New York May Have Actually Lost Transit Riders by Building An $11 Billion Train Station

In January, New York’s Metropolitan Transportation Authority unveiled “Grand Central Madison,” a re-branding of the decades-delayed and eye-wateringly expensive East Side Access project that now brings Long Island Rail Road service to Grand Central in addition to Penn Station. For a price tag of some $11 billion—originally slated for about $3 billion—it was supposed to be one of the most consequential transportation projects in the region’s recent history, bringing some 160,000 passengers to the new station every day and increasing service by 50 percent while slashing commute times by up to 40 minutes per ride, according to MTA estimates.

Those estimates are in line with what major transit agencies ought to expect from double-digit billions of investment in new service. Indeed, it is not just expected but necessary, because the entire premise of such projects is agencies borrow lots of money by issuing bonds to build more service in order to attract new riders who will pay off the costs of the project over time. If such mega-projects fail to attract new riders, it could hobble an agency's finances for decades.

Unfortunately, with the ESA project, all signs point to a dud. Three months after full service on ESA started, it is far from clear ESA has resulted in any new ridership at all. In fact, there is some evidence to suggest that LIRR may have actually lost riders as a result of ESA because it made some people’s commutes worse.

“So this big $12 billion everyone billed as saving 40 minutes a day off your commute, people are actually doing the math, looking at the clocks, saying my commute is actually longer,” said Patrick O’Hara, who runs the website The LIRR Today which chronicles the goings-on of the nation’s largest commuter rail service. “In some cases, quite a bit longer.”

In order to make the new schedules, the LIRR had to radically change the way service to its Brooklyn terminal works. As a result, they eliminated timed, scheduled transfers at a major merge point in Queens which many riders relied on to switch trains in a quick, efficient way. Now, riders can end up just missing connections and waiting for the next train.

As tempting as it may be to blame Governor Kathy Hochul who controls the MTA or the current slate of executives running the agency, the fact is they all inherited ESA from previous regimes of governors, officials, and local politicians going back decades. Their task was to finish a project that had languished for years. And the current MTA officials have consistently argued one of ESA’s main benefits is as a resiliency project, creating redundancy for Long Island commuters heading to Manhattan in the case of another Hurricane Sandy-like storm damaging tunnels or even to aid in routine maintenance. This is a real benefit that cannot be measured by ridership numbers.

And there’s no question that at least some people are using Grand Central Madison. In April, the MTA celebrated ESA’s one millionth passenger, claiming an average weekday ridership of about 50,000, or about a quarter of the LIRR’s total weekday ridership. But to put this number in some perspective, the BX12, a bus route in the Bronx, carried 40,000 passengers per day before the pandemic. 

However, these are all lipstick-on-pig arguments that try and make the best out of a project that, at its best, is an awkward fit into the city’s transportation landscape. Not only is 50,000 riders less than a third of what the MTA predicted ESA would transport—a huge miss even taking pandemic ridership patterns into account considering LIRR ridership has rebounded to about 70 percent of its pre-pandemic levels—but it is likely nearly all of them are existing riders who simply shuffled over from Penn Station or Atlantic Terminal, adding little to no new revenue to the MTA’s coffers to help pay for this massively expensive project.

Because transit ridership patterns remain screwy as a result of the post-pandemic situation, it is difficult to say with any precision how ESA impacted LIRR ridership. However, because the MTA operates a separate commuter rail service serving different territory, Metro North, which did not get any new train station and dumps commuters off some 16 stories directly above Grand Central Madison, we have a pretty good “control” group in this natural experiment. And, according to O’Hara’s analysis, LIRR’s average weekday ridership has increased 14.7 percent since ESA opened while Metro North’s ridership increased 21.4 percent. There are ways to slice the data that don’t look quite as dire. For example, in April and May LIRR averaged a higher percentage of its pre-pandemic ridership (74.2 percent) than Metro North (68.5 percent), although the same pattern held true (69 percent versus 66 percent) for the last two months of 2022, before ESA opened. In other words, ridership on the commuter rail service that didn’t get an $11 billion new station looks an awful lot like the one that did.

Another, less scientific way to break down the data is to simply look at a chart of ridership data for the two railroads over time. Normally, a major new service like ESA comes with a ridership bump. Below is a chart of daily ridership figures for LIRR and Metro North since the beginning of 2022 (the chart moves up and down because weekend ridership is lower). One railroad is in red, the other in blue. Can you tell which railroad got an $11 billion service expansion in early 2023?

When asked about these trends and if the MTA is pleased with the return it is getting on ESA, MTA spokesperson John McCarthy said the agency is “happy when we see growth” and “we always want to see more. We have capacity available and we’re going to continue to push for more growth.” He is hopeful that over time ridership will grow as people settle into the new service patterns and more reverse commuting takes place.

For his part, O’Hara is more sanguine. “At best, it’s staying the status quo,” he told Motherboard. “More likely, it’s having somewhat of a suppressive effect.”

How did this happen? How did the MTA manage to spend $11 billion on a flashy new train station, add all this service, and get no new riders? 

Most obviously, Grand Central Madison is only about a 15 minute walk from Penn Station and the project included no other new stations or lines. It was never clear how this would substantially change the calculus for Long Islanders coming into the city in the way that, say, London’s Elizabeth Line has, which cost about twice as much as ESA but added dozens of new stations and miles of track to areas previously poorly served by transit. As a result, the Elizabeth Line has been a smashing success, averaging some 600,000 passengers per weekday and increasing Transport for London’s fare revenue by some $60 million through eight months of service.

