Imagine if someone asked you: Highly complex software project A costs $x/KLOC, highly complex project B costs $10x/KLOC, and therefore isn't project B wasteful? It's impossible to answer based on that information, and you probably know far more about software projects than you do about underground urban transit projects.
Many of the comparisons in the article seem to be similar to that, and with projects that seem far more complex, technically and politically, than what most of us deal with. Other than the fact that all the projects mentioned are called 'trains' or 'subways', and I assume are mostly underground, I don't have enough information to say they are comparable at all.
Sometimes, there is no way for the layperson to analyze the situation on their own.
Yeah, no, the US is definitely an outlier, building virtually any public infrastructure costs way more than it should. Tokyo can still build underground subway lines at 1/8th the cost:
The answer why is complex, but the two big factors seem to be red tape, specifically that standards are overengineered and it takes forever to get approvals, and lack of experience and existing construction infrastructure, so instead of having a well-greased machine like Singapore's that churns out line after line in rapid succession, you need to ramp up and ramp down for every single project.
What the article refers to is a cost estimate for the Koto Waterfront line, which hasn't been approved at this point, let alone started construction. Whether it will be built on budget is anyone's guess.
The route would also be in a relatively recently built area, so it's not hard to imagine why it would be cheaper to plan and execute that a route under Manhattan.
I don't know very much about NYC, but the comparison does not seem fair.
Tokyo's newest subway the Fukutoshin line, spanning some of Tokyo's most congested areas (Shibuya, Shinjuku, Ikebukuro), cost around $250m/km, so still around 1/4 of the NYC cost.
I have a very good friend who actually worked on this project. He was a supervisor level guy and would send me pictures from the tunnels. I believe his exact quote was "the MTA burns barrels of money like it's going out of style." The waste was unbelievably extreme. I kept telling him to write a book but he said if he did he'd never work in civil engineering again.
Sadly true, but I wonder what that means because the layman in a democracy is called upon to be knowledgeable enough to identify whether their elected politicians are mismanaging projects, or if in fact this public transit project is doing well and on the way to success.
I know the ideal is that we can decide for ourselves, but the reality is that in most things in life, we lack the expertise.
I'm not a plumber, doctor, police officer, etc. Sometimes they are obviously right or wrong. Otherwise, I can try to second-guess them based on some limited or amateur knowledge here and there, but I know that when laypeople try that in regard to my profession, it very rarely helps them and can be a recipe for disaster. They don't even know what questions to ask.
In the end, IMHO we have to decide which experts to trust - this plumber or that one; this subterranean urban transport engineer or the other one. Thanks to the open societies in modern democracies, plus modern communication, we have a plethora of experts and voices to choose from, but it ain't perfect.
The layman is only called to be knowledgeable enough to identify whether their elected politicians are mismanaging projects after hearing expert testimony.
Which is the exact same way politicians can know whether projects are not mismanaged or over-budgeted by the construction companies, etc.
that's why good and fair journalism is so important. It should connect the dots and try to explain to the layman what's going on, while ideally being honest about the opinions of the paper and not having an explicit agenda (i am not really sure if i used that word right, for me it feels kinda extreme).
I see your point, but I my feeling is that the "unknown unknowns" of the New York situation might make up a 50% difference? But we're looking at a >200% difference.
As far as I'm aware, the biggest issues with building deep tunnels like this is everything that gets in the way - other tunnels, utilities, old buildings that might fall down, and rivers. The Crossrail project in the UK has a high pice tag (£15bn) but is also much larger, and given how old London and its infrastructure is, I would imagine has had to cope with a similar number of obstacles, if not more. I imagine Paris is in a similar place given its age.
Phase One of the Second Avenue line cost $4.5bn [1]. That's about as much as it cost us to build our "Stegasaurus" subway station downtown [2]. Until elected officials lose elections as a result of cost overruns it will be prudent for leaders to divert resources to efficiently-voting public unions.
(MTA officials say the Second Avenue Subway cost as much as it did because of Manhattan's "complex underground infrastructure" as well as the fact that the New York City Subway runs all the time [3], the latter not being a requirement of Paris or London's systems.)
> the fact that the New York City Subway runs all the time
The Second Ave subway is a new line. It didn't even cross any existing lines. They obviously had to hook it up with the existing Q at one point, but I don't see why the 24/7 argument applies here?
