Id.).
The proposal suggested extending tracks south of GCT under Park Avenue, to 33d Street, 60 feet below existing buildings. Haikalis proposed a four-track railroad, saying: “approximately 6000 feet of double-track tunnel is required; perhaps $80,000,000 would be a reasonable estimate of the cost involved” (Id.). Haikalis’s estimate would have risen to slightly more than $3.85 billion by 2017, according to the Producer Price Index kept by the Bureau of Labor Statistics. That was the year the cost of the “Grand Central Madison” terminal that was eventually built was estimated at $11.2 billion, and the final cost was somewhat higher. The cost of building the current East Side terminal for the LIRR was almost three times the amount that Haikalis had estimated for a project that would have connected the two stations, but the final product that the riders got was vastly different. It is a stub-end terminal, still with no connectivity to any other rail lines, and the time required to get from its platforms to the historic GCT platforms or the Lexington Avenue subway adds eight to ten minutes to a rider’s travel time.
Haikalis defended the cost-effectiveness of his proposal: “The cumulative appreciation of travel time and convenience for Long Island, Westchester-Connecticut, and West-of-Hudson passengers resulting from the suggested improvement would amount to $17,400,000 annually by 1985. Capitalized at 10% per year, the benefits are more than double the estimated cost of the investment” (at 3). Calculating that estimated benefit using the Consumer Price Index, it would amount to almost $33,500,000 today.
We don’t know whether Haikalis’s study helped bring about the MTA’s 1977 vote to approve the East Side Access project that would have brought LIRR trains into GCT. The FTA’s predecessor, the Urban Mass Transportation Administration (UMTA), issued a Record of Decision (ROD) approving the project for funding, but it was never built. David A. Andelman reported in the New York Times on October 11, 1980: “Workers have been laboring beneath the streets of New York City and the East River completing a $160 million tunnel for the Long Island Rail Road that will never carry a train.”
20 Years Later
During the 1990s, there was again hope that the two railroad stations could be connected. The Access to the Region’s Core (ARC) Project, which began its study in 1995, was under consideration, and one of the alternatives would have connected the two stations directly. Project managers had appointed a Regional Citizens’ Liaison Committee (RCLC) and many active advocates from New York City and New Jersey, including this writer, were on it (that was before I started reporting actively on the rail transit beat). Most, if not all, RCLC members favored the alternative that would build some tracks between Penn Station and GCT.
Significantly, the report chose Alternative AA as the preferred alternative. That was the plan to run “Penn Station – Grand Central Through Operation” and is described in detail at Pages 16-22 of the report. The Introduction continued: “This Milestone Summary Report represents a transition from a broad-based investigation of Midtown Manhattan access needs and opportunities to a refinement of the most promising approach moving toward a recommended Locally Preferred Alternative. Phase III of ARC will (1) verify the constructability and operability of Alternative AA with the aid of computer simulation, (2) identify and analyze long-term variants to Alternative AA which may achieve the same aims more cost effectively, and (3) identify and recommend nearer-term solutions to meet the growth in demand at Penn Station. Phase 3 will include assessment of options for constructing the long-term project in stages which would provide a continuum of improvements for commuter rail and Amtrak services converging on Midtown Manhattan.”
A “Milestone Summary Report” of Phases I and II of the ARC Project was issued in May 1999 (download below). NJ Transit, New York MTA and the Port Authority of NY&NJ (Port Authority) were participating agencies. The study considered several alternatives and said in the introduction: “The study confirmed that demand to access the core from east and west will continue to grow. From the east, critical projects including the Queens Boulevard Connection and LIRR East Side Access are advancing. From the west, additional capacity is needed which directly impacts Penn Station New York and its access network.”
