Should New York Rebuild the Subways?

After Hurricane Katrina, some people argued that we shouldn’t rebuild New Orleans, not simply because it was below sea level but because the city was economically and politically dysfunctional. The same argument could be made for the New York City subway system, which was so heavily damaged by Sandy that repairing it could cost “tens of billions of dollars.”

You could always swim to work.

It’s not just the subways, of course: the entire transit system has been damaged. But in the suburbs, at least, buses on streets can easily substitute for rail.

The issue with the subways is not that the cost of repairing the system is so high but that maybe it doesn’t make sense anyway. New York is the only city in the country that truly depends on rail transit. More than 60 percent of commuters in Brooklyn, the Bronx, and Manhattan, and 51 percent of those in Queens, get to work on transit. This transit, and particularly the subways, supports ridiculously high population and job densities: Manhattan is about 20 times denser, and Brooklyn and the Bronx 10 times denser, than they would be if we built them today.

Transit, particularly subways, is vital to those densities because the city’s street network–particularly in Manhattan–simply cannot support all the density by itself. New York subways are the only rail lines in America that move more people per rail mile than urban freeways move per lane mile. Having them thus significantly increases the ability to move people into, around, and out of Manhattan and the other boroughs.

As long as New York already had a subway, it probably made sense to maintain it. But building new subways, such as the Second Avenue subway which is costing more than $2 billion a mile, makes no sense. Will it make sense to perform costly repairs of the subways heavily damaged by Sandy?

There are those who argue that density has a great economic value and that all cities would be denser if it weren’t for barriers put in the way of density. On the other hand, if densities were lower, the damage from storms such as Sandy or other events such as earthquakes would be a lot lower.

Operating and maintaining New York’s transit system costs $10 billion a year more than fare collections. While increasing fares by an average of $2.50 per ride could cover those costs, this wouldn’t be enough as the system isn’t being maintained to a state of good repair. Most of the subsidies come from auto users, out of either federal gas taxes or bridge tolls that are diverted to transit.

There are two alternatives to rebuilding the subways. The drastic alternative is to simply let the city fend for itself without subways. A more realistic alternative would be to convert the subways into underground busways. Electric buses could move just about as many people as the subways do with far less infrastructure.

Battery-powered buses in particular would require almost no infrastructure other than rechargers (and the tunnels themselves, of course, which as far as I know weren’t damaged by the storm). At eighteen feet in height, the tunnels are tall enough for double-decker buses, which should be able to move about as many people per hour as the subway trains. With minimal added infrastructure, the buses could even be driverless, making them far less expensive to operate and maintain than rails.

It may turn out that only a few of New York’s 400-plus route miles of subways were harmed by the storm. But if it was significantly more, the city should seriously consider beginning a transition from rail subways to bus subways.


40 thoughts on “Should New York Rebuild the Subways?

  1. msetty

    Randal, on some issues I sometimes think you can be very thoughtful, but you’ve “gone off the deep end” (sic) with this one.

    Sure, you can move 800+ buses per hour per lane as in the Lincoln Tunnel, BUT YOU HAVE TO HAVE THE TERMINAL CAPACITY to handle those volumes, too. The Lincoln Tunnel ONLY WORKS because you have the Port Authority Bus Terminal on the other end, with 200+ bus bays handling about the same number of riders than several (e.g., less than 10) rail platforms handled at the Times Square station alone. Get real, BUSES DO NOT HAVE, AND NEVER WILL HAVE THE CAPACITY OF RAIL WHEN THE MASSIVE REQUIREMENTS FOR TERMINALS ARE CONSIDERED. This is a law of geometry and physics, NOT POLITICS OR ANYTHING ELSE.

    You anti-rail jihad has now gone utterly insane, which I was afraid would be predictable after this latest storm. Of course I expect some of the usual crew of f—ing bullshitters here to try to back you up, regardless of geometry and physics. So be it.

    I can also guarantee you that the the billions needed to rebuild the subways WILL be spent, along with probably $30 billion+ on storm surge barriers surrounding New York Bay. Politics now will ensure this outcome. Wall Street will ensure it and maybe finance large parts of it, too.

    Frank Reply:

    Bring the rhetoric and name calling down a notch, Mr. Setty. Maybe a trip to the family ranch in Napa would help you relax a bit.

