Hacker News new | past | comments | ask | show | jobs | submit login
Why New York City Stopped Building Subways (citylab.com)
272 points by jseliger 11 months ago | hide | past | web | favorite | 261 comments

NYC should stop building subways with tunnel boring. Instead, tear up the street, build the darn thing and then rebuild the street over it. This method is called cut and cover and it's what many countries do because it's substantially cheaper and faster:


Imagine if the 2nd ave line was done this way - sure you'd get a lot of pissed Upper East Siders, but their property value is going to skyrocket from the project anyway. As far as I'm concerned, they could deal with a year of inconvenience instead of getting it built over decades for magnitudes of money more.

Of course, the City Council would never agree to this plan. It's easier for them not to disturb anyone if they want to be re-elected...

You may be overestimating how easy cut and cover is. I live in the city of Antwerp where they decided to dig up the central crossing that connects the main shopping street with the main traffic hub, to redesign it so cars would go underneath and shoppers can walk across the crossing unimpeded. It was planned for 18 months, and now is projected to be closer to 3 years.

This is what it looks like when you do such a thing:


The stores in the neighborhood are complaining that they'll probably go out of business before the works are done.

> You may be overestimating how easy cut and cover is.

You're right. In fact, the Market Street passage of BART/Muni in San Francisco was a cut and cover. The closure of Market Street and delays in construction caused a lot of damage. Not the least of which the collapse of businesses[1]

The Theaters went out of business, became porn cinemas and liquor stores. Much of what Civic is today was a result of that era.

Not to say Cut and Cover was the 100% cause - but it was a major, major contributor.

1: https://hoodline.com/2016/07/in-their-words-the-convergent-h...

FWIW, the Crown Zellerbach Building, completed in 1959, was "controversial due to the decision for the building to face Bush St. instead of Market St., Market St. being in decline during the time it was built."[1][2] 1959 was years before the Market St subway construction began.

Aside: I'm not a student of architecture, and Modernist (and especially International Style) architecture didn't really click for me until someone explained the design of that building. All building services (restrooms, stairwells, elevators, etc) are located in the giant, black, windowless monolith. To my mind that feature is most descriptive of the style and its motivations, particularly regarding the relationship between form and function and the emphasis on the latter. Its shape and materials are, at least superficially, almost entirely dictated by its pure function. Yet it's the defining characteristic of the form of the building and clearly intentionally imposing.

[1] https://en.wikipedia.org/wiki/One_Bush_Plaza [2] https://noehill.com/sf/landmarks/sf183.asp

EDIT: You can't see the monolith from the Wikipedia or Wikimedia photos. Which is actually sort of strange given how imposing it is in real-life. Here's a better photo, https://www.som.com/FILE/16472/crownzellerbachhq_788x900_gab... , from the SOM website (see https://www.som.com/projects/crown_zellerbach_headquarters).

> FWIW, the Crown Zellerbach Building

I've always loved that building.

Very different neighborhood from Civic though -- even though they're only 20 minutes walk apart.

During and post-war that end of town was much less desirable. That changed significantly when the shipping moved. Also that building marks arrival of modern building techniques suitable for an earthquake zone. That triggered a construction boom in FiDi.

Mid-Market on the other hand was in a post-War retail boom. Civic was a major hub of retail and Theaters -- and was impacted a lot more by the BART construction.

All that said, the arc of San Francisco neighborhoods does baffle me at the best of times.

Vancouver too. The Canada line expansion killed the vast majority of cambie's family owned / small businesses.

Having recently witnessed the boom in construction around the Canada Line stations, I'm not sure I agree. Many businesses on Cambie are doing very well, including some that hung on through the difficult time when the line was being built. Yes, some went out of business and that is sad, but many survived and fortunes seem to be up as a result.

These are the vultures who got fat on the carcasses of those that didn’t survive.

In contrast, it didn't kill the downtown businesses on Granville St. While any long term construction is probably going to negatively impact adjacent businesses that doesn't mean cut and cover is going to be the wrong decision to make in every case.

> Not to say Cut and Cover was the 100% cause - but it was a major, major contributor.

I don't think it's the cut and cover method per se. Here in Minneapolis when they built a surface light rail line from Downtown Minneapolis to Downtown Saint Paul iirc nearly 200 businesses closed along the LRT and cited the construction as the primary reason. Any construction that impedes travel to businesses for a long time is going to be negative for businesses. Here we subsidized about half the affected businesses along the line during the 4 years of construction but it apparently was not enough.

Simultaneously, BART construction had a similar detrimental effect in downtown Oakland. Arguably, it is only now recovering.

I stopped going to theaters because everyone brings their loud kids. Are you sure there isn't a larger draw for porn cinemas and liquor stores?

This is now reminding me of this blog I found on HN yesterday:


Once you have enough sewer/natural gas/electric/cable/telephone/fiber/water/who knows what else/ lines buried under the streets I can totally understand how nobody would want to touch that with a ten foot pole, just like legacy code :)

> The stores in the neighborhood are complaining that they'll probably go out of business before the works are done.

Even with the tunnel boring they did on the Upper East Side of NYC for the 2nd Ave Subway, half the avenue was dedicated to construction entrances and other above-ground activities. Ultimately, may stores did go out of business. My point is, tunnel boring is also quite disruptive to what happens at the street level.

The tunnels were built with TBMs. The actual stations were cut and cover.

You can't really build stations any other way. New York is way too expensive to buy property alongside the road as an access shaft, and in any case you need to be able to access both tracks from any given entrance. So either you have an island platform directly under the middle of the street or two side platforms on either side, both of which would need a connecting passageway the width of the street, from street to platform.

I concur; just replacing some sewer pipes on Cedar St in Somerville is taking 2+ years of construction, and plans have been underway since 2013 and still aren't done. It's a residential street without too much odd about it.

https://www.somervillema.gov/cedarstreet https://www.somervillema.gov/sites/default/files/cedar-stree...

I'm not sure it's possible to dig a hole in Massachusetts without it taking 10x longer and costing 10x more than planned...

Or touch a bridge for that matter

If the cost savings on the construction are enough you could probably just pay off the local businesses.

In NYC that'd be tough. Those businesses go through a lot of money to survive. A single restaurant near me on the 2nd ave route, back in 2010, was paying $50k/month just in rent. That was just 1 diner. And on taxpayer's expense?

Rents go down too if there is construction.

By what, magic?

Most businesses sign long-term leases, often 10 years. Landlords are under no legal obligation to offer rent breaks while construction is underway.

In fact, landlords would probably prefer the businesses go under so they can sign more lucrative deals post-subway construction -- see the current NYC store vacancy problem: http://www.nydailynews.com/opinion/change-math-keeping-nyc-s...

You're assuming the area will become undesirable during construction and landlords will be forced to offer cheaper rates to lure tenants, but as we see above more likely they'd prefer to wait. Especially if a subway is about to be built there.


That's how it goes in other countries: if the construction work lasts longer than N days and affects a business, there is a compensation.

In Vancouver, BC, many business did go out of business when a skytrain line was built through a few communities. It can be devastating.

Where are you thinking of? Many skytrain stations are either built away from businesses or now have businesses depending on them.

Canada Line in 2010

> You may be overestimating how easy cut and cover is.

Not only that, it's not always feasible depending on the depth of the tunnel. The 2nd ave tunnel is 80 feet deep, and in a city with heavy underground infrastructure already in place.

You probably also underestimate the amount of infrastructure under the street - not just power, water, sewerage but also network, fibre, etc. It is not remotely a small job to do such a thing.

Not to say that it can't work out (or that some of those services don't have to be moved for tunnels) -- but cut and cover was a lot easier when we had less infrastructure.

These problems are obviously 10x in such a dense city as NYC.

I've read before that in many places, they don't even know exactly what's below the street. These sorts of surprises are the bane of budgets and schedules for massive public works projects.

The problem is that the alternative is not necessarily better. The Second Avenue Subway stub took about a decade to build, and even though New York opted to use less disruptive surface methods, it ended up taking much longer. 3 years is better than 10.

Isn't this temporary? The government should be concerned with the long term benefit to the city overall. The businesses that come in afterwards would likely benefit from subway access.

Sure it is temporary and the businesses established afterwards would benefit greatly. However you're discouraging the establishment of new businesses in general if your policy includes not caring about the short term effects.

What if that subway needs major maintenance in two years because of a mistake during construction? Why would I open a business on that street if the city has already shown it'll put me out of business to work on it?

Nope. The Bart trench appears to have turned Market St into a slum for generations (and this is despite all the municipal money being poured into transbay and other nearby sites).

Well, it IS getting better, slowly. Maybe it will take 75 years to recover. How long-term are you thinking?