Alternative AA (below) was a modified version of the former Alternative A, which also called for through-running between Penn Station and Grand Central. The plan called for two more tracks from Secaucus to the approach tracks to Penn Station, a new two-track Hudson Tunnel, the Secaucus Loop to provide one-seat rides from the Main/Bergan and Pascack Lines to Penn Station (and effectively rendering obsolete the yet-to-be-opened Secaucus Junction Station and, to some extent, the historic Hoboken Terminal), a new two-level station at 34th Street for the LIRR and Metro-North to share, extending the six low-numbered tracks at Penn Station that are primarily used by NJ Transit, a new West Side Yard for Metro-North (and later built by the LIRR), and use of the 63d Street Tunnel under Queens (at 16).
Significantly, the report stated: “ARC determined that construction and operation of Alternative AA appears feasible” (Id.). The report continued: “In addition, a conceptual service plan has been developed that supports all three railroads gaining access to Grand Central and Penn Station, although unique rolling stock would have to be developed and acquired by NJ TRANSIT … to meet the operating requirement of Alternative AA” (at 17). The report described proposed breakout from the south side of Grand Central and how the new line would go to Penn Station, (at 17), operability (including peak-commuting hours) (at 18), new all-mode rolling stock (at 18), capital costs (estimated range in 1997 dollars of $4.1 to $4.6 billion (at 19-20), operating and maintenance costs (at 21), ridership forecasts (at 21), and passenger circulation at Penn Station and Grand Central (at 21-22). The ridership forecast for 2020 called for an increase of 30% over the “no build” ridership number, and said those riders could be accommodated because of the two additional proposed tunnels. NJT riders were expected to break 69% to Penn Station and 31% to Grand Central; within Penn Station’s capacity (at 21). The evaluation continued by saying that there was more to do concerning Alternative AA (at 22-23).
The Big Letdown
Over the next four years, ARC’s planners considered alternatives whose designations had changed. The changes were reported in the “ARC Summary Report 2003,” a 36-page document with five pre-report pages (download below). A modified Alternative AA was still on the books, as was the new Alternative G. That plan would have brought NJT into the Lower Level of GCT, but there was no mention of the LIRR in the basic description (at ii). There was also Alternative P, which would have built a new underground station near Penn Station area without any connection to the East Side, and Alternative S, which would have built a connecting line to Sunnyside Yards to speed throughput of train sets, especially during peak-commuting periods (Id.).
Then came the zinger: “Alternatives P and S, and the near-term capacity improvements 1 and 3 are recommended for advancement into a Draft Environmental Impact Statement. These alternatives provide the highest level of incremental train capacity, accommodate forecasted levels of ridership demand, and provide capacity relief to alternate travel modes” (at iii). The “near term capacity improvements” were linear storage yards for NJT and the LIRR. The former was never built, but the latter was built as the LIRR’s current West Side Yard.
According to the report (at 13), Alternative AA was rejected because “Between ARC Phases 2 and 3, the location of the LIRR East Side Access Terminal had been changed from the Madison Avenue Yard adjacent to the existing Grand Central Terminal lower level to a deep site directly beneath the existing
terminal, with a bi-level configuration. That required modifications to be made to the Grand Central Terminal configuration of ARC Alternative AA.” After describing the modified plan, the report said: “After considerable conceptual engineering and operational cost analysis, Modified Alternative AA was dropped from further consideration for reasons of higher construction complexity, operations risk, and costs” (at 14).
It appeared by that time that the LIRR was out of the picture, by its own action of changing the East Side Access proposal. The report described Alternative G, which would still have brought NJT trains to GCT, in detail (at 15-20, with ridership forecasts at 28-29). The report made its conclusions about Alternative G (at 34) and the other proposals: “Penn Station and Grand Central Terminal were both designed to be expandable to accommodate future operating changes. Penn Station was built to enable Tracks 1-5 to be expanded eastward into a future tunnel under 31st Street. Grand Central Terminal was designed to allow for a breakout from its lower level southward. Alternative G was configured to take advantage of these opportunities, accommodating in concept both Metro-North and NJ TRANSIT at both stations and providing one-seat service to both the east and west sides of Midtown Manhattan for their customers.” Then the tone of the report suddenly changed: “there would be impacts on NYCT subway structures and the operations support systems at Grand Central Terminal. Uncertainties over the extent to which these impacts could be mitigated could not be resolved satisfactorily during the Phase 3 effort. Alternative G would offer the smallest incremental increase in trans-Hudson train capacity among the alternatives and create complex train operations that would affect the operational reliability of the respective railroads.”