    C. P. Zilliacus Reply:

    msetty wrote:

    I can also guarantee you that the the billions needed to rebuild the subways WILL be spent, along with probably $30 billion+ on storm surge barriers surrounding New York Bay. Politics now will ensure this outcome. Wall Street will ensure it and maybe finance large parts of it, too.

    Mr. Setty, New York (City and State) have been warned repeatedly that a storm surge presented a threat to tunnels (rail and highway). See this N.Y. Times article for some details.

    Regarding billions that may be spent on surge barriers below New York City, I will leave that to others to discuss. But why didn’t the NYMTA and the PANYNJ have flood barriers in their highway and transit tunnels? In the scheme of things, such barriers are relatively cheap.

    Dan Reply:

    But why didn’t the NYMTA and the PANYNJ have flood barriers in their highway and transit tunnels? In the scheme of things, such barriers are relatively cheap.

    This is the same question thousands of engineers ask every day about every aspect of our public infrastructure. There is little political will to do maintenance and prevention.

    And to make investments to harden infra against events exacerbated by man-made climate change? Kiss of political death as soon as the lobbying firms hear about it.


    metrosucks Reply:

    Proof of “man-made” climate change exacerbation, and don’t like to some left of Che website either, asshole planner.

    metrosucks Reply:

    I agree with Frank. I think msetty should go visit his family ranch, and maybe get some anger counseling too, before he blows an artery. After all, these are just ideas here on this blog; no government goons are going to be goose-stepping to Manhattan to tear down subway rails because Randal suggested it.

  2. Neal Meyer


    While I was working in London 5-6 years ago, I purchased several books on the Tube system (here is one). One of the things I learned while reading up on the history of the Tube system was that political rulers in London (and presumably Paris and elsewhere) had to face the exact same decision during the early – mid 20th century that New York City faced: what to do when your subway or rail system cannot financially support itself? Edwin Mills asks the same question in his textbook Urban Economics.

    Now, you yourself (and readers of this blog) already know the answer as to what happened during the 20th century and urban rail. Rail lines were torn up in the vast majority of American cities, save the really big cities – like New York. London, Paris, and other European cities that were already large before the onset of the automobile also held onto rail and subway systems.

    A big difference between then and now is that what has happened is what I call the federalization of American life. It’s one thing to pose the question of what to do when something like a subway or urban rail system can’t support itself? It’s another when you start asking a question like, “how many taxpayers are going to be forced to pay for the propping up or other rehabilitation of a rail or subway system like New York’s?” If your taxpayer base is the New York City (8 million), or the metropolitan area (16-17 million or so), then that’s one thing, since New Yorkers are the ones who use the New York subways. However, thanks to 20th century American “Progressivism”, people like myself who harbor old fashioned political ideas like federalism are thought of as having a Confederate flag hanging on my bedroom wall. Hence, every issue in American life is now federalized, so that means that it is highly likely that 310+ million Americans are in one way or another going to be paying for the rehabilitation of the New York subway system, and there’s something really, really wrong with that.

    In other words, if you are a defender or a believer that “government is good”, then fine. Then tell me which issues are local, which ones are regional, and which ones are national or federal? And, more importantly, why is that the case?

    bennett Reply:

    That’s a great question and one in which volumes have been written. Surely a comment on a blog post won’t even begin to scratch the surface. But for the sake of conversation, let’s take the issue of disaster relief. Is disaster relief a local, regional or federal issue? The answer is yes 😉

    Obviously the impacts of a disaster are felt on a local and regional scale. The primary response, plans and recovery has to (and does) come from local, regional and state governments. However when the resources needed to respond to the disaster exceed the ability of local, regional and state governments the feds step in to offer assistance. This is the way it currently works and IMHO rightly so.

    When I compare Frankenstorm to Katrina 2 things stand out. 1st, the mayors of the mayors cities effected and the governors of NY and NJ had specific plans in place to deal with such an event, even if the event was unprecedented. 2nd the federal assistance for Sandy was prompt due to foresight and preparation. Obviously the intensity of the two storms were different, but almost 2,000 people died during Katrina in New Orleans alone. It looks like the death toll for all the states affected by Sandy may not even reach 100. What’s the difference? IMHO, functional governments (at all scales) vs. dysfunctional governments (at all scales).