In order to support a line running 24/7, you have to build a lot of bits and pieces to include more redundancy. For example, the two miles of Second Avenue Subway has three crossovers, all of which need multiple switches, so that trains can run on either track, allowing maintenance crews to cut power and safely work on the train lines during more quiet hours of the day with minimal disruption. Conversely, BART in San Francisco has very few crossovers. That was cheaper, as signalling and switch infrastructure is expensive, but also makes it much less granular, and as such, harder to do any sort of maintenance.
Tunnel boring, digging and other soil/earth works are the big cost factors, the required logistics, and regulatory processes (permits, permits and making sure that you have a plan when you close important lanes on streets and avenues of a big and busy city - so traffic engineering).
The crossovers probably cost so much because they're underground and require extra tunnel space. (They require extra (as in between the two-or-more tracks that are being switched) foundations above ground too, which is probably more expensive than the switch itself.)
So are subway lines switching over constantly in NYC? Or is the redundancy all over? That sounds like a high price to pay for 24/7 service.
Even the busiest lines in Tokyo are mostly single-rail. Of course, when a train breaks down, then nothing can happen, but at least you don't need 4 sets of rails, only 2 (one for each direction). There's probably way less in terms of delays because of that too. Each train is on its own set of tracks, so little cross-cutting.
>"Even the busiest lines in Tokyo are mostly single-rail."
But aren't most of the train lines in Tokyo owned and operated by separate competing companies They don't have to interoperate or not that much correct?
Once upon a time the trains in New York City were also owned by separate competing companies, there was the IND, BMT, and the IRT but they now operate as one entity under the MTA. As a result all these lines were originally built differently with different tunnel widths and platform to accommodate the different size cars. I think this level of interop adds to the complexity of the system as a whole.
> Until elected officials lose elections as a result of cost overruns it will be prudent for leaders to divert resources to efficiently-voting public unions.
Why the focus on unions, which are very weak in the U.S.? How about public officials who give work to their wealthy, powerful allies and donors?
> Why the focus on unions, which are very weak in the U.S.?
Weak in general, yes; but very strong in the public sector. They push through all sorts of measures because they control a large voting block (and often public sector unions support each other without questions; for example, if the fire fighters union is pushing for something, you can bet the police officers union will support it too). This really magnifies their power.
New York City's unions, particularly public and construction unions, are strong. In my limited experience, at hearings and meetings involving New York infrastructure, the "ayes" are supplied by the unions. The "wealthy, powerful allies and donors" are more present at zoning and related private-development discussions.
Unions are visibly and numerously present at New York's many off-season elections, e.g. judicial primaries and participatory budgeting polls.
> New York City's unions, particularly public and construction unions, are strong
Good point; unions seem stronger in NY than in other parts of the country.
However, if you are a New Yorker, I don't think you mean to claim that the wealthy and powerful aren't the beneficiaries of corruption.
IMHO, unions give workers power, balancing the scales a little. Power can be used for good or bad (or almost always, for both), but I certainly wouldn't exclude working-class people from it for not being angels. Why should they be held to a higher standard than other powerful people? If we are going to fight corruption, why pick on the working class?
> Is a heavy crane operator making $400k/year plus benefits "working class"? That sure sounds like wealthy to me.
It's absurd to imply that such a person, if real, is representative of the incomes of union members in New York. And again, why focus on that guy and not the person making tens of millions?
EDIT: Reading the articles you linked to, I understand those are the high-end of salaries of crain operators building skyscrapers in in NYC.[0] I'm not sure that's inappropriate; it's high skill work with many lives and enormous investment on the line, like an airplane pilot. Why do people think their incomes are inappropriate (if that is the implication) and not the person who makes tens of millions from their casual gaming software startup.
And that does not qualify as wealthy in NY, at least not in a way that will buy you any influence.
[0] The International Union of Operating Engineers Local 14-14B lays out the wage floor for its operators at $73.91 per hour. That’s $150,000 a year before overtime, plus benefits equal to $32.50 an hour. When the operator is behind the controls of a tower crane, he gets a $2-an-hour bump. On weekends, when cranes are moved, pay doubles. With overtime, many union members earn half a million dollars a year.
