So far in this series, we examined the concept of through-running for local “transit railroads” that serve a city’s downtown core and the suburban areas that surround it. The idea is that trains come into the hub station from one set of outlying points on one rail line and go to other outlying points on another. We reported on how it works in Philadelphia and Toronto, how Los Angeles appears to be preparing for it, and how there are two different concepts of “through-running” in the New York area: one that has been in use for more than a century, and the other that doesn’t seem likely to get started. We have spent the past three articles describing various plans for both types of through-running in the New York region.
One plan we described would bring trains into Penn Station from New Jersey on NJ Transit and out on the Long Island Rail Road (LIRR) and vice versa. Amtrak makes a similar operatiang plan work by running between points south of New York (typically originating or terminating at Washington, DC these days) and Boston and other New England points. In theory, that concept is straightforward, but it would probably not translate to local operation very well. At least for now, the types of equipment that run on NJ Transit, the LIRR, and Metro-North are not compatible with each other. We’ll have more to say about that in the final article in this series.
The other proposed plans would connect Penn Station on the West Side of Midtown Manhattan with Grand Central Terminal on the East Side of Midtown. That is not feasible now, and such an operation would require building a track connection between the stations. It might have happened a long time ago, since such a plan was selected as the “Locally Preferred Alternative” as recently as 1999.
The situation changed in 2001, when the Federal Transit Administration (FTA) issued a Record of Decision (ROD) that called for a new LIRR terminal below GCT, instead of the previously preferred connection between the two Midtown stations (download below). The document said: “As the tunnels approach Grand Central Terminal, they will fan out to provide new tracks and platforms below Grand Central Terminal’s lower-level tracks. … Each platform will have stairs and escalators rising to mezzanine level cross-passageways above and perpendicular to the platforms” (at 2). That was the start of the planning process for the current stub-end deep-cavern terminal and the end of the initiative to connect the existing stations. Joseph M. Clift, former Planning Director at the LIRR, told Railway Age that there was once talk of connecting the LIRR deep-cavern station mentioned here with the deep-cavern station that NJ Transit was proposing at the time as part of the ARC (Access to the Region’s Core) Project, but that idea went nowhere. He also said that such a plan would not have delivered any additional service. In the long run, rather than connecting the two stations and providing a new destination to attract more riders, the LIRR added capacity deep under GCT, while NJ Transit was planning to add capacity deep under Penn Station as Alternative P of the ARC Project, forerunner to today’s Gateway Program.
It took only four years to make the swing away from connectivity complete. Instead of using Madison Yard, which was directly under Madison Avenue at the northwest corner of GCT, the LIRR decided to start on its “East Side Access” terminal far below the street, as described. After long delays and severe cost increases, the deep-cavern terminal finally opened as “Grand Central Madison” in January, 2023. Its performance so far has not met expectations, but it’s what the railroad and MTA wanted, so it was built.
Connectivity between the three railroads serving Midtown seemed feasible at one time, with different plans that could have brought NJ Transit, the LIRR, or both, into GCT to join Metro-North and connect with it. Nonetheless, those plans all fell through, and it now seems highly improbable that the two sides of Midtown Manhattan, with all of its office buildings and the workers who go there, will be able to get off the train from any of the three railroads that serve the area at either Penn Station or on the East Side, and have a short walk to the office or other destination.
As for getting off a train from one railroad and onto one operated by the other, that is now feasible at Penn Station, where it’s physically possible to transfer between NJ Transit and the LIRR. It’s a two-seat ride and not that easy, because neither railroad cares about its riders connecting with a train on the other railroad. It’s also relatively expensive, because there is no fare integration. Each carrier requires a separate fare. NJT and the LIRR could make transferring easier, but that decision would be up to them.
It’s difficult sometimes to discover who made the decision that prevented an innovation like through-running from being implemented. In a column several years ago, I wrote that Metro-North was the obstacle. Howard Permut, who was the president of that railroad at the time, wrote a response denying that his railroad slammed the brakes on the through-running plan. Yet, George Haikalis, who has advocated for this type of through-running for 55 years, told this writer that he talked with Permut, who claimed that he did not want other railroads in GCT, which was and remains Metro-North’s hub. Clift mentioned another theory. He had been informed that MTA head Virgin Conway wanted trains from Rockland and Orange Counties to have access to GCT, rather than allowing NJT into the station.