    C. P. Zilliacus Reply:

    Funny how all of the “rugged individualists” in places like Florida don’t seem to object to the federal government coming in to help them after a hurricane strike.

    Jardinero1 Reply:

    I would like to put to rest some of the myths about disaster relief in the USA.

    First of all, FEMA is the smallest player in the disaster relief game. The largest players are the insurance companies who provide the bulk of the payments after a disaster. Second, moneywise, are local responders, and national guard units, third are not for profits. FEMA does little to nothing, moneywise or organizationally during or after a disaster. FFIP is administered through FEMA, but adjusting of losses and payments are handled through private insurance carriers, not the government. Federal disaster loans and grants are handled through the SBA and FEMA respectively. The number of grants handed out is minuscule compared to insurance payments.

    My personal opinion and that of my neighbors is that FEMA is something we could probably do without and everything would turn out about the same. I say that as an insurance agent and as one who went through Hurricanes Rita and Ike.

    bennett Reply:

    On the flip side, if you are a believer that “government is bad”, then fine. Tell me what is the proper response, if any, to a natural disaster. Is this another Tea Party “let ’em die” moment?

    Frank Reply:

    Putting on my minarchist/constitutionalist hat, what the federal government does is supposed to benefit all directly. Under this paradigm, states have the responsibility/authority to respond to a natural disaster.

    Putting on my anarcho-capitalist hat, private agencies, with which residents voluntarily associate, would respond to a natural disaster. Without crowding from government, these agencies might have more resources with which to respond.

    Please don’t bother responding using the tinfoil hat fallacy. Or the Somalia fallacy. Or any fallacy.

    bennett Reply:

    My understanding is that states have the responsibility/authority to respond to a natural disaster and that FEMA is there to assist the states. I also don’t think it a huge cognitive leap to understand how FEMA assisting states in need is a benefit to all, particularly in the financial epicenter of the country, not to mention seaports, air hubs, etc.

    As for your anarcho-capitalist hat, sure, it’s a great theory. Here in the real world we have to operate outside of the theoretical vacuum. The reality is private agencies and markets can’t and don’t fix everything.

    Consider my fallacy a response to your fallacy.

    Frank Reply:

    If the government didn’t crowd private entities out of disaster relief, perhaps more entities would exist. If people weren’t taxed for disaster relief, perhaps they would obtain disaster relief through voluntary transactions from free-market entities. Of course this system can’t be perfect; no system is. But it’s preferable to coercion, IMO.

    Please note that I’m not speaking in absolutes. I have couched my arguments with conditionals like “perhaps” and “might”.

    You have not shown how I have committed a fallacy in my hypothesizing, btw. However, the use of the “real world” fallacy is a cliched and tired attempt to discredit hypothetical thinking.

    Frank Reply:

    And after reading MJ’s comments, I have to agree that this is another “opportunity to experiment with more bottom-up (and possibly entrepreneurial) responses”.

    BTW, how many houses did the federal government effectively replace post-Katrina compared to private organizations like Habitat for Humanity, Brad Pitt’s “Make It Right”, etc.?

    And how do apartment dwellers in Oregon or Washington benefit from federal subsidization of millionaires who live on tenuous strips of sand and have beach front property?

    Jardinero1 Reply:

    See my reply to CP Zilliacus above. Nearly all disaster relief is bottom up and privately financed.

    Fred_Z Reply:

    The Tea Party likes to “let ’em die”?

    I had no idea Obama was in the Tea Party.

    gecko55 Reply:

    And guess which state has the largest net tax outflow? Yep, that would be New York. $71,732,015,000 in 2008.

    gecko55 Reply:

    Also, I guess Mr Meyer lives in Texas. How does it stack up in terms of taxes. Oh, I see. No 5 … in inflows. $20,768,262,000 in 2008.

    Maybe you should rethink this bit: “… it is highly likely that 310+ million Americans are in one way or another going to be paying for the rehabilitation of the New York subway system, and there’s something really, really wrong with that.”

    Neal Meyer Reply:

    Consider this a reply to all other replies.