>By definition, a person using a union to extract above market wages (I.e. rents) is a rent seeker.
The definition already assumes that "market wages" reflect some optimal (for society) balance. So it's just circular reasoning.
In the real world "market wages" also reflect power struggles, and in those power struggles those paying the wages have numerous means to get the upper hand -- from forming cartels and controlling the media, to downright buying politicians and favorable legislation.
Unions help balance those power struggles to the (much more numerous) worker's side. You know, the people that, unlike Ayn Rand's heroes, actually produce something.
Not to mention that the rich are the ultimate rent-seekers, especially in this stage of the economy. And "too big to fail" to top.
>He resorted to name calling, but which of his names do you think is actuslly incorrect?
Rent seeker is already incorrect as I showed. Then there's racism (doesn't even apply in this context), nepotism (as if union members hire their siblings?), cronyism (ditto), etc.
Unions attempt to obtain a higher income than that which would be necessary to keep a factor of production (in this case labor) employed in it's current use by manipulating the social or political environment without creating new wealth. Thus, they are by definition rent seekers.
...those paying the wages have numerous means to get the upper hand -- from forming cartels and controlling the media, to downright buying politicians and favorable legislation.
You seem to be confused - it's unions, not wage payers, who have these powers. Unions buy politicians and get favorable legislation, and a union is by definition a cartel.
In contrast, cartels of wage payers are illegal, and the media is rarely favorable to their interests. But in the rare cases when wage payers form a cartel, this too is rent seeking.
You know, the people that, unlike Ayn Rand's heroes, actually produce something.
I'd suggest once you finish reading the definition of rent seeking, you might also consider reading Ayn Rand. You might be shocked to discover that all of her heroes are productive workers.
>The definition of rent seeking has nothing to do with your favorite socially optimal balance (...) "Unions attempt to obtain a higher income than that which would be necessary to keep a factor of production (in this case labor) employed in it's current use by manipulating the social or political environment without creating new wealth. Thus, they are by definition rent seekers."
Unions represent those that par excellence create new wealth, the workers. It's the managerial class that doesn't create any new wealth -- and it's that class, land owners, industry leaders, etc, that is more prone to rent seeking.
Of course a manager getting a $1 million bonus or a CEO ensuring their golden parachute while the company goes down the drain (thus hardly "creating any new wealth") is not called "rent seeking" while a construction worker getting $1 per hour more due to union bargaining is.
Note that "a higher income than that which would be necessary" is mumbo jumbo without clear definition, except if we are to accept subsistence pay.
>You seem to be confused - it's unions, not wage payers, who have these powers. Unions buy politicians and get favorable legislation, and a union is by definition a cartel.
You seem even more confused.
First, you say that "it's unions, not wage payers, who have these powers".
By this, you already seem to accept the notion that individual workers do not have these powers -- unless they organize in a union. Which was my point precisely.
If you read my comment, I didn't say that that the unions don't have power. I only said that wage payers have much power over the workers, and that unions help to alleviate that.
Your statement is agreeing with me on this, with the exception that you seem to believe that the union then gets too much power over the wage payers (which I believe is wrong).
The second problematic thing is that I enumerate which powers I believe that wage payers have: "from forming cartels and controlling the media, to downright buying politicians and favorable legislation" -- and you say that those are the powers that unions have.
Sorry, but those are by no means the powers that unions get. The unions at most offer collective bargaining powers (and usually their leadership tends to be corrupt and in bed with wage payers anyway). The Rockefellers, Hearsts, Trumps, etc of the world are more known to "forming cartels and controlling the media, to downright buying politicians and favorable legislation" than unions are.
Unions represent those that par excellence create new wealth, the workers.
But they create no new wealth themselves while seeking to capture more wealth. Hence, rent seekers.
Of course a manager getting a $1 million bonus or a CEO ensuring their golden parachute while the company goes down the drain (thus hardly "creating any new wealth") is not called "rent seeking" while a construction worker getting $1 per hour more due to union bargaining is.
Correct, a manager/CEO/construction worker negotiating higher compensation by threatening to leave is not rent seeking. That's because this factor of production would not be employed at a lower wage.
Rent seeking is a technical term with a specific definition, not a general term of derision. Unions meet that definition, CEOs (in their role as labor) mostly do not.