Clift also told Railway Age “We have spent a generation avoiding the obvious, which is improving connectivity for Midtown Manhattan by providing access to both the East and West Sides. It is, in my mind, professional malfeasance for all the planners that have come since then. For instance, claiming that providing additional capacity on the south side of Penn Station is an improvement.” He added: “We started out to improve access to both sides of Midtown for everybody. What we have gotten is less mobility with every step, with no new places riders can go, and with the inconvenience more time between platform and subway or street. Going from inside Grand Central to far below it was, physically and timewise, a major step down. A further step down occurred when the LIRR discontinued timed transfers at Jamaica, so the apparent (and sometimes actual) frequency of access to riders’ destinations was cut in half because connections at Jamaica are now random, rather than timed.”
Clift blamed “Lack of concern for the customer’s perspective, and emphasis on making projects bigger and, therefore, more expensive.” Even today, he poses the question: “Why spend all the money to have more trains at Penn Station, when the lower level of GCT is still available? There are eight tracks in the center of the lower level that are designed for a buildout going south, so Alternative G from the old ARC Project could still be done somehow. That would allow more trains with a new destination, which is why New Jersey advocates liked it so much.” He added that GCT still has capacity for that project.
While the LIRR’s push for a deep-cavern terminal on the East Side put an end to organized advocacy for a connection between the two stations, Haikalis did what he could to keep the concept alive. Early in this series we noted the exhibit Making the Connection, which Haikalis and his colleages at the Institute for Rational Urban Mobility (IRUM) developed, and which was on view at the Municipal Arts Society and other venues in the area. The IRUM web site, www.irum.org, has a link to its content.
Today, at 89, Haikalis continues to hope that his ideas for mobility and connectivity will someday be built and used by riders. He believes that the Port Authority of New York and New Jersey (PANYNJ) should take the lead in such an initiative and told Railway Age:“I certainly think the Port Authority has the statutory authority to advance this connection. What they’re missing is the will. The PANYNJ is controlled by two people: the two governors. If the governors order the PANYNJ to come up with a plan to connect the two stations, they have access to the talent that would be needed. If the governors had the will, it would be under construction right now.” Haikalis also urged advocates in New Jersey to help do that. He added that the New Jersey advocates should be using the political process to get their governor behind this, and so should the advocates in New York.
So, what is the “real story” behind the failure to build a connection between the two stations? We can’t force anybody to disclose any details about the past that they do not wish to reveal, nor would that be a good idea in this case. Suffice it to say that there are often inconsistent stories told about big projects, and all we know is that a certain result occurred. In this case, the reality is no trains serve both stations.
Anywhere Else?
Earlier in this series, we looked at Philadelphia (where it works well), Toronto (where it makes sense because Union Station is downtown and happens to be in the middle of the line), Los Angeles (where it could happen), and Boston (where it will probably never happen, despite prior efforts). We have already described the New York area situation in detail, but there are a few more places to consider.
In theory, any regional-level passenger-rail system that has more than one line could be a candidate for through-running. Chicago might not be such a candidate, because of the city’s geography around the downtown Loop. The four separate stations used by Metra are all located near the CTA’s elevated Loop and slightly outside of it. There are some projects under consideration to build connecting tracks between a pair of stations, but that’s a different solution to a different need.
Disregarding metropolitan-style subway and elevated lines and concentrating on railroads, what else is there? Not much, although a relatively new single-line operation should be mentioned in this context. That line is the New Mexico Rail Runner Express, which bears some similarity to GO Transit in Toronto, which started on the Lakeshore Corridor before other lines were added. Toronto Union Station is in the middle of the corridor, and trains run through there to both sides of the corridor, where they turn and head back. Rail Runner is similar in that trains go south to Belen and north to Santa Fe from Albuquerque. Belen is on the historic Santa Fe’s Southern Transcon (now part of BNSF). Going toward Santa Fe from Albuquerque, trains run on former Santa Fe track (also used by Amtrak’s Southwest Chief) to a point east of Bernalillo and onto new railroad built along I-25 for Rail Runner service. The line ends at Santa Fe, the state capital, at a station and some lead track on the Santa Fe Southern, now a short line, but once a Santa Fe branch that took riders between Santa Fe and the main at Lamy.