    At bottom, socialism never treats anyone fairly. All of the utilitarian type arguments about inflows and outflows of taxes, who manages to capture what from whom, and which governmental entity is supposed to be responsible for what (and there are a bottomless number of websites and articles that argue over all these issues), never quite seem to get around to touching that fundamental observation about government. And, I would add, it is something that many people have a vested interest in not admitting to.

    And yet, therein lies quite a bit of the appeal of the power of the state to many – that wielding government power will somehow level things out, tip the scales in their direction, or make life more fair. The story we are told is that sometimes you give, sometimes you take.

    Yet, what irks many people is when the results still don’t match up to the intentions, watching the powerful still manage to recapture the state’s spoils for themselves, or watching interest groups steer it in their direction. Then the only thing people can come up with is growing leviathan even further.

    Rinse and repeat.

  3. bennett

    I wonder what the cost would be to retrofit the subways to accommodate buses. How much will it cost to remove 400+ miles of rail? If a weather event like this were to happen again, how much lower would the repair costs be? How much would newly designed buses tailored to fit the NYC subway cost, seeing as vehicles with that specific chassis do not exist and would have to be designed and manufactured from the ground up ($500k each?)?

    My guess is that overall the savings would be marginal as many of the costs associated with maintaining the subway are not rail specific but also electrical, station and tunnel specific. The up-front capital costs would be huge and probably large enough to make antiplanner heads explode.

    I’m with msetty on this one. This seems like a post in which the intent is to provoke transit advocates and rail-transit-haters alike. Not only are the financial savings likely non-existent (at least in the short term), the logistical feasibility is borderline impossible.

    I give Mr. O’Toole credit though. It’s only once or twice a year where his posts seem to dive off the deep end. If it were my blog that frequency would be much higher for sure.

    Jardinero1 Reply:

    I like Randal’s idea. I would go one further and toll, heavily, all the streets in Manhattan to fund the infrastructure changes. This would clear most of the single occupant autos off the streets and allow more buses to run at surface level and at a higher rate of speed.

    The Antiplanner Reply:

    Already there. See Ending Congestion by Refinancing Highways.

    C. P. Zilliacus Reply:

    In a perfect world, impose a London-style congestion cordon charge and give a discount to HOV-3 (or HOV-4) vehicles.

    But for now, it is apparently going to be HOV-3 at most entrances to Manhattan:

    N.Y. Post: Mayor mandates car passenger minimums in Manhattan

  4. Dan

    I wonder if any New Yorkers will read this piece, and whether it will provoke a mild chuckle or some coffee through the nose. Na ga happin. But thanks for the laugh Randal!


  5. Dave Brough

    “…two alternatives (no) subways, and convert the subways into underground busways could move just about as many people as the subways do with far less infrastructure.”
    Even better, convert to a dual-mode rail system – robocar suspended from the ceiling traveling at .5 second headway and on-demand. This solution would not only circumvent the flooding issue, it would double the capacity of the system and make it 24/7 hr w/no interruptions. Moreover, when not flooded, the surface portion could still be used and, with suitable connector points, could also carry freight, this further increasing capacity and utilization. The suspended technology is proven and available.
    The dual-mode technology is realistic and do-able.
    The only issue would be that of dealing the unions, which is why we enlist Jesse Ventura to implement it.

    LazyReader Reply:

    Elevated microcars? It would be just like the cars. Vehicles that fit only a few people at a time trying to rush around town. If commuters try to get to one location like say a central business district then it’s still gonna have limited capacity. It’s no different than all the other personal rapid transit systems and the advocates come off as some sort of crank. If were gonna replace the automobile why do it with something not only like the car but really not as good as a car. It would be highly impractical in cities (not enough capacity) and suburbs (guideway too expensive as the cost whether in the city or the suburbs of building and maintaining elevated guideways). Then in suburbs of course there is the aesthetic detail and the possible noise it may generate. If it’s elevated how do you evacuate without a fire ladder or handle disabled passengers.

    Dave Brough Reply:

    “Elevated cars that fit only a few people at a time trying to rush around town.”
    Exactly. Except not ‘trying’ to rush around town, because they’re grade separated, SUCCEEDING in rushing around town.

    “It’s no different than all the other personal rapid transit systems”

    Exactly. Well, because it will also run on the surface (dual-mode), it is better than guideway-only PRT

    “If were gonna replace the automobile why do it with something not only like the car but really not as good as a car.”