That's just an example of a specific union. Not a general characteristic of unions, which have welcomed black, hispanic and asian workers as far back as the 1915 or so -- at a time when repression and racism was so systematic that there were even state and federal laws against specific minorities.
Sure, $400k/year is a damn good salary, but realistically, how many of these people are managing that? We're talking 40+ hours of overtime a week to reach that salary. So it would seem that there's a shortage of people qualified to do the work.
You know personally of someone making such a salary? Can you back this up or is this just rumor? Sounds like the kind of story that gets passed along by your drunk uncle starting with "a friend of mine said that a friend of his.."
Thank you for the links- Extraordinary claims and all that. Still none of the articles claim that >400k is representative of an average operator. As the article says, France has stronger unions than we do but their costs are way lower. Personally, I've never met, or talked to anyone who has ever met an active member of a labor union. So for me they are semi-mythical organizations to begin with.
Unless union labor was a majority/plurality of the project's cost, then they're not the biggest problem with this project. Seems like you're just using them as a bugaboo to pin all the world's problems on.
Infrastructure schminfrastructure. Dig deeper. That's what we did in Los Angeles. When you get on the Red Line, you descend down something like 60-90 feet of stairs/escalators, compared to 18-30 feet in NYC.
The Vox article [1] specifically cites depth as a contributor to the Second Avenue Line's increased cost:
"the majority of the Second Avenue subway’s price tag actually comes from the stations rather than the tunnels. The line is being built unusually deep for a subway tunnel, which doesn’t do much to change the basic logistics of boring a tunnel, but does make excavation of station caverns unusually difficult."
Geology and hydrology are enormously different between those locations. Not saying you definitely can't do that, since I'm no expert either, but it's kinda silly to just say "dig deeper" like it's a no-brainer.
2nd avenue subway is muuuuch deeper than most of the 100 year old stations in manhattan, for just this very reason. And also because no one would stand for cut and cover anymore.
"Cut and cover" being what exactly? Trenching down from the surface and then filling in above the tracks? Hardly seems practical beyond a fairly shallow depth.
Yet with the exception of London, that's how most of the world's pre-existing subways were built. London is a bit of an outlier as it was very highly developed even in the late 19th century and there was a great fear of damaging existing "tall" buildings. Although even there a large amount of cut and cover tunneling exists.
> Berlin’s U55 line cost $250 million per kilometer, Paris’ Metro Line 14 cost $230 million per kilometer, and Copenhagen’s Circle Line cost $260 million per kilometer.
Are these useful comparisons? How can you possibly compare a train running the length of Manhattan to anything in Copenhagen? I want to know how much a new train line costs in Tokyo.
I wonder what those other cities are built on. The Second Ave subway's bedrock is schist [1] -- which is an extremely strong, durable type of rock.
A quick glance at this [2] engineering book indicates Copenhagen is limestone. Apparently the same is true for Paris. Any engineers here who want to chime in about how that affects cost?
> Tokyo Metro Fukutoshin Line: ¥250 billion for 8.9 km of new track. This is $280 million per km. Tokyo Metro has claimed future lines will be $500 million per km as a reason to not build future extensions.
Also note that Tokyo is much less dense than Manhattan. A better comparison would be Hong Kong, except that labor and other costs there will be dramatically less.
Thank you for providing Central Tokyo's density; it is certainly the best comparison to Manhattan in Japan. However, you are incorrect about Manhattan's density, which is 27,812.2/km²
Just to clarify for folks who don't know Tokyo: Wakoshi is the first stop on the new Fukutoshin line, and it is located in Saitama, which is not in the 23 wards composing Central Tokyo.
The answer to your Tokyo question is in the "compendium of international subway projects", linked in the article (which also contains those numbers you quoted)
As an outsider, it feels to me like the budget of any project tends to expand to the amount of capital available. When the budget is less, people tend to be more frugal. Which also means that at least some corner-cutting is done. But when the resources are vast, people insist on doing everything by the book. I see this not only in cases of public infrastructure but also in case of space programs as well(case in point, Mangalyaan). America has always been resource-rich and therefore I observe that it has a culture in which spending a lot of money is acceptable(since there is a lot of money).