> The government should be concerned with the long term benefit to the city overall.

The government should first be concerned by not crushing individuals' rights. That is what Rule of Law is for. Many societies have been utterly destroyed by putting "the general good" as a pretext for policies.

What “rights” do individual property owners have to public streets adjacent to their property?


That's just a standard good governance principle.

Doesn't the government subsidize these businesses while construction goes on? I guess I was under the impression they would be subsidized.

Give subsidies to the businesses that are affected...

@Joerie the 2nd ave/Q lines has literally taken more than a decade.

> in the city of Antwerp

So, Belgium.

> and now is projected to be closer to 3 years.

Right. Belgium. :P

National slights aren't ok on HN, so please don't post them here. (Unless I've misinterpreted this comment and it means something I've missed completely, in which case sorry.)

You're being downvoted for profanity.

From what I've seen in Seoul, "cut and cover" doesn't take a year, but more like five years. And Seoul isn't exactly known for slow construction problems, so I think it probably can't be made much faster.

It can still be cost effective, but it can be rather unpleasant during the construction period.

Part of the problem with that is you uncover all sorts of other stuff under the roads: water mains, sewer, gas, steam, electric, telephone, fiber, etc, and especially in a city like NYC it's not all documented. In some cases, old abandoned pipes are used to run new wires.

When you dig, you have to deal with all of this. If you break something, you have to fix it -- and in the case of fiber or multi-hundred-pair copper, you can't necessarily just "splice" it or easily reroute around your construction site.

Having watched new gas lines go into my neighborhood last summer, with non-trenching, small-scale directional drilling/boring, I can only imagine at the tangled mess some places may be or be becoming.

Here, the gas company (contractor of same, really) put in detection wires so that the location of the new, plastic pipes can be sensed. However, the contractor ran them in the most convenient fashion for them, meaning often on angles from a single starting point, and also rather close to the surface in places. Their are minimum depths they are supposed to maintain at various positions from the main to the meter. I'm not sure they always did so. Some homeowners who landscape or garden more aggressively, may be in for a surprise.

You have to do this with subway construction anyways, because New York ended up boring relatively close to the surface. If you're boring twenty or thirty feet (eight to ten meters) down you're still in the danger zone of misplaced utilities.

Going deeper is possible but does have drawbacks; access shafts must be deeper and it takes a longer time for people to move between platform and street level. New York's older cut-and-cover sections also do not deal with ventilation systems, because regularly installed grates to the surface essentially use the piston effect to circulate air in and out of the system.

Going deeper is possible but does have drawbacks; access shafts must be deeper and it takes a longer time for people to move between platform and street level.

Quite a few of London's lines are deep, but they generally have escalators which I don't see a lot of in Manhattan.

Cut and cover worked back in Sherlock homes days as there was very little buried plant at that time.

Cut and Cover is cheaper, for sure. But tunnel boring is also cheaper everywhere else than it is in NYC. The problem doesn't seem to be the construction method so much as the beaurocracy and politics surrounding the construction. Moving to a more disruptive construction method might help projects get completed faster, just because there will be the political capital to complete the project quickly to resolve the disruption. but i wouldn't expect it to solve the cost problems.

Indeed, a more disruptive construction method is likely to drive costs up, since it gives unions more extortion material to work with.

> Of course, the City Council would never agree to this plan. It's easier for them not to disturb anyone if they want to be re-elected

Oh no, democracy... On a more serious note, boring causes far less disruption to street-level commerce. I haven't seen an analysis which includes the lost tax revenue and economic activity from cut and cover.

Moreover, something deeper is going on as even boring in NYC costs more than in other countries.

> Moreover, something deeper is going on as even boring in NYC costs more than in other countries.

Corruption. Rampant, unabashed corruption. That's what's going on. https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...

You are certainly right with "deeper". I wouldn't dare digging anywhere near the foundations of a skyscraper, and cut&cover isn't doable in a city with dozens of decades of uncharted pipes, tunnels, cables and whatever you're gonna find once you start digging.

And I certainly don't want to know the insurance costs for this kind of project, no matter the method...

Deeper is possible, but expensive. See NYC Water Tunnel #3.[1] Under construction since 1970. Several sections already in use; completion 2020. As deep as 800 feet down, 24 feet in diameter.

[1] https://en.wikipedia.org/wiki/New_York_City_Water_Tunnel_No....

Parts of the DC Metro were built with cut and cover. We moved to the Maryland suburbs of DC during construction, and it was a nightmare. There were no accurate maps of DC. Maps would show a street as one way East, but Metro construction had changed it to one way West! It was so bad, my father, who loved maps, completely gave up on driving in DC, ever. I can understand the economic argument for cut and cover, but the up front cost in inconvenience is high.

> NYC should stop building subways with tunnel boring

Plenty of other cities worldwide use tunnel boring and get far better cost per mile than in the states. That's not the issue.

Which cities? What is the issue?

> What is the issue?

Overreaching unions. Layers of consultants. Squabbling jurisdictions. And as the article mentions, the fact that they build so infrequently that they essentially have to skill up from scratch every time and can never optimise or learn from previous mistakes.

There's a lot of information about this out there but that about covers it I think. Here's a nice quote from a nytimes article:

"Trade unions, which have closely aligned themselves with Gov. Andrew M. Cuomo and other politicians, have secured deals requiring underground construction work to be staffed by as many as four times more laborers than elsewhere in the world, documents show."


Paris is old, narrow, and managed to build a subway tunnel through/under the Catacombs for far less.

The major issue seems to be that construction unions do not negotiate with the public authorities paying for it, but the construction contractors, and the contractors are more than happy to just pass along the cost and pad their margins. https://www.nytimes.com/2017/12/28/nyregion/new-york-subway-...

It's not just that, it's an entire stack of unaccountable entrenched interests from top to bottom. Unions, contractors, MTA management, all the way up to the squabbling governor and mayor.

One of the most discussed is Madrid, but most every other city outside the U.S. has lower expenses, even those with comparable (or worse) labor and other costs. A simple Google search will provide plenty of articles and studies, such as http://lab.rpa.org/building-big-less/

The elizabeth line in London is about to start service. from http://www.crossrail.co.uk/news/crossrail-in-numbers it includes. "42km of new rail tunnels"

Crossrail (the project which is to be branded "Elizabeth Line" although all the other "lines" are lines within some larger modal network, so arguably this branding makes no sense) was a huge project. From funding to today it's ten years and the basic idea (East-West tunnels through central London) is from the mid-20th century.

It's taken so long that projected peak system congestion _with_ the extra line is predicted to be worse than it was without the line when it was funded, even though its capacity is about 10 passengers per second through the core. The lines it's "relieving" are hopelessly over-capacity, so it will fill immediately when it opens AND it's expected to increase total usage because hey, there's this line from outside London right into the core, why not live there and commute...

But certainly it's a comparable problem, in the sense that e.g. Crossrail had to "thread the needle" fitting the new tunnels between existing escalators and the Northern Line underground rail tunnel that both remained in use during construction.

> From funding to today it's ten years

This seems pretty reasonable to me, to be honest. And most of it is done ahead of schedule isn't it?

How quickly do you think it should take from zero to a 21 km tunnel underneath London? It seems insanely fast now I think about it.

Very fast. And not just that. A big part of the cost and complexity of Crossrail isn't the tunnelling but rather the (re)builds of the stations, the remodelling and reconstruction of the above-ground street areas to increase capacity there too, and many other subtleties that people don't tend to think about.

It's the biggest civil engineering project in Europe, or so they say. And a big part of that is because you don't just massively increase transport capacity into the core of a major metro area. That would just overload all the other infrastructure.

> A big part of the cost and complexity of Crossrail isn't the tunnelling but rather the (re)builds of the stations, the remodelling and reconstruction of the above-ground street areas to increase capacity there too, and many other subtleties that people don't tend to think about.

The BBC's documentaries on the Crossrail are great at showing all of the complexity of building the thing and dealing with historic buildings, super-hard brickwork, archaeological discoveries, concert halls and more.


Crossrail is a well-done project. But just to be clear: there were fully 68 years of feasibility and planning studies prior to getting the funding for construction. Since then, it's been laudably smooth sailing -- but if one considers the pre-construction planning work, there's substantial room for improvement.

What's the estimate economic cost of a cut & cover, which shuts down a major NYC street for 1yr, vs the projected cost of a tunnelled subway line x2 (because modern government is always 2x the budget, at least)?

Some shops on the street could very well go out of business due to lack of local traffic... would they be compensated?

Yeah I would assume that there is no way that they would be compensated for the lost revenue. I would also imagine calculating the lost revenue for those shops, and subsequently trying to get the government to reimburse them for it, would be prohibitively expensive both politically and financially.