Not only did ESA have no clear path for attracting tons of new riders, but it quietly made service for some existing riders worse. To understand why, we must look to Brooklyn.

The LIRR is a branched system, meaning each line fans out as it heads east down Long Island and converges as they approach the city. Before ESA, LIRR had three terminals in the city: Penn Station, Atlantic Terminal in downtown Brooklyn, and Hunterspoint in Long Island City. All but one of the LIRR branches pass through Jamaica station in eastern Queens. While not the marquee destination, Atlantic served about a third of LIRR riders pre-ESA, with its quick subway connections that made it easy to get to Manhattan’s financial district without putting up with midtown crowds.

Prior to ESA, the LIRR had what O’Hara described as a “finely orchestrated act” in which the branches would arrive in Jamaica and execute timed transfers where passengers could walk across the platform to switch to a train going to a different end point. So, for example, if you were on a train terminating at Penn Station but you wanted to go to Atlantic instead, you could easily get off at Jamaica and walk across the platform, losing very little time to end up where you wanted. The same would occur in reverse in the afternoon rush hour to get to the branch line you need to get home. This made service more efficient. As any longtime LIRR rider would tell you, it was far from a perfect system—Jamaica was a notorious bottleneck and choke point that often caused service issues—but the schedule hadn’t been substantially changed for more than 50 years because it more or less worked.

But ESA presented a problem. The LIRR now has four terminals, not three. And the LIRR simply doesn’t have the capacity to run substantially more trains through Jamaica, especially during rush hours, to serve this new train station in Manhattan. So the LIRR decided to turn the Jamaica-to-Atlantic line into a shuttle service, essentially ending direct service to Brooklyn, and built a new platform at Jamaica for Brooklyn-bound trains.

Given the available options, this was not an unreasonable decision. But O’Hara said it was never clearly communicated to riders that ESA would come with this massive tradeoff for Brooklyn-bound riders.

The biggest impact this had was eliminating timed transfers, or when two trains going to different places are scheduled to arrive at the same station—ideally across a platform—so people can easily switch from one to the other. Timed transfers at Jamaica were central to the LIRR because of the system’s design, with branches on one end, multiple termini on the other, and one meeting point in the middle. The key for a timed transfer is that if one train is delayed a small amount of time, say a matter of minutes—a frequent occurrence on the notoriously imprecise LIRR—the other will hold at the station so people can still transfer. For the last 50 years, that’s what LIRR did. 

But ESA eliminated the timed transfers at Jamaica. Now, if, say, a Brooklyn shuttle is three minutes late getting into Jamaica, the connecting train going to people’s final destination might not wait. Moreover, because of the new platform for the Brooklyn shuttle, the transfers are no longer cross-platform. They are called “up and over” transfers where people have to go up stairs, across an overpass, and down to the scheduled track, which can take several minutes, adding to the possibility of a missed connection. And if the shuttle is crowded, there could be hundreds of people trying to get up and over, resulting in crowding and more delays and missed connections. If you miss your connection, you have to wait until the next scheduled train, which could be anywhere from 10 minutes to a half hour or more, depending on the time of day. The end result is that some people are getting better service to the east side of Manhattan while others have unreliable and potentially disastrous commutes and have to worry about whether they will make connections by a matter of seconds.

McCarthy said the MTA is aware of the timed transfer issue. In the morning rush, because the MTA runs a lot of service to both Penn and Grand Central, he said he thinks there is no tradeoff and is instead “about getting customers to understand the opportunities for transfer at Jamaica that are most often cross-platform.” By not holding trains, McCarthy said, they can run more service. But he added the agency is continuously tracking metrics, conducting customer surveys, and talking to people on platforms to address any outstanding issues and may consider bringing back timed transfers, especially off peak, in the future.

There have been other issues with the ESA rollout too, including overcrowding on trains, especially to and from Penn Station where a higher percentage of riders still want to go than the MTA anticipated, and patchy late night service out of Penn Station leaving patrons of MSG events waiting longer for trains home. But these types of service hiccups are common when creating an all-new service, especially one that gives riders roughly equal choices, and it’s hard to predict demand with much certainty. Initially, McCarthy said, the LIRR split train service roughly 55/45 between Grand Central and Penn, but the rider split was closer to 70/30, although over time it’s getting closer to 60/40. Faced with these numbers, McCarthy said, the MTA has to decide whether to stick or twist; do they wait longer and see how people get used to the new station or change service patterns quickly? The MTA has somewhat mitigated these issues by ironing out service kinks and making schedule adjustments, but O’Hara’s LIRR Today Twitter account is still littered with replies and retweets from riders documenting overcrowded trains and missed connections.

There is no quick fix for the Jamaica conundrum. As service stabilizes and the dust on a multi-decade project settles, we can see the benefits and drawbacks of the New York region’s costliest transit project in decades. Undoubtedly, some commuters from Long Island who work near Grand Central are pleased. But there is no avoiding the fact the MTA also spent $11 billion to make some people’s commutes worse. 

“That's the biggest concern to me,” O’Hara said, “that the press coverage will die down and people will begrudgingly accept their commutes are slightly worse and it’ll be this way for another 50 years.”