Some trains only run between Albuquerque and Santa Fe, others run only between Albuquerque and Belen, and still others run the length of the line. While the through-running model works on the Rail Runner, the level of service is low. On weekdays, trains run between Albuquerque and Belen only at peak-commuting times, but in both directions. There are eight trains in each direction, divided with half running in the morning and the other half between mid-afternoon and early evening. There are ten trains in each direction between Albuquerque and Santa Fe spread through the day, with an extra train from Santa Fe on Friday evenings. The Saturday schedule calls for six round trips (five trains from Albuquerque to Belen), most, but not all, of which run through from end to end. The Sunday schedule calls for four trains from Albuquerque to Santa Fe and three to Belen. As with Saturday, most, but all run through. More than any other factor, it appears that most of the trains run through for the obvious reason that Albuquerque is the state’s principal city, and it’s located in the middle of the line.
The same holds for the Utah Transit Authority’s Front Runner in Salt Lake City. The line runs between its south end at Provo and its north end at Ogden, through Salt Lake City. Trains run half-hourly for most of the service day on weekdays and hourly on Saturdays, with no service on Sundays. While a few trains turn at Salt Lake Central station downtown, most run through between Provo and Ogden.
Denver’s RTD runs four lines form its terminal at Union Station, but it can’t offer the advantage of a train coming in, discharging and picking up passengers, and then going outbound again. That’s because Union Station is a stub-end station, so short dwell times are not available. Even Amtrak’s Callifornia Zephyr, the only long-distance train serving Denver, backs into the station and has standing time there.
Sound Transit’s Sounder commuter trains run north from Seattle to Everett and south from Seattle to Lakewood through Tacoma. Trains run only at peak-commuting times, so through-running would not make much sense. It might generate some operational convenience, but little benefit for riders, if any.
Besides hosting Amtrak and its General Offices next door, Washington’s Union Station hosts trains but, again, almost entirely at peak-commuting hours. Virginia Railway Express (VRE) runs service through Alexandria to Manassas (Broad Run) and Fredericksburg (Spotsylvania), but not throughout the day. MARC trains serving Maryland run from Martinsburg (W.Va.), Brunswick, and Frederick toward the Nation’s Capital in the morning and back out after the workday, but not at other times. The Camden Line runs to and from Baltimore’s Camden Station during peak-commuting hours only, but in both directions. Only the MARC Penn Line, which runs on Amtrak’s Northeast Corridor (NEC) line runs at times other than commuting periods, although weekend service is limited. Given the heavily commuter-centric nature of service in the region, through-running would not deliver much benefit for riders. Only the MARC Penn Line runs at “off peak” times, and it could not be paired well with a line that only runs for commuters, like the MARC Brunswick Line.
On Aug. 8, 2024, VRE and the Maryland Transportation Administration (MTA), which owns and operates MARC trains, announced that customers who ride on monthly, weekly, or 10-trip tickets on one railroad are now allowed to use the other railroad on the other side of DC Union Station at no extra cost. With current schedules, it is difficult to fathom who would take advantage of the deal for riders who use both railroads to complete a trip. In theory, perhaps an Alexandrian might commute to Baltimore, but that would be a very time-consuming trip, while Washington, DC is much closer to home. Looking toward the distant future, MARC might someday be extended beyond Baltimore to Newark or Wilmington, Delaware, where it could connect with SEPTA trains to Philadelphia. Similarly, VRE could be extended beyond Fredericksburg to Richmond. If both extensions are built and placed into service, they would both go far enough to be useful as a through-running operation, of almost corridor-length. For the through-running model to work well, though, it would be necessary to expand service considerably. That would include service running at regular intervals, seven days a week. That could happen someday, at least in theory, but it certainly will not happen anytime soon.
In Canada, Toronto’s Lakeshore Corridor hosts through-running trains frequently, and there might be other opportunities for more similar operations there someday. The other multi-line Canadian operation serves Montreal. Some lines there use VIA Rail’s (formerly CN’s) Central Station, while others use Lucien L’Allier station, a truncation of the CP Rail (now CPKC) lines that once served the historic Windsor Station, which is still standing, but no longer hosts trains. Through-running is not possible there, and the low levels of service, especially outside commuting times, would not justify the expense.
So that’s it. We have covered all the places where through-running could happen and does not, in addition to the places where such operations are running. We will examine the benefits that a thorough-running operation can deliver and under what conditions, as we conclude this series in the next and final article.