Second Avenue Subway construction lasted a decade and the stations still had to be built as cut and cover (you do need access shafts to the stations).

In light of that cut and cover sounds downright peachy.

The alternative solution is to build stations "inline" with the bored tunnels. This can be done quite cheaply, assuming modern ventilation/engineering are used. Note that $2bb of the $4bb for the 2nd Ave line came from station construction...

> Note that $2mm of the $4mm for the 2nd Ave line


Good catch, thanks.

You still need access shafts. New York builds subway stops that have access from the sidewalk; purchasing property for that would be too expensive. And if you build entrances from the sidewalk you also need a way from a sidewalk to access tracks going in both directions. So you end up digging up the entire street anyways.

In addition to what others have said about the disruptiveness of cut-and-cover, it's worth noting that Manhattan's bedrock is almost entirely schist[1].

I don't know whether that was ever factored into the decision to bore instead of cut, but I'd bet that cutting through schist is a lot harder than the chalk and clay of London and Paris.

[1]: https://blog.epa.gov/blog/tag/manhattan-schist/

The original subway tunnels were built with cut and cover, so that isn't an issue.

I did some research, and it looks like the IRT was constructed with a mixture of cut-and-cover and boring[1]. The schist apparently isn't too much of an issue when you have dynamite[2] ;)

[1]: https://mhdh.library.columbia.edu/exhibits/show/irt/laying-t...

[2]: https://www.nypl.org/blog/2015/05/04/subway-construction-the...

Better idea: ditch subways and bring back elevated lines. Works just fine in Chicago.

Monorail! Monorail! Monorail!

Yeah! Specifically I'd like to see more Schwebebahn installations.


When I was in London many parts of the underground was elevated, so I guess it works for them as well.

As far as I can remember none of the London Underground is elevated.

Because London's transport is all integrated under control of the local government (an option which is prohibited in most UK cities to facilitate "competition" ie price-gouging by foreign owned corporations) you might not have realised that you weren't using the Underground but one of the other modes, almost all of which are covered by the same ticketing system (the "Dangleway" isn't and neither are the boats).

If you travelled on ordinary seeming trains above ground, those weren't the Underground but instead ordinary "heavy rail" trains on National Rail infrastructure within London. Many of those lines are indeed far above the ground in parts of central London, but those are just called viaducts and aren't usually considered an "elevated railway" per se.

If you travelled on small trains in East London that didn't have a driver (you can sit at the front like a roller coaster), that's the Docklands Light Railway, almost all of which is elevated, it was built towards the end of last century during regeneration of the "Docklands" area.

If the trains seemed to be sharing a road with cars at street level, in South London, that's the Tramlink.

If the trains were smaller and red, that's a bus.

There are parts of the Underground that are, in fact, above ground and elevated (especially farther out like Ealing and Stratford). Then there’s also the Overground, which I think counts as light rail, but is generally on ground level.

The Overground is just a (branded) part of the normal National Rail heavy rail network that serves London. Same track, same stations, same rolling stock, just operated by TFL (Transport for London) like the rest of London's public transport.

TFL control two heavy rail networks - the Overground and TfL Rail - although the latter will become part of the Elizabeth Line when it opens.


Pretty much all the elevated train lines in London were built a long time ago - often before any of the houses which surround them. I don't think any new elevated lines (with the exception of bridges and the occasional longer section to join existing elevated lines) have been built for decades, perhaps significantly more.

You're probably thinking of the Dockland's Light Railway.


It's not part of the Underground, but it's also managed by Transport for London as part of its integrated network of public transport.


We have some parts of the subway in Paris which are elevated, but the street under the rail can't be used for trafic, it's usually parking space or empty. So now the city is building tram or bus lanes above ground.

There are elevated trains in the New York subway system as well. For example...


This was featured in the Blues Brothers right?

Are you kidding?

The Blues Brothers was set and filmed in Chicago (although they did shoot a scene in Milwaukee). It showed the "L" rapid-transit system.

> "Are you kidding?"

I don't think that was necessary. Yes, The Blues Brothers was centred around Chicago, but it might have been a while since GP saw the film, so their memory may have been hazy.

Aside from that, The French Connection featured a famous car chase that included the elevated trains in New York City:


Yeah, that's fair. My surprise came because The Blues Brothers is, in my mind, the Chicago movie. It couldn't really be more Chicago if it tried.

As long as they are quieter than the elevated BART lines.

They're only too loud if you're under them. Maybe if you live right next to it with a real old window facing it as well.

I'd be interested to know where you find this to be true. I have noticed that BART trains through Rockridge don't make much noise. But the track noise is so bad through Fruitvale, you can hear the trains over a mile away.

Looks like it wasn't that clear, but I was referring the elevated lines in Chicago (El) not BART.

I could hear the elevated Bart line at Daly City station two miles away in Colma, where all the graveyards are.

In addition to the problems related to working around utilities, subways built with cut and cover create a lot of street level noise when operating. Also, buildings vibrate every time a train runs by. Bored subways are imperceptible from above ground.

> Bored subways are imperceptible from above ground.

Not entirely true - the Crossrail in London goes under the Barbican arts centre, and that section of the track has special floating tracks to reduce noise and vibrations:


One of the biggest problems with that approach in NYC is that they simply don't know what's under the streets.

There's multiple articles about the issue, but this article from the NY Times does a good enough job illustrating the problem: https://www.nytimes.com/interactive/2016/08/18/nyregion/new-...

Cut and cover also doesn't work as well when you have hundreds of mapped and unmapped utility lines and pipes only a few feet below the surface.

As someone who was in the UES a lot, right in the area where 2nd ave construction was happening, I can't imagine cut and cover being much worse than what they did there. Constant explosions, tons of dust, a lot of restaurants had to close because no one would go there.

Cut and cover only works well if there is accessible territory underground. Manhattan is a Swiss cheese of tunnels, pipes and wires that were dug using cut and cover a century ago. If you want a tunnel now, you have to put it further down underneath everything else.

> cheaper and faster:

Faster is reasonably objective. But one reason it might be "cheaper" is that the extra cost is paid as an externality by all the people who just have to lump being inconvenienced.

Toronto is doing cut and cover on Eglinton street over the past few years. It seems to be going quite well, I'm sure it affected some businesses, but most of Eglinton is not quite gentrified yet.

No, Eglinton line is bored tunnels, only the stations are cut and cover. http://www.thecrosstown.ca/stay-informed/fact-sheets-backgro...

Ah right, my misunderstanding

2nd avenue already got pretty torn up with access tunnels, it's not like tunnel boring helped a ton.

Also a good chunk of NYC subways were built with cut & cover (sections of IRT 2/3 lines as an example).

in Amsterdam they used a drill. Very soft soil, no elevation soever. But underneath a lot of buildings and not always following the streets so Cut and Cover would not have been an option. Most interesting thing: freezing the area (with often wooden foundation poles in it) before drilling through it for stabilization.


And the Noord-Zuid Lijn almost bankrupted the city after all the historic houses at the end of Vijzelstraat started collapsing and people had to be evacuated in minutes for multi year hotel stays! The cut-and-cover stuff they did when they built the original metro was much more destructive to places like Waterlooplein, but also much lower risk...

What's your point? This project is a massive failure.

For a good short documentary: https://www.youtube.com/watch?v=tMhjxp_CZng

This is how the original subway tunnels were built.

sure you'd get a lot of pissed Upper East Siders,

The owners of the forever destroyed street-level businesses, you mean?

i think sf is doing cut and cover on stockton for the new muni line and it's been in construction since i moved here 3 years ago

those are just the stations, the tunnel is bore

Beyond the constant delays and overcrowded trains, the biggest impediment I experience is the lack of subways to the airports. LGA doesn't have any train lines, while JFK and EWR have "airtrains" separate from the MTA or PATH.

Chicago's "L" goes directly to both ORD and MDW. My brief time there showed that things could be far more connected than New York's system.

American cities are the worst when it comes to this.

Seattle just built a train to the airport (yay) but for bureaucratic reasons, it doesn't actually go in to the airport. It goes to the far edge of the airport, so you have to lug your luggage outside through the entire parking lot.

Meanwhile, everyone who took an Uber can step right into the departures area.

And, as you mention, New York is much much worse in this regard. The AirTrain from JFK costs $5, more than it costs to go from anywhere to anywhere else in all of NYC on the subway.

> The AirTrain from JFK costs $5, more than it costs to go from anywhere to anywhere else in all of NYC on the subway

I'm okay with this. If I'm going to the airport, I'm identifying myself as a New Yorker (or visitor) with the means to travel by air. It doesn't make sense to use a minimum wage worker in the Bronx's taxes to subsidize my vacations.