This is something I've been harping on for ages. The educated public is generally aware that the US spends way more for comparable healthcare services and has a general understanding of both the public policy causes and implications for this as it relates to health outcomes but there's far less that's been written about infrastructure (seriously, pretty much every article references the pedestrianobservations blog because that's the only real writing that's easily googleable that's been done).
American infrastructure routinely costs between 5 - 10x that of comparable developed countries and, when you dig into why, it's like a litany of every failed project management idea from software, all rolled into one.
The reasons are myriad but, at it's core, the main driving factor is that America would rather spend an extra $10 non-corruptly than $1 corruptly. The obsessive focus on stamping out corruption puts rigid rules in place that stymie cost efficiency.
Until there's serious political pressure demanding serious government procurement reform, nothing will change. There's enough motivated special interests on the other side to keep America in a perpetual infrastructure deficit. But for that to happen, more smart writing and analysis needs to be focused on the problem of infrastructure procurement and how changing the way we do things could unlock huge savings and bring about the infrastructure revolution that America needs.
I'm having a hard time reading between the lines here, but it feels like you're saying that allowing corruption would lower costs, but that really goes against my intuition.
Several Japan infrastructure projects also cost huge amounts of money, and construction companies and gov't are very close.
Japan has better infrastructure. But the way they do it is by spending huge amounts of money. The only way the bullet trains exist is because of massive government spending (during Japan Railway's privatization, about $150 billion of Shinkansen debt was taken on by the tax payer).
He's not saying corruption costs less, he's saying that US procurement rules compliance eliminates the competitiveness of contract bidding.
There's basically two different construction industries in the US: contractors that do private construction, and contractors that do public works construction. On the private side, you have thousands of companies ranging in size from some random guy who does drywall all the way up to pipeline and oil rig construction contractors. On the private side, all the small guys have been weeded out because of absurdly high regulatory compliance costs(which have been designed to eliminate corruption), and so it's just a handful of extremely large companies. And knowing they have no competition, their bids are obscenely overpriced.
Every infrastructure project is its own shitty snowflake in the US—many arguments within the article stem from this. If we simply committed to more, budgeting slightly less on the assumption that kinks get worked out, and threatened cost overruns highly public, I bet things could work out.
The multitude of governments and jurisdictions problems is more worrisome however. That kink won't fix itself.
Here in London we are building the crossrail, which is 118km in length deep under London. It costs around £15bn, so that's £125 million per km. That's including the 40 planned stations.
London is far far older than NY, it's also more congested underground and a lot harder to organise logistically.
Blows my mind that the NY subway costs this much to extend.
Crossrail is not directly comparable, for starters, only 21km of crossrail is underground; Most of it's 118km length is above ground so the per-km cost of the below ground sections get averaged out over the above ground sections.
This section of the 2nd Avenue subway (it's just 2.7km long) is deep under Manhattan itself. To get a comparison on cost you would have to only compare only the section of crossrail running under central london itself.
Only a small portion of that 118 km will be in newly built tunnel and there will be relatively few newly built underground stations on Crossrail, which is why it's much cheaper (underground stations are very expensive to build). Also, the use case for Crossrail is very different from the use case of the Second Avenue Subway.
Hong Kong is mostly granite, which is significantly more difficult to tunnel than the shale-like schist bedrock of manhattan...and MTR construction reflects this fact, being more expensive per km than London, Paris, or Madrid. Even after this fact, MTR construction is still 50% cheaper than the average subway construction costs in the US.
In addition to the other comment-- As far as I understand most London trains are not 24/7. See the other HN post about this, but building 24/7 tracks is significantly more expensive because you need to build multiple independent tracks so that work and maintenance can be done on one while trains run on the others. This in turn requires more switches and complexity.
For NYC the trade off is worth it. The city absolutely requires 24/7 public transport.
Wow! Here in SF they're building cross-town BRT for $5m/mi. Blows my mind that London's subway costs that much to extend, dozens of times more expensive. And we have earthquakes to worry about, London doesn't.
Is every subway project going to be failure? Here's example from Helsinki & Espoo, Finland.
https://en.wikipedia.org/wiki/L%C3%A4nsimetro#Cost
Constant cost & schedule slips. Nobody knows when it will be actually ready. Here's the official news feed: http://www.lansimetro.fi/en/home/news.html
Yet that's still 13 new stations, 21 kilometers for €1.2 billion (estimated).