In the third paragraph from the end of the report, ARC planners gave connectivity the ax (at 36): “The drawbacks related to Alternative G (concerning the risks surrounding construction and operations) indicate that the alternative, as devised, is not acceptable for implementation. As a result, Alternative G is not recommended for further development in this Draft Environmental Impact Statement.”
The advocates were dumbfounded when they got the news. They found it difficult to believe that the alternative for which they had pushed for decades had sunk form the “Locally Preferred Alternative” to elimination in only four years, when the only recited change in the facts was that the LIRR had decided to build a deep-cavern terminal under Maidson Avenue, west of GCT. The game seemed over, with the railroads going their own way, and with apparent disregard for the riders who wanted access to both the east and west sides of Midtown Manhattan without the hassle of spending the possible equivalent of the train fare for a taxi ride between railroad segments (in addition, there is no fare integration between NJ Transit and the railroads on the New York side and little between Metro-North and the LIRR), take a two-seat and two-stop ride on the subway, or have no other choice but to walk between the stations.
Calls Continue
I covered the opening of the new LIRR “Grand Central Madison” deep-cavern terminal 25 months ago. One reporter asked if the more than $11 billion spent on the facility couldn’t have been better spent on improving the subways, instead. The featured guests, Gov. Kathy Hochul and MTA head Janno Lieber, ignored that question. Still, the facility seems to be underperforming. Ridership has not met expectations, some trains have been moved back to Penn Station or Brooklyn to meet demand, and actual choice of departure times to many destinations from either Penn Station or the new terminal has not increased under the new schedule. Some riders have complained that service is worse.
It has now been more than 21 years since the connection between the stations got thrown under the proverbial bus. Still, advocates like Haikalis and Joseph M. Clift, former LIRR Planning Director, still call for that sort of through-running. It’s not just advocates who want it, though. Mysore Nagaraja (former head of MTA Capital Construction), Howard Sackel (a former Amtrak and Port Authority Director at ARC, and Bob Prevedi (a former planner at NYC Transit) co-authored an OpEd in the New York Daily News on July 3, 2023; five months after the LIRR East Side deep-cavern station opened. They touted the success of London’s new Crossrail/Elizabeth Line and said: “This does not reflect well on the LIRR’s new $11 billion East Side Access project, nor Amtrak’s proposed $40 billion Gateway plan. The public needs to demand something more ambitious and connective like the Elizabeth Line if New York is spending more than $50 billion.”
The OpEd explained: “Grand Central Madison’s shortcoming is that it only helps Long Islanders reach East Midtown Manhattan and does so by requiring the LIRR’s scheduling department to split the destinations from each branch, making what was previously a train every 30 minutes on the Port Washington and Babylon branches into hourly train service.” Then, sounding like the advocates who have been saying the same thing for 30 years, they said: “The biggest problem here is the parochial thinking of Amtrak and the commuter rail agencies and lack of political leadership that puts customers’ needs over agencies. Someone, the mayor, the governors of New York and New Jersey or perhaps Transportation Secretary Pete Buttigieg or even President Biden, ‘Amtrak Joe,’ needs to step up and champion this kind of thinking.”
They never did, but now it’s up to POTUS 47, Elon Musk and Sean Duffy. Could they “step up and champion this kind of thinking”? While nobody else can know for sure, it seems very difficult to believe that they would. So, as even Amtrak advises, there are three ways to get between Penn Station and GCT: take an expensive taxi ride, take a two-stop, two-seat subway ride, or walk. That’s just the way it is.
There are no more case studies to examine, but we will have further thoughts about through-running and other “transit railroads” in the next article in this series. Then we will conclude the series with some final comments on through-running, what it has done, what it can’t do, and its future.