Broadly, however, I agree that our lack of downtown-to-airport express rail lines is embarrassing.

What if you work at the airport?

Their employers should pay them more or buy them tickets. They've been pushing, instead, for a subsidy [1]. But at the end of the day, that's still using lower-income New Yorkers' wages to subsidize a system which principally (though not exclusively) benefits higher-income New Yorkers.

[1] https://www.nytimes.com/2005/01/12/nyregion/train-to-jfk-sco...

You still have to pay. There are a ton of airport workers at the Q10 terminal at Kew Gardens.

How are they "subsidizing" your vacations? That's one less dollar you're spending at the restaurants they work at or at the other businesses that employ them. So they're subsidizing the airport because the more you pay at the airport, the less you spend in the city. I'd rather spend that money in the local businesses than give the Port Authority another dollar they can squander.

That's one less dollar you're spending at the restaurants they work at or at the other businesses that employ them.

In reality people don't visit a city and spend a precise, calculated number of dollars.

$1 towards government services is still a $1 of non-subsidy not being collected somewhere else.

Realistically speaking, who is making leisure or business trips where a few dollars is a major concern?

Airport transit investments are, sadly, incredibly hard to justify. Since airports tend to be located in far-flung, less-populated parts of cities (or entirely outside of cities), it faces the deadly combination of being expensive to build (greater length/mileage of line) and having low ridership (relative to the rest of the transit system).

It's a highly salient point of failure for visitors, and generally is bad PR. But besides looking bad, it's not clear that investing in single-seat-to-terminal trains to the airports are actually the best use of very limited dollars.

The JFK AirTrain for example has a ridership of ~11,000 per day. There's an argument to be made that those riders have greater economic impact than average, but ultimately that ridership resembles some of the more far-flung residential neighborhood subway stations (see: Graham St L, Bergen St F, 145th St 1 in Harlem, etc).

Compare that with parts of the system that are in dire need of maintenance/renovation with much greater ridership (see: Herald Square - 125,000 riders a day, Union Square - 106,000 riders a day, Times Sq - 202,000 riders a day).

In a world where costs are under control and funding abundant, I am all in on airport links, but right now? Right now I'm not convinced this is where money should be going, as much as it sucks for someone who flies a lot like me.

> Airport transit investments are, sadly, incredibly hard to justify.

Somehow other countries manage to justify it. Tokyo, London, Paris (TGV too!), Amsterdam, Frankfurt, Beijing... the list goes on. But somehow building infrastructure is hard to justify in the US. Pennywise and pound foolish.

Very few of those places built the airport extension first, or spent the entire bank account on just the airport extension. Given limited dollars, airports are at best a middling concern.

Limited dollars you say? How are the limited for the most powerful country in the world with the strongest economy?

A combination of low appetite for taxes, poor spending, and the disadvantages of being the first movers when it comes to infrastructure.

China is able to rapidly build infrastructure because it didn’t have very much. In twenty to fifty years when everything is dur for replacement, that will be the true test of infrastructure.

The AirTrain has so few riders because it fucking sucks, this phenomenon doesn't require detailed analysis.

NYC can and should build trains that go from dense center city locations right into airport terminals. Like real countries do. If they did they would be wildly popular.

The problem is that there is little remaining rail capacity into New York as it is and it would almost certainly be more productively used for commuters rather than occasional travelers.

Cause or effect though. Are there fewer riders because of the lack of decent rail-service to the airport, or the other way around?

It's a good question - and we can gain some insight by looking at cities with much better rail service to their airports.

The baseline is the JFK AirTrain, which carries 11,000 riders per day, out of an estimated total of 5.6M daily subway riders (not counting commuter rail), against a total flyer volume of ~60M a year.

The Heathrow Express, which does take you straight to the terminal, for example carries about 17,000 riders per day, out of an estimated 7M total daily Tube riders (not counting commuter rail), against a total flyer volume of ~78M a year.

Another data point: the BART SFO station (also direct-to-terminal) carries about 6,500 riders per day, out of an estimated total 423,000 daily riders system-wide, against a total flyer volume of ~50M a year.

It's not all doom and gloom - we also know that Narita Airport was able to double its proportion of passengers arriving/departing by rail by improving service. But even if the same results can be replicated in the US - the baseline numbers are so low that the increased ridership would still represent one of the smallest parts of the overall transit system.

FWIW I'm one of the beneficiaries of these rail links, and I'm very grateful they exist - but these projects are ultimately more about signaling/politics/PR than they are about shrewd allocation of infrastructure spending.

About the Heathrow Express. Its ridership is low because the Piccadilly Line (1 of London's busiest lines) also terminates at Heathrow Airport. Most people I know take the Piccadilly Line instead of the Heathrow Express to the airport to save money.

Yeah, Heathrow Express makes sense if you're ending up near Paddington but otherwise I don't really understand it. If there are a couple of you with luggage, get a car and it's door-to-door. Otherwise take the tube and save a bunch of money. I go to London a lot and mostly take the Piccadilly Line or sometimes a cab at off-hours or from inconvenient locations. But I honestly don't understand why the Heathrow Express would be interesting outside of a narrow range of circumstances.

BART SFO is technically direct to terminal but only because you change to a specific spur line. This is similar to the JFK air train and seems designed to support the taxi union. Each such change cuts the number of poss Le riders dramatically, especially when lugging bags.

I take BART to and from SFO all the time. But then I'm essentially always just with carry-on. Pretty much any public transit from airports (with the possible exception of busses explicitly set up for the purpose) aren't a great fit for travelers lugging a bunch of suitcases.

The arrangement at Rome Ciampino Airport is pretty silly: there's a railway line running next to it, with a station about 300 metres from the runway. But the entrance to the airport is on the other side, on the motorway!

To get from the station to the airport, you need to take a bus or a taxi round half the perimeter of the airport. Which, if you don't have long before your flight, and the bus drives off before you board, leaving several passengers fighting over the handful of taxis in this tiny town, can be quite stressful, i can assure you.

What's worse is that airport _workers_ have to pay additional for the AirTrain I think there's a discounted rate now, but I'm pretty sure workers had to pay $5 at one time. I did some work with the PA in the mid 00s and someone I worked with told me that he still had to pay $5 when the AirTrain first opened.

At least the whole walk is covered. Count your blessing.

That kind of makes sense though. Most public transit is subsidized. Why subsidize people who aren't necessarily paying taxes in the area?

This was a mindblowing discovery for me when I flew into NYC for the first time. One of the largest subway systems in the world and some of the busiest airports in the world a few miles from each other but not connected was/is confusing.

Is there a well understood reason why?

Depends, are we talking about JFK, EWR, or LGA?

I've looked into this some - but if someone is better versed in the history of this stuff, please jump in.

For LGA, there have been attempts for years to extend the N train north from its current terminus in Astoria, to LGA. This is the most economical way to connect the airport to the mass transit system (there is currently no train whatsoever, the only way to LGA via public transit involves a bus).

Most proposals for this line extension called for the N train to become elevated on its approach to the airport, and the neighborhood has loudly opposed any of the nuisances that come with living next to an elevated train. This political pressure has been impossible to overcome, while tunneling has not been considered feasible or affordable.

On top of this, the state government which runs the NYC subway is reluctant to endorse this plan also, since connecting LGA to the NYC subway doesn't serve the (very important) suburban electorate. The state government has consistently pushed instead for the airport to be connected to a nearby suburban commuter rail station instead (Mets-Willets Pt on the LIRR). The political calculations behind this are pretty obvious, despite this option being far less connected to the greater transit network and promising far less ridership, and also perpetuates the need for a "monorail" transfer, like JFK and EWR.

The history of the transit situation for JFK and EWR are also interesting (and like most things NYC-infrastructure related, extensively documented), but this post is running pretty long. Happy to go into it if people are interested, but the tl;dr is: funding instability, financial downturns, squabbling fiefdoms, jurisdictional battles between NJ/NY, and the ever-present NIMBYism.

That's a great comment and I would love to know your thoughts on JFK/EWR situation if you don't mind writing it up.

My father (retired railroad management in New York and then Los Angeles) used to blame taxi and limousine commissions and lobbies. He was resigned to the fact that rail will never get built to most American airports, at least not in the way it's built in Europe.

> Is there a well understood reason why?

LaGuardia, by all rights, shouldn't exist. It needs to be shut down - aside from the fact that the airport itself is already falling part, its mere existence actually creates delays across all three airport systems in the area (JFK, LGA, and EWR). Because of the way the traffic zones intersect, it would actually be more efficient to reroute all existing traffic from LGA to EWR and JFK, and we'd end up with fewer delays with the same overall capacity.