    Exactly, why not replace it with something better than

    “It would be highly impractical in cities (not enough capacity)”

    Au contraire. Each subway tube can carry 4 lanes of pods: two running on the roadway (these subject to pre-emption during flooding, and two hanging from roof-mounted guideway. The cars and either join and platoon (like existing subway cars) or run at .5-second intervals. Because there are two subway tubes, double the number. Do the math and you’ll come up with greater capacity than any subway – and every passenger gets a seat and goes direct to destination.

    “the cost whether in the city or the suburbs of building and maintaining elevated guideways (is too expensive).”
    Hardly. The cost of suspending a rail from the roof of a tunnel is a mere fraction of anything else. For the surface car, the only cost is ripping up the rails. The cost of installing it in the burbs is about 1/10th that of contemporary rail.

    “In suburbs of course there is the aesthetic detail”

    Negative. You’re not running the guideway down residential streets. You run it along the busier roadways, along floodways, along freeways: you use robocar mode to get you the few miles from home to guideway.

    “and the possible noise it may generate.”

    Stand beside any freeway or commuter rail line and tell me it’s not noisy. The thing about guideway is that the rolling portion, the wheels, would be secreted within the channel of the guideway, thus keeping the noise down. And remember, there’s virtually no noise from our EV robocar getting us from front door to the guideway entrance.

    “If it’s elevated how do you evacuate without a fire ladder”

    You evacuate WITH a fire ladder, and more likely, a bucket truck.

    “or handle disabled passengers.”

    Who says the pods couldn’t handle wheelchairs? They will.

  6. TommiM

    I have better idea. Why don’t you flood all the subway tunnels and use electric submarines? There would be minimal infrastructure costs, submarines can be driverless and submarines can use nuclear power!

  7. MJ

    On a more serious note, I think there is a strong argument to be made for the federal government not to be involved in rebuilding the subway network. Much like New Orleans in the wake of Katrina, this event provides an opportunity to experiment with more bottom-up (and possibly entrepreneurial) responses to urban issues, transportation included.

    Randal alludes to the outrageously expensive 2nd Ave. Subway project, but I think the question can be asked more generally about how the subway network would be operated in the absence of federal subsidies. There are at least two possible outcomes. The network could be left to local and state authorities, who would then have to make more carefully considered decisions about pricing, subsidy, and service levels. Alternatively, NYC could experiment with private ownership and operation of the network. Again, I would expect some dramatic shifts in terms of pricing and service offerings (perhaps along the lines of time-of-day pricing), but I would also expect to see more entrepreneurial efforts toward marketing to and attracting passengers, improved maintenance practices, and less tolerance for crime and antisocial behavior aboard trains and at stations.

    Frank Reply:

    Agree 100% on both counts.

    As for subways, let Virgin and others bid on contracts to repair and/or operate the NYC subway. After trains in the UK were operated by concessionaires, new amenities were added, such as being able to pay for tickets with a mobile phone.

    I haven’t been on a NYC subway in over five years, but I typically avoided the subways for taxis on trips to the city due to the foul odors and intense heat.

  8. Matt Young

    Rip out therails, put in rubber wheeled technology. No need for batteries, a third rail for electricity works just as well for the rubber wheeled as the steel wheeled.

    But the real advantage, the last mile, is what buses provide. Getting them out of the tunnel to make the last mile is likely impossible in NYC.

    In San Jose CA, however, we have useless Light Rail with a surface right of way. Ripping out those tracks and replacing them with rubber tire makes a whole lot of sense because on the surface the buses can just exit the right of way, run on the city streets and get the last mile.

    The last mile of commute is where the action is.

  9. Fred_Z

    I read the articles at the two links, and their description of the vast damage, incredible effort, huge expenses, blah, blah, blah, and my only sensation was the overpowering stench of pork.

    All of these people see endless jobs, contracts, fees, conferences, consultations and brand new hot secretaries with big boobies.

    metrosucks Reply:

    It’s always boggled my mind how laying down 2 or 3 rails plus accessories has commanded the billions of dollars it does.

    T. Caine Reply:


    Andrew Reply:

    No, most of the work is heavy civil grading, tunneling and structures. Just like building a road, the cost is not in placing the wear coat on top of your aggregate and asphalt or concrete base.

Leave a Reply