I think it's a mistake to place the blame on weak unions, not only because there's really no evidence presented for it, but also because New York City actually has very strong unions, unlike the rest of the US.
When groups that are used to struggle from a position of weakness chance into power, they rarely use it as responsibly as they should. Like the dilemma of revolutions creating even more oppressive governments, but generalized.
Now in this case, we are talking about a long tradition of established power, not about a weak group suddenly rising. But I think the pattern might still apply: The rare powerful unions are not recent underdogs, but they feel a cultural oneness with the weak ones in other places. It must be very tempting to believe that you are fighting the fight of the typical weak union when you are actually expanding on the excesses of the rare powerful ones.
Look at what the East Side Access is costing. Current estimate above $10bn, completion 2022. That involves building another level of railroad station underneath Grand Central without disrupting operations above.
If we were as efficient as Paris, Seattle's recently passed $53B ST3 package would give us 140 miles of fully underground subway. Instead, we get 62 miles, mostly above ground, and on a 30 year timescale. However, this argument was basically dismissed by transit advocates and anti-transit advocates alike. The cost problem in the US is infuriating because those who care about costs don't care about transit and those who care about transit don't care about costs.
But above ground doesn't matter as long as it is grade-separated, which is what every new foot of rail laid for Sound Transit 3 will be. (That's also one reason for the long time-frame: A new tunnel needs to be dug under downtown and a new bridge built over the ship canal to Ballard.)
A very large chunk of the Sound Transit 3 cost will be interest paid on borrowing. Because Washington State has no state bank or infrastructure bank, Sound Transit has to "borrow as it goes" by issuing bonds, building, paying back, issuing more bonds, building, and so on. It doesn't get the funds in one up-front chunk--like Vancouver did for its light rail system, from the BC and Canadian federal governments--so it's fund-as-we-go.
The borrowing limits affect the timeline, but I'd be surprised if they affect the cost in any measurable way. Bond interest rates are so low for something like this that after adjustment for expected inflation they might as well be zero.
Above ground does matter. Grade separated still mostly works from a transit perspective, but there are other concerns: how they affect surface traffic, how they can be affected by marine traffic, etc. Underground is more expensive but still demanded for a reason, and we are still paying >2x per mile for above ground vs what Paris paid for underground lines.
You should read the article (or even just the subtitle of the article) which takes the position that all that spending on one tiny little expansion is what keeps us from being able to spend larger amounts on projects that really would benefit a lot of people.
125th Street really, really needs some cross town service.
Something connecting all the way from Riverside Drive/Henry Hudson Parkway passing through the 125th Metro North Railroad station. (...and maybe even shuttling to Randals Island, why not?)
It's faster walking than it is to take the buses that run that route. Cold weather means waiting for the buses sucks, and the only time it's worth a trade is when you're carrying something heavy.
Car service in Harlem is slightly schizophrenic, even with car hailing apps and "boro cabs" (because normal yellow cabs don't operate in Harlem, for reasons I still don't understand...).
The same is true for much of the city. Other than the S at 42nd, crosstown is a big weakness of the system.
But as the article points out, there's many ambitious plans that could be done if it weren't for the insane price tags and construction times. Heck how about some reasonable way to get to LGA?
Manhattan's topology gets really strange after about 106th street on the West Side. That's not an excuse in itself, but the fact that Manhattanville is lower than its surroundings on all sides makes for some interesting tunneling. Elevated lines (like the 1 after 116) are also a problem in terms of maintenance and noise/light pollution.
> because normal yellow cabs don't operate in Harlem, for reasons I still don't understand
This is an ugly relic of NY's more violent (and more explicitly racist) past. I went to school with the kids of cabbies, and anything past the UWS was essentially off limits thanks to 30 year old stereotypes about the people living there.
Many of the comparisons in the article seem to be similar to that, and with projects that seem far more complex, technically and politically, than what most of us deal with. Other than the fact that all the projects mentioned are called 'trains' or 'subways', and I assume are mostly underground, I don't have enough information to say they are comparable at all.
Sometimes, there is no way for the layperson to analyze the situation on their own.
EDIT: Minor edits