Unfortunately, LGA is the most convenient airport for people who live on the Upper East Side. They're wealthy and politically well-connected, so they would strongly oppose any efforts to shut it down. They also are only a $20 cab ride away from LGA, so the fact that it doesn't have heavy rail/light rail transportation there doesn't bother them.

This post is literally ridiculous, I don't think a single thing in it is true.

The constraining factor for flights in and out of NYC, or nearly any place, is runway capacity, not available airspace. For all practical purposes airspace is essentially unlimited. You could bulldoze Brooklyn, strategically place a couple hundred runways on it, and land thousands of planes per hour if you wanted to, maybe a couple less if it got foggy.

For reference I have a pilots license and have flown across and through NYC's class B airspace many times. This post makes no sense.

> This post is literally ridiculous, I don't think a single thing in it is true. The constraining factor for flights in and out of NYC, or nearly any place, is runway capacity, not available airspace.

What you're saying flies in the face of a decent amount of reporting on the matter, as well as what I've been told by a friend of mine who happens to work in ATC at one of the aforementioned airports.

I honestly try to take the L whenever I can, and do take it to the airport about 30% of the time. The trouble with the L approach is:

1. Normal rapid transit cars have no space for luggage

2. There are no express trains, so it’s not guaranteed to be faster than driving (it often is during rush hour however)

3. The airport stations are a 10 minute walk to security instead of a 1 minute walk like the car dropoff gets

4. When traveling for business I can charge in an Uber, spread out in the back seat and get some work done. On the L I’m in a cramped seat where using a laptop is hard.

So on paper I appreciate the approach New York and London (and others) have taken with a premium airport-only service.

I think the lack of elevators is worse. In fact, elevator availability is sort of a precursor to good airport access since schlepping luggage up and down slick stairs can be a nonstarter, especially if you are disabled, elderly, or have small children.

Transit to airports is generally a secondary concern. The only regular group of riders that would use such a line would be airport workers, because by definition there are not that many travelers, and on top of that many of them with luggage or in large groups could find it easier or cheaper to pay for the door to door ride of a taxi.

The O/D pairs of people going to and from airports also tends to be very dispersed, which is not great for rail unless you have an extremely comprehensive, rapid rail network covering nearly all possible trips.

Airports also, by their nature, tend to be out of the way and away from busy neighborhoods, which limits their potential catchment. Even when they are linearly arranged, terminals are generally set back from these linear routes due to their circular road networks and parking requirements. Look at how much additional mileage is required to serve the terminal at Dulles in DC, for example: https://i1.wp.com/www.thetransportpolitic.com/wp-content/upl...

Finally, airport layouts themselves tend to be really bad for transit; common configurations for airports, laid out in a circle (JFK) or a very long terminal (LGA) will require some significant chunk of passengers to walk to and from any rail station sufficiently long enough for a regular train, to the point where you need a people mover around the airport anyways (SFO, ORD). And if you need a smaller people mover to shuttle people around anyways, why not just shuttle them to a central train station a bit of a distance away, so that the rail can effectively serve both further destinations and the airport passengers?

I lived in New York until the early '80s (I was 11 at the time), and I recall the commercial "Take the Train to the Plane". I always thought that they were part of the MTA.


It was part of the MTA. You'd still need to take a bus from the train, though.

From memory:

JFK: just have to transfer from AirTrain to regular subway in Jamaica (?) Queens.

Newark: take the bus to Newark station and take PATH train to lower Manhattan.

Plus you get to see real New Jersians and New Yorkers. They are not overly friendly and they dress way too fancy compared to the Californians.

> Newark: take the bus to Newark station and take PATH train to lower Manhattan

From Manhattan, it's an NJT to the Newark airport stop from where the cutest little monorail [1] takes you to the airport.

[1] https://en.wikipedia.org/wiki/AirTrain_Newark

Also one of the slowest monorails I’ve been on, even the Tokyo monorail built in 1964 was faster :)

JFK is actually even easier, in that you have LIRR trains departing from Jamaica to Penn Station (and in the future, Grand Central) every few minutes at nearly all hours of the day, and the transfer is a single elevator ride. You can be in Manhattan within 35 minutes of leaving the terminal.

Transit to the airport is less helpful than you'd think. The main people who are going to want to use transit are the people who work there (being both lower income and not having to carry lots of luggage around), and the working hours of subways is generally inconvenient for airport workers.

cough respectfully disagree. In most major Asian cities the airport connection is super useful, going in and out of NYC felt like going to the 50s and it was better than most of the other cities when visiting the US.

The NEX in Tokyo is great, since Narita is way outside of Tokyo. As is the connection to Incheon directly from the center of Seoul.

Turns what would have been a painful ass commute into something relaxing.

In Seoul I could even check in the luggage for my flight before hopping on the train.

I guess if you hate convenience....

Seconding this. I did not need a car at all in Tokyo, not even to get to the airport. The NEX and services like it are fantastic.

Yet in the rest of the world many people use the train to get to the airport. When I fly into Moscow DME, why would I want to sit in a traffic jam for 90 minutes when I could get the train in under half that time. Taxi from Amsterdam Schipol? That would be crazy. From Tokyo Narita? You're having a laugh. London Heathrow? Get the express to Paddington and a taxi from there. Or just take the tube the whole way.

> From Tokyo Narita? You're having a laugh.

When I told a cab driver in Tokyo I wanted to go to Narita he was extremely confused and had the hotel bell hop explain to me that, basically, I was nuts and should take the train.

You're lucky he didn't stick you on one of those awful hotel-affiliated coach buses to Narita, the drivers of which are forbidden from applying steady pressure to the accelerator.

Nothing wrong with taking the bus in certain cases. I used one last time from the Ginza area. It was either being compacted in peak hours Tokyo Metro, or walking to a nearby hotel and waiting for a bus in the lobby. The choice was clear and I don't regret it.

> London Heathrow? Get the express to Paddington and a taxi from there. Or just take the tube the whole way.

That's about as convenient as taking the LIRR to/from JFK, or the NJ Transit to EWR, which many people in NYC already do.

JFK requires an airport train to take you to the train station. Same at Newark, and BWI from memory.

> JFK requires an airport train to take you to the train station

It's the same train that takes you from one terminal to the next. The alternative would be to force the LIRR or the A train to stop at literally every terminal on its way to Jamaica/Far Rockaway. That would be way worse.

It's not like transferring between them is any more inconvenient than having to transfer between two subway lines, so it seems like a distinction that not particularly meaningful.

Cross rail is a commin' http://www.crossrail.co.uk/route/maps/

> The main people who are going to want to use transit are the people who work there

I live in Manhattan and regularly take the LIRR to Jamaica, from where I catch the Airtrain to JFK. It's a predictable-duration trip. It costs $10 to Juno to Penn Station, $7.50 (off peak) to $10.25 (peak) for the train ticket, and $5 for the Airtrain. And I don't get sick reading on a train like I do in a car.

Transit to the airport is hugely helpful for travelers. A subway ride from LGA to a Manhattan hotel would be far cheaper and more efficient than taking a cab.

If cost is an issue, there's the NYC Airporter bus, which is a flat $15 to Midtown. Or you can take the Q70 bus to the Roosevelt Ave - Jackson Heights subway station (in Queens).

Airports are a particularly visible omission, for a certain class of people, but they aren't in the top 5 or so most valuable possible extensions to the subway system.

> Airports are a particularly visible omission, for a certain class of people, but they aren't in the top 5 or so most valuable possible extensions to the subway system.

Speaking of class, some transit systems suffer more than others from lacking a healthy mix of ridership affluence (MTA isn't one of them). Descending into a spiral of "only for the poor" is one of the failure modes of transit systems. If an airport line "injecting" its share of non-destitute riders is what keeps the while system from falling off that cliff it would be worth far more than just their tickets or the directly prevented road traffic.

High status destinations have indirect benefits for the whole system.

Airports aren't in the top five most beneficial routes for middle- and upper-class residents, either. There are all kinds of outer borough routes that would be of far greater benefit to the residents there.

This is sort of like the case of the couple who buys more house than they need because the in-laws visit once a year and they insist on having a room for them. It's silly to optimize for rare events. When you have to go to the airport, then it's particularly noticeable that there isn't a route to the airport. But going to the airport isn't on your everyday agenda. There are far more beneficial routes.

This article touches on but doesn't explore in great detail the history of elevated lines in NYC. Some of their traces have been incorporated into the current system, most notably as the 7 train in Queens and the J/M/Z (including the Ridgewood/Middle Village spur), but many were torn down.

Some lines in Brooklyn that people aren't even aware of (beyond the 2nd and 3rd Ave lines):

* The Myrtle Ave elevated, running from Downtown Brooklyn down Myrtle and connecting with the M's Ridgewood spur. Torn down in the 1960s.

* The Lexington Ave elevated, running from the Brooklyn Bridge waterfront across Downtown Brooklyn and Bed Stuy, connecting to the current J/Z line near Broadway Junction. Torn down in the 1950s.

* The Fulton St elevated, since replaced by the A train, that ran over Fulton St all the way into Downtown Brooklyn. Torn down in the 1940s.

You can see these elevated lines crossing Northern Brooklyn in this map from 1949: https://www.nycsubway.org/perl/caption.pl?/img/maps/system_1...

I personally would have found these lines quite useful, but what took their place was a collection of bus routes, running along old streetcar routes (read: weird routes that stop suddenly at borough and neighborhood boundaries). And NYC buses cannot be trusted...

> You can see these elevated lines crossing Northern Brooklyn in this map from 1949: https://www.nycsubway.org/perl/caption.pl?/img/maps/system_1...

Using a Perl script to wrap a (non-animated) GIF of an old map. So much history in that one link.


This one would have been convenient for me if they'd kept it.

> You can see these elevated lines crossing Northern Brooklyn in this map from 1949: https://www.nycsubway.org/perl/caption.pl?/img/maps/system_1...

One of the commuter tips is "Try to shop between 10 and 4". Were shoppers overwhelming the peak-hour commuters?

EDIT: other interesting things on the map: the 5-digit telephone number.

> other interesting things on the map: the 5-digit telephone number.

It's not 5 digits, it's 7: MAin 5-6200. The first two digits are 6 (M) and 2 (A): 625-6200. Telephone exchanges used to be named, this particular one was called MAin.

See also https://en.wikipedia.org/wiki/Telephone_exchange_names

> This article touches on but doesn't explore in great detail the history of elevated lines in NYC. Some of their traces have been incorporated into the current system, most notably as the 7 train in Queens and the J/M/Z (including the Ridgewood/Middle Village spur), but many were torn down.

You don't even have to leave Manhattan to see elevated lines, just take an uptown 1.

As a native Detroiter one of the saddest things to me was that the city never got a subway. I don't believe you can have a truly world class city without a subway.

There were multiple efforts to create a Detroit subway starting in 1910 but each one fell short. Finally in late 1928 the subway was approved! Two stations were built but construction stopped with the great depression. Today sadly no one even knows the location of the two stations.

Here's the winner of a dream subway contest for Detroit:


I don't know the detailed history at all, but in an intuitive & oversimplified way, I wouldn't think big transit projects ever had much of a chance in the home of the auto industry. Seems like they would want to demonstrate what an automobile utopia looks like.

Well the auto industry did fight the subway which is why despite overwhelming support it took so long.

Detroit did have an extensive trolley system. I found out that when the trolley's were secretly purchased and discontinued by GM in 1956 that 25% of Detroit families didn't have a car. The majority switched over to the bus lines which were expanded. Course the buses were made by GM ;<).


My god, did that happen to every trolley line? Makes me want them back all the more.

Well my dad was called a conspiracy theorist and worse for advocating that GM bought up Detroit's trolley line. I am just glad that he lived long enough to see the truth come out and be proven right.

It happened in a lot of cities nationwide, from what I understand. Didn't know about Detroit, but am not surprised. I think possibly the fact has even been established that it was the result of a, well I'm tempted to use the word conspiracy, but let's say "effort," of Standard Oil, Firestone Tires and Ford, but don't quote me on that.

A lot of places felt the need to take the suspiciously over-enthusiastic step of burning the trolley cars... ostensibly to strip the wood & recover the scrap metal.

It's funny how so many people don't know this. It is true of almost all major midwest to east coast cities.

That makes sense to me? Buses are environmentally equivalent to trolleys, cheaper, both have to deal with traffic, etc. In most cases the fact that buses are more flexible is good but inflexibility can sometimes be useful in terms for getting people to believe you're really going to stick with a new route and therefore it makes sense to start a business on it.

Now if they were replacing subways or light rail with buses that would be horrible.

Having a dedicated line to travel on helps and preference at traffic lights helps a bunch. I think you underestimate the amount of delays buses have to deal with when they also have to be in traffic.

Both trolleys and buses can have dedicated lines of travel or you can force them to share their lanes with other traffic. And both both buses and trolleys can be given light priority or they can wait with everybody else. The issues you're raising are totally orthogonal to bus versus trolley. Cities with good bus or trolley systems do these and cities with worse but cheaper systems don't. And another good way to speed up buses or trolleys is not forcing them to pull out of traffic when they stop to pick up customers by having special islands or having the curb extend into what would otherwise be parking spaces.

It probably does depend on the infrastructure choices the city implements. I don't see a lot of buses systems that have isolated islands, dedicated lanes, etc. I do see them frequently for street trolleys.

I'm curious as to why neither the Cadillac line nor the Michigan/Woodward lines extend up into the Pointes, yet St. Claire Shores has a major hub. Grosse Pointe Park in particular is quite walkable, especially in the older areas that were build primarily for house servants.

One of the things the NYC region could do is unify operations and planning of NJ transit, LIRR, and Metro North into one through running network with a tunnel from GCT south with two more Manhattan stops then across the Hudson to another main station where the Hoboken terminal is (a bit more inland likely). Overtime the systems and fleets can merge.

Instead you get this absurd patchwork of expensive stand alone projects, ARC, the huge building over of of a cap on the LIRR Hudson railyards which required maybe a billion in spending. Talks about a new Penn Station. Etc.

As a plus you reduce the demand on the existing Penn tunnels used by Amtrak.

If this had all be planned as one system the money already spent or being proposed to be spent on these disparate/disconnected things could provide do wonders.

>unify operations and planning of NJ transit, LIRR, and Metro North

Everyone agrees but insiders who know about the current state of the transit organizations consider this suggestion so absurd as to write an April Fools post[1] about it.

1: https://pedestrianobservations.com/2018/04/01/new-york-area-...

> insiders who know about the current state of the transit organizations consider this suggestion so absurd as to write an April Fools post

The difficulty comes down to regional politics. The regional rail systems (LIRR, Metro-North and NJTR) were set up to give those on Long Island, upstate/in Connecticut and in New Jersey, respectively, a dedicated regional transit authority. A politician attempting to unify operations would need to convince those voters should risk giving up their regional transit control.

Oh that’s golden, thanks for sharing. :-)

> As a plus you reduce the demand on the existing Penn tunnels used by Amtrak.

Amtrak has owned and operated the HRT (tunnels built in 1905) since 1976 when it took over the NEC from the failing Penn lines in the early 70's. It's the busiest segment of track (Interlock-A) in the world and out of the busiest train station in the world (NYP).

You could reduce demand on the existing tunnels by building ARC.

An excellent point. This is also congruent: https://www.vox.com/policy-and-politics/2017/1/1/14112776/ne....

Oddly, the GAO effort to figure out why infrastructure costs so much isn't in the budget, last time I checked: https://www.vox.com/policy-and-politics/2017/5/24/15681560/g...

Seriously. I visited NYC last year and spent a few days hanging out with some friends out on Long Island. Traveling on LIRR is a very unfriendly process for anyone who doesn't do it basically every weekday and integrating LIRR travel with subway travel can be a non-trivial optimization problem. It's not super complicated once you get the hang of it, but otherwise it is just not obvious where you can get on and off trains, where the trains go and when, etc.

> GCT south then across the Hudson to... Hoboken

Effectively the PATH & 4/5 train already do this with a transfer at Fulton Street. And while unifying the system may look good on paper, I'd hate for the PATH to descend to the MTA's low standards of reliability and cleanliness. Better part of the system work well than all of it work poorly.

I'm always amazed at how clean the Path train and stations are especially when transferring onto them from the subway. It's a completely different experience

> One of the things the NYC region could do is unify operations and planning of NJ transit, LIRR, and Metro North into one through running network

How would this address the principal problem identified in the article, the rising cost per mile of building new track?

That unfortunately is an American centric problem, it’s particularly bad in NY but other parts of the country aren’t immune to it either. There’s a lot written about that elsewhere.

My point in this instance is that even at the inflated construction costs you could still accomplish a lot more with better planning and linking these systems up. A tunnel and new stops in manhatten connecting them is going to accomplish a lot more for the same price as capping a huge rail yard. Building new tunnels that terminate three separate systems in different locations. And building a new Penn station.

Does any cross state combined public transit effort even exist in the US? Maybe DC, I am not familiar with Washington DC.

It’s a nightmare in my city just between neighboring counties. Only 2 counties were included in the initial build out of public trains and not any significant expansion since. The other counties vehemently reject any such proposals. One county is going through expansion talks again but the train will only go one stop to their bus hub. I bet they will end up rejecting all of it.

NIMBYism isn’t very strong in my city but it certainly is regarding public transport, trains especially.

There's not a lot of places in the country with "true" interstate integrated transit needs. Places like Chicago have far suburbs and exurbs within the commute shed in other states but are still generally within state. Even cities like St. Louis generally have a clear dominant locus in one state, which limits the ability of the other state(s) to play off the other state.

The general exceptions to this rule are Kansas City and Washington, DC, where the workplace loci are truly distributed among multiple jurisdictions so that no one state is dominant. In DC, the standard subway (WMATA) runs across all three jurisdictions, although there is some amount of political football over funding. The commuter rail systems are not integrated, but that is largely due to limited capacity over the Long Bridge. (MARC, the Maryland system, is participating in the Long Bridge replacement project in part because they do intend to extend service to Alexandria once there's capacity to do so).

NYC is a truly special case because it's simply so massive. In terms of geographical boundaries, there are roughly three natural independent sheds: trans-Hudson, trans-East, and trans-Harlem rivers. Each of these sheds is massive enough to let each of the three commuter systems that focus on their own sheds to not have to coordinate. These sheds also happen to largely coincide with different states: NJ is trans-Hudson, NY (or, rather, Long Island) is trans-East, and CT (as well as parts of NY) is trans-Harlem.

Even in DC, there is a lot of public transit not managed by WMATA. While WMATA runs the MetroBus system, each jurisdiction also has its own bus system beyond that, in part because none of the jurisdictions wants to have to negotiate with all their neighbors to make service changes to routes that start and end within their own borders.

DC's MetroRail system has a very large share of rides that begin in one jurisdiction and end in another, so it makes a lot of sense for it all to be managed centrally by WMATA. But I suspect there aren't a lot of commuter train rides that begin in Connecticut and end in Suffolk County (or that start in Baltimore and end in Prince William County). Given that, just how would a MetroNorth/LIRR merger help riders?

The main benefit from the proposed merger would be solving issues with Penn Station: you could eliminate some of the separate concourses, easing overcrowding; by through-running trains, you could also increase the number of trains serviced without having to build an expensive new station.

Sure, agencies are created whenever multiple states need to share two sides of a public transit solution (or bridges, tunnels, etc.). e.g., Port Authority of NY & NJ, PATH, Delaware River Port Authority (PA & NJ), etc.. they are found virtually everywhere where cities exist near borders.

I should have clarified. I was talking about public transport specifally for light rail. The article was about subways specifically.

NJ Transit and SEPTA share the link between Trenton and Philadelphia. Metro-North and NJ Transit also have shared service into upstate regions of New York on two lines.

> Does any cross state combined public transit effort even exist in the US?

Yes, AMTRAK [0]

[0] https://history.amtrak.com/amtraks-history/historic-timeline

Amtrak is the federal government. I am talking about one states public rail system connecting with another states rail system. New York to New Jersey was given as an example in the parent post.

Amtrak is great for vacations or other travel but I don’t thinks its use case is daily commuters. At least it’s not the use case for the majority of the US.

The Amtrak Cascades line is actually state owned and operated: http://www.amtrakcascades.com/about One example, at least!

Chicago's Regional Transit Authority operates the Metra which spreads into northern Indiana and Southern Wisconsin

> Metra which spreads into northern Indiana

You're thinking of the South Shore Line [0], which is entirely independent of Metra [1] and the RTA.

> and Southern Wisconsin

Metra has just one stop on one line in Wisconsin (Kenosha on the Union Pacific North line).

[0]: http://www.mysouthshoreline.com/

[1]: https://metrarail.com/maps-schedules/system-map

You are correct, but the planned extension of the CTA Red Line includes a new station that is shared with the South Shore, creating a brand-new connection ability that never existed before. The planning for the new Union Station to McCormick Place link involves connecting the CTA, Metra and South Shore lines.

If you are standing in downtown Chicago, you can jump onto a train and conveniently go to O'Hare International (Blue Line CTA), Midway International (Orange Line CTA), Milwaukee International (Amtrak), and South Bend Regional (South Shore). South Bend has daily non-stops to NYC so I'm not sure why it's called a regional airport, but, well, there you go.

I'm confused by what possible efficiencies can be generated by connecting the Red Line and the South Shore line. I presume that most riders on the SS line are commuting to work in or near the Loop. The SS line is undoubtedly faster than the Red Line. Furthermore, there's already Metra lines in existence around the area of the proposed Red Line extension.

Furthermore, I'm not sure what good a Union Station to McCormick Place link would do. Union Station is already within a 20 minute walk to every single CTA/Metra/SS line. Furthermore no one really lives/works near McCormick Place. It's primarily a convention center.

It has an express, and it still counts.

Although it really sucks when someone misses the correct train and you have to drive to Winthrop Harbor to pick them up.

Commuter rail ought to optimize for seating and ride comfort on long journeys with infrequent stops. Subways ought to optimize for capacity and efficiency on short, mostly standing-room trips within densely populated areas. Unification with subway rolling stock would be awful for commuter rail riders’ comfort and willingness to use the system. Unification with commuter rail rolling stock would be awful for subway crowding.

No one is suggesting unifying subways and commuter rail. Note that LIRR, Metro North, and NJ Transit are all commuter rail. (FWIW commuter rail is a bad term we use here in the states - something like “regional” rail would make more sense.)

It's a decent article but at the end I found it glaringly silent on the current state of the "lure of the suburbs." Isn't that waning? That seems important here.

To summarize, the article is premised on "three broad lines of history":

- lure of the suburbs

- battles over control

- deferred maintenance costs

It concludes by noting how battles and costs are still big factors, but nothing on the "lure of the suburbs." It just mentions in passing that the city population is growing, but doesn't connect any dots. The diagram makes it seem like that yellow "line of history" is still sky high off the charts.

But there is in fact a big migration back into the cities. People don't want McMansions in the burbs as much. Empty nester baby boomers are moving back too. This is undoubtably why the article is written in the first place and what is placing more pressure on overcoming the other two problems.

Everyone slices the numbers differently, but it's pretty hard to say the suburbs are dying out based on the data. https://www.vox.com/2015/1/22/7871687/death-suburbs-myth

"This shows that at almost every age level, the country continued to de-urbanize from 2000 to 2013. Various trend pieces about empty nesters moving to the big city simply aren't reflected in the data."

My experience is that people are overall more willing to trade space for urban convenience than they used to be, but still not so willing/able to trade both space and affordability for it. In places where the suburbs are also stupidly expensive - NY and CA - you don't see it as much. Everywhere else in the US - where property in the suburbs gets dramatically cheaper once you're five, ten, fifteen miles outside the city center - that's hard to resist.

Didn't say that. I said that lure of the suburbs is waning. Skyrocketing costs in cities are a symptom of that and the Vox article concedes as much.

You said "there is in fact a big migration back into the cities" and that's the part that the numbers disagree with.

Instead, there has been a slowdown in the continuing move to the suburbs. And a suburb-centered population is one that still won't move so quickly on transit and the like, at a state and national level.

No, population growth in the burbs alone does not disprove migration into the cities. As I said, skyrocketing costs in the cities are evidence of this. With respect to city transportation infrastructure, wealthier people moving in and developers building up capacity are significant forces in affecting policy.

Indeed, a somewhat underappreciated story from the mid to late 20th century in America is that cities were dying. I grew up near the tail end of that era so I well remember it. If you look at the population peaks for a lot of big cities in America in the 20th century (New York, Chicago, Seattle, San Francisco, Detroit, etc.) you'll see a peak in the early to mid 20th century followed by a long decline. There are many reasons for this phenomenon, not the least of which is rampant racism, but a major consequence of it was that shrinking big city populations led to economic decline, rising crime, diminished tax base, diminished spending on core city infrastructure, and overall a feeling of helplessness and crisis in those cities. It was only with changing demographics and a transformed economy where urban centers became desirable places to live again and economically vital again (starting around the '80s and really taking off through the '90s and early 2000s) that big cities rebounded and started growing again.

But the infrastructure debt from all those years of decay still looms, and cities will be paying that off for decades to come.

What’s fascinating to me about this history is that most of NYC’s subways system was built by two private competing firms, the IRT and BRT. The city then became heavily involved first by creating the competing but government financed IND, then with price controls, and finally a full takeover. Once it was a publicly owned monopoly the NYC Subway’s long decline began.

Pretty much the same thing happened in SF but with street cars instead of subways.

This is spot on and the same observation I had, yet I'm puzzled as to why it was downvoted. A system that profits simply runs better. The best rail systems in the world in Japan are for-profit.

https://www.citylab.com/transportation/2012/05/secret-tokyos... (<-- also from citylab.com)


The big difference is not for-profit status, but the fact that Tokyo's and Hong Kong's subway operators own much of the property around the stations, and make a ton of money leasing it: http://money.cnn.com/2015/03/30/news/hong-kong-mtr-subway-pr.... Economically that makes a lot of sense. Transit infrastructure creates a positive externality: it benefits not only the rider, but the shop or office that the rider goes to. NYC's MTA can only recover from one side of the transaction: the rider. JR and MTR, as major landlords around the stations, can recover from both sides, capturing some of the positive externality.

I think that is a stretch of a statement. There are plenty of public services that work great as well as many private services that absolutely suck.

It's a little more complicated. The first NYC subway was built by the city and operated by IRT. Part of the IRT system was then built with private financing. Most of the system was built under the "dual contracts" with IRT and BMT, where the city designed and built the systems and the IRT and BMT operated them under lease: https://en.wikipedia.org/wiki/History_of_the_New_York_City_S....

Your sample size is pretty small though. Stockholm and Vienna are very good examples of working public subway systems.

That's not quite correct. The original subway lines were operated by IRT and BRT, but they weren't private ventures. The city paid to build most of the lines and leased them to the operators. The city always maintained majority ownership even when they brought in private investors.

A name mentioned twice might have warranted a paragraph. Robert Moses, the subject of "The Power Broker" a Pulitzer Prize-winning biography (worth an audiable credit).

He ran the Triborough and build a lot of bridges, parks, parkways etc. Most of those bridges do not have rail decks, because he also believed the future was cars, and rail would compete with his source of revenue, toll fees.

It is hard to overstate the impact that one person had on the NY infrastructure, but this article very much understates his impact.

User costs for transit were pretty low in the early 1900's. The article says that in 1947 they finally increased the fair from 5¢ to 10¢ -- $0.61 to $1.12 in 2018 dollars. New York's fare is now $2.75, and in my city it's closer to $5.

On one hand, transit should be heavily subsidized and not profitable because profit motives in infrastructure choke the force multiplier economic effects of having said infrastructure. You need some kind of mechanism to disincentivize overuse though, so the 5 or 10c fare worked at the time for that purpose.

On the flipside, the NY metro is so decrepid and underbuilt for its demand if you didn't have inflating fares driving potential travelers to clog up other, less efficient transit systems the lines for trains would be out in the streets nearly 24 hours a day.

Profit motives in infrastructure don't necessarily choke anything. If you want cheap fares, ensure there is some competition. And if you really want to, you can subsidize fares. They don't need subsidies in Tokyo though, all of the companies running train lines are generally profitable.

This was an April fools joke but hit's every good policy idea NYC could do spot on. https://pedestrianobservations.com/2018/04/01/new-york-area-...

From a local controlling agency, from unifying our commuter rails, from joining with the path. All these things would be good for the city and people who live here but I cry because I'll never see a sliver of it in my lifetime.

If you want to compare the development of Moscow subway from 1935 till now and planned 2020 https://stroi.mos.ru/mobile/metro

And the map as a pdf https://stroi.mos.ru/uploads/media/file/0001/61/81fbd34f4412...

Apologies for the strong tone, but how many articles do we need to actually get something done? It's exhausting to see article after article without action.

Ah. You've not lived the experience that's Toronto transit:

1. Eglinton Crosstown started in the 1990s. They started boring, and when Mike Harris became premier he cancelled the project, so they filled in the existing tunnel.

2. Sheppard subway (aka. the Stubway) built despite objections of planners, who said that it would be woefully underused. Built anyways because of political pull. 16 years later, still woefully underused.

3. Transit City proposed and supported by provincial government. First, funding was reduced by the provincial Liberals. Then, Rob Ford, claiming a "war on the car!" cancelled it, pushing all the money into...

4. The Scarborough Subway. Originally estimated at a 3-stop 2B line, it's now a 1-stop 3.5B+ line - and its cost keeps climbing. All politicians support it because Scarborough is vote-rich territory, despite its ridership projections being lower than that of the Sheppard subway. It's a frickkin mess.

5. Meanwhile, the Relief Line (which everyone acknowledges is needed) languishes because of the perception that it'll serve the "downtown elite".

I mean - it's not all doom and gloom I suppose: the Crosstown is finally being built, again, 25 years later.

It's remarkable that an article that in-depth managed to not make a single mention of the word "union", when unions played the central role in driving up costs.

I think that points to the meta problem hamstringing development of subways, and so many other things: the hold unions have over not just politics, but culture, and especially academic culture.

That's a pity. NYC subway could greatly benefit form circular lines or lines that go from west to east in addition to those from north to south. It's often required to take completely convoluted subway routes because of lacking connections.

Is there an possibility of a smarter above ground (read: bus) system? What would be the effect of small buses running more often? Perhaps with a technology component that allows them to effect the timing of the traffic lights?

If you could cut back on car traffic - which is a positive regardless - then buses are going to benefit.

A bus can carry about 100 people max. A subway train can carry 1,000.

About 5.7 million people ride the NYC subways every day, buses aren't going to cut it.

For a real-world NYC example, check out the planning for buses to temporarily replace the L train: https://www.nytimes.com/2018/04/03/nyregion/l-train-shutdown...

Yes. But other than rush hour or special events how often is that 1000 ppl capacity reached? Shouldn't the whole system be more flexible? Does everyone have to leave work at the same time? How economical is it to run subway trains with few passengers?

It seems that a whole lot of resources (i.e., money, including ongoing maintenance) is being tossed at outlier situations.

> But other than rush hour or special events how often is that 1000 ppl capacity reached?

A single train can carry 1,000 people. It's not a "rush-hour event."

> Shouldn't the whole system be more flexible? Does everyone have to leave work at the same time?

The work day is still roughly 9-5, nothing the MTA does is going to change employers' opinions on scheduling. Good luck telling your boss you want to come in 2 hours late to avoid subway overcrowding.

> How economical is it to run subway trains with few passengers?

Train frequency is slowed down significantly during non-peak hours.

Don't get me wrong, there are efficiencies to be had -- like not having a live human at every station 24/7 -- but unions fight those measures.

> It seems that a whole lot of resources (i.e., money, including ongoing maintenance) is being tossed at outlier situations.

I'm not sure what you're proposing. Buses? Again, buses won't work -- see above.

They could massively increase fares but that is essentially a tax on poor people.

What they need is for both the state and city to cough up more money to address crumbling infrastructure and overcrowding.

> A single train can carry 1,000 people. It's not a "rush-hour event."

> Train frequency is slowed down significantly during non-peak hours.

Agreed. So with the exception of the outlier events there is A LOT of idle capacity. Capacity that was VERY expensive to build and maintain.

Traditional buses? Or a rethinking of above ground transportation, congestion, flexible work hours, whatever, etc.

BTW, I didn't notice anything here about off-hours / peak-hours pricing.


Given the cost to add more (excess) subway capacity, I don't see the warm in a pause and asking "what if"? What they tried once, where the only goal was a twist on the status quo, isn't the type of mindset I'm proposing.

p.s. If climate change goes as planned, (lower) Manhattan is going to be prone to more and more flooding, etc. Does it make sense to sink more money into the ground, for unrealized capacity?

I understand your example. Unfortunately, it's red ocean. We need a blue ocean approach at this point.

> there is A LOT of idle capacity

The MTA's biggest problems right now are overcrowding and old, crumbling infrastructure, not idle capacity.

This article just pointed out we're chasing our tails so much on those 2 problems we don't even think about (gasp) adding new subway lines, although it argues because we're more committed to cars as a form a transit.

> rethinking of above ground transportation, congestion, flexible work hours, whatever, etc.

Since the Bloomberg Era the city transportation dept has added thousands of miles of bike lanes but there's not much more it can do to "rethink" above-ground transportation.

Flexible work hours are just a nonstarter for most companies. Starbucks going to let you start your shift late? Goldman Sachs? No.

> where the only goal was a twist on the status quo,

What constitutes a twist? What specific policy proposals would change the reality of public transit? DeBlasio proposed a Brooklyn trolley line -- the kind of "out of the box" thinking you're proposing -- and it's gotten nowhere.

> I didn't notice anything here about off-hours / peak-hours pricing

The pricing is always the same it's just that the trains run at different frequencies. During rush hour they can come every 3-4 minutes, in the middle of the night every 30 minutes.

There are no magical fixes to NYC's transit problems. It just requires more money (via higher taxes) and better management.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact