The MTA Sticks to Its Oversize Stations
In our construction costs report, we highlighted the vast size of the station digs for Second Avenue Subway Phase 1 as one of the primary reasons for the project’s extreme costs. The project’s three new stations cost about three times as much as they should have, even keeping all other issues equal: 96th Street’s dig is about three times as long as necessary based on the trains’ length, and 72nd and 86th Street’s are about twice as long but the stations were mined rather than built cut-and-cover, raising their costs to match that of 96th each. In most comparable cases we’ve found, including Paris, Istanbul, Rome, Stockholm, and (to some extent) Berlin, station digs are barely longer than the minimum necessary for the train platform.
MTA Construction and Development has chosen to keep building oversize stations for Second Avenue Subway Phase 2, a project that despite being for the most part easier than the already-open Phase 1, is projected to cost slightly more per kilometer. Nolan Hicks at the New York Post just published a profile diagram:
The enormous size of 125th Street Station is not going to be a grand civic space. As the diagram indicates, the length of the dig past the platforms will not be accessible to passengers. Instead, it will be used for staff and mechanical rooms. Each department wants its own dedicated space, and at no point has MTA leadership told them no.
Worse, this is the station that has to be mined, since it goes under the Lexington Avenue Line. A high-cost construction technique here is unavoidable, which means that the value of avoiding extra costs is higher than at a shallow cut-and-cover dig like those of 106th and 116th Streets. Hence, the $1 billion hard cost of a single station. This is an understandable cost for a commuter rail station mined under a city center, with four tracks and long trains; on a subway, even one with trains the length of those of the New York City Subway, it is not excusable.
When we researched the case report on Phase 1, one of the things we were told is that the reason for the large size of the stations is that within the MTA, New York City Transit is the prestige agency and gets to call the shots; Capital Construction, now Construction and Development, is smaller and lacks the power to tell NYCT no, and from NYCT’s perspective, giving each department its own break rooms is free money from outside. One of the potential solutions we considered was changing the organizational chart of the agency so that C&D would be grouped with general public works and infrastructure agencies and not with NYCT.
But now the head of the MTA is Janno Lieber, who came from C&D. He knows about our report. So does C&D head Jamie Torres-Springer. When one of Torres-Springer’s staffers said a year ago that of course Second Avenue Subway needs more circulation space than Citybanan in Stockholm, since it has higher ridership (in fact, in 2019 the ridership at each of the two Citybana stations, e.g. pp. 39 and 41, was higher than at each of the three Second Avenue Subway stations), the Stockholm reference wasn’t random. They no longer make that false claim. But they stick to the conclusion that is based on this and similar false claims – namely, that it’s normal to build underground urban rail stations with digs that are twice as long as the platform.
When I call for removing Lieber and Torres-Springer from their positions, publicly, and without a soft landing, this is what I mean. They waste money, and so far, they’ve been rewarded: Phase 2 has received a Full Funding Grant Agreement (FFGA) from the United States Department of Transportation, giving federal imprimatur to the transparently overly expensive design. When they retire, their successors will get to see that incompetence and multi-billion dollar waste is rewarded, and will aim to imitate that. If, in contrast, the governor does the right thing and replaces Lieber and Torres-Springer with people who are not incurious hacks – people who don’t come from the usual milieu of political appointments in the United States but have a track record of success (which, in construction, means not hiring someone from an English-speaking country) – then the message will be the exact opposite: do a good job or else.
Pedestrian Observations 
My goodness. These clowns not only have no respect for taxpayers’ money, but also seem to have contempt for them. Building an underground cathedral for more back office spaces at such a high cost is a nice fuck you to all the people footing the bill.
But maybe this isnt so unusual because here in ontario, the govt is spending $1 billion/km on the ontario line 😡
Bond villain lair?
The first challenge here is to persuade the New York political scene that it makes sense to run the New York Subway well. This should be a winnable argument politically. If it was well run the Democratic Party would have a functional big city to point to in terms of persuading people in the rest of the country to vote for them. Plus it would help them in New York City itself and to win elections like George Santos’ congressional seat more easily.
In terms of hires though you need to hire someone who the New York political establishment would be comfortable. Tokyo, London, Paris, Seoul, Hong Kong are all cities that New York should consider nearly its equal with high ridership, multi-line metro systems. Seoul is probably a bit more brave as a pick than the others but does appear to have a very good system.
Beyond that there is a second tier where the systems are smaller but where it could be possible with cities like Singapore, Osaka, Taipei, Madrid, Santiago and Berlin. Maybe some of the lower level positions could come from those cities.
Obviously some of those cities have super high construction costs which I have ignored – but while the list can be shortened because of that some degree of pragmatism will be needed around which cities make the list so you can have a reasonable sized hiring pool.
To be clear the second tier is much less clearly defined in my head than the first and there are limitations of the medium. Stockholm with 3 lines and 460m annual passengers maybe should be included. Difficult to judge.
Madrid? The only reason that was cheap was because the build quality was utter shite. And having worked with different Spanish Contractors on design-build projects I’m surprised the system still standing with all the corners they try to cut to maximize profit. Istanbul has built 350km of metro in the last 15 years at a reasonable cost and there are some good lessons to be learned. However, the 24/7 working in urban areas and the don’t give a fuck about the surrounding communities, the low wages for workers and the blatant corruption where work is awarded to AK Party cronies means that costs need to be taken with a pinch of salt. and means that the way work is planned and executed there is not something you could easily translate to somewhere like New York.
This idea that agency-employed jobsworths who can’t be fired and get index-linked pensions, healthcare, free travel etc. for life after retirement is somehow superior/cheaper to using consultants who get none of these is directly linked to high costs is farcial. Designs have to be performed to the design codes no matter who performs them, agency standards have to be complied with, local building codes have to be complied with etc.
The city hasn’t been under AKP control in a number of years and is building its own metro lines, still at low cost, with negative support from the state.
And nowhere with functional infrastructure construction do people treat the civil service as jobsworths who get index-linked pensions, except in countries where everyone gets an index-linked pension as part of a national social security scheme. An underpaid civil service is one that gets little respect from the general public. Ditto a civil service that’s subjected to a class system, in which the engineer with 20 years of experience still has to lower their head when Torres-Springer or one of his personal staffers enters the room; in the low- and medium-cost world, that public-sector engineer with 20 years of experience runs the show and the public policy grad is the one who lowers their head.
My experience is that design “standards” and “codes” are actively weaponized by the consultant mafiosi. They are the ones that select the “standards” which are to be rubber-stamped by the pathetic and useless rump agency staff, then they eliminate alternatives that don’t meet the “standards”, then they gold plate everything in sight, pleading that their hands are tied by standards that were apparently handed down from God on stone tablets and that cannot be questioned, only obeyed.
Fucking rent-seeking scum.
The problem is that every city on this list where the working business language is English has extreme costs, through cross-pollination of bad ideas about how to privatize the state and replace it with consultants. The people who came up with the globalized system of project delivery think it worked; one of them, who’d worked in Singapore, literally didn’t believe me when I said Singaporean construction costs are high. Such people also get paid very well, which gives them clout, in the “if you’re so smart, how come you ain’t rich?” system that permeates the US as well as post-Thatcherite Britain (not just England, cc Borners).
Yeah, I know. Its a British not English system, which extends to the Republic of Ireland. Scots brag about lower electrification costs because the Sewell convention protects them from treasury micro-managing, but trams etc shows the same elevated costs. No civil service reform in Scotland nor planning reform either. Scottish nationalists are an extension of Scottish Unionism’s assumptions about their unique superiority. Ireland does the same despite all the rhetoric of being Liberal Europeans against those evil reactionary English (which doesn’t stop them from having racist housing “rent control” or having no universal healthcare, or refusing to have a non-xenophobic census). The post-1923 system is deeply based on assumptions that the UK is a uniquely super-special polity in order to justify keeping the English imprisoned in it to keep the gravy train going. Which is why the most hard left “decolonisation” people are the most brutal British nationalists.
I don’t think London has much to teach NY because TfL has many of the same weaknesses and its good bits i.e. operations cost discipline and high frequency signalling etc are developments from legacy ideas. They wouldn’t know construction costs are a problem and they barely have concept of interlining discipline (yes northern and Bakerloo but that’s it).
I attending an event where the head of Transport for greater Manchester who is ex-TfL was present. I tried to get him to read Alon’s stuff, but he rebuffed me quite aggressively. Hence Manchester keeps demanding expensive expansion of Piccadilly and Victoria rather than focusing on through-running and cheaper shoulder stations in the brownfield sites east and west of central Manchester. Its not just privatisation of the state, its also the deep conservatism of the craft-based Anglo-saxon worker culture especially in the context of soft-budget constraints.
Pointedly all the push to reform British transit systems by learning from the abroad comes non-industry people e.g. Centre for Cities or Britain Remade. The industry is filled with bad actors. Which is also true of urban planning the UK-Ireland in general, where all the reform energy is coming from economists and industry not within planning circles who can’t admit we have been a net negative on human society.
If the head of Manchester Metrolink isn’t flexible enough to handle Alon’s research then he’s putting Andy Burnham’s Bee Network at risk because if Metrolink aren’t flexible then that won’t be achievable.
And fundamentally it is very clear the treasury has seen Alon’s research, so ultimately the industry figures will have to move on and work out how to reduce costs.
Really? Who in the Treasury? I haven’t seen that!
TfL people are understandably defensive, given so much of what gets thrown at them on costs is lies. And the construction cost disease is not just them, its more the Civil Service etc fault. But that’s no excuse, especially given anyone who rides the Jubilee can see that more recent stations are…more expansive than their predecessors.
Professional Northerners being flexible, knowledgeable and competent rather than prideful, nostalgia-poisoned and utterly unwilling to ever take responsibility fiscal and moral is a big ask. Its goes against their proud history of fucking themselves over and then blaming southern England for not being as bad as them and not giving them a large enough allowance to waste.
Seriously they could have set up Local Development orders to build a station with back-up terminal/passing platforms here among the towers but instead they wanted “affordable housing” and not to bother thinking about land prices and sensible operating patterns.
https://www.google.com/maps/@53.4787787,-2.2666469,367m/data=!3m1!1e3!5m1!1e2?entry=ttu
According to the Guardian the UK government figured out on their own that the costs were ridiculous for HS2 – https://www.theguardian.com/uk-news/2023/sep/25/why-has-hs2-ended-up-being-so-expensive
Plenty of bad decisions in Manchester. https://www.nationaltrust.org.uk/visit/cheshire-greater-manchester/castlefield-viaduct is another one.
Inner City Labour is about Inner City Labour membership feeling good about themselves nothing more. They have no competition, no discipline and no ethics other than moral onanism. Exemplars to idea that you can steal so much more than somebody’s paycheck.
On HS2 recriminations they always talk about political uncertainty as just something imposed on them because the Tories are evil (which they are, but that’s what they were hired to be). They don’t want to change, its still all passing buck. The comparisons are almost certainly thanks to Britain Remade’s crew.
And the Guardian journalist is in “oh that’s worth a sentence not a thoughtful paragraph” mode.
Until people get serious about Civil Service reform either by moving to ending cross-department placement or more professional agencies its all bullshit. The system is rotten, it was made to be rotten because that’s the only way Ukania can live.
Inner City Labour is always vulnerable to the Greens and the Lib Dem’s. The challenge is getting good decisions made outside London and the South East – and strictly that isn’t party-political as northern powerhouse rail is a shitshow and that isn’t a pure urban project.
I do think the approach of getting the individual shadow ministers to come up with policy in their areas should work well – and it is clear improved public transport will be a focus under a Labour government.
Certainly the projects started under New Labour were much more comparable to French costs at the time. Perhaps a little higher, but certainly not massively.
Getting New York to retire off the 10 worst senior managers and replace with 5 from TfL, 3 from Paris and 1 each from Tokyo and Seoul would be a definite improvement on the status quo.
Wait, which projects are you thinking of? The Jubilee extension is a bit older and was twice as expensive per km as M14. Crossrail was already in the late New Labour era budgeted at close to its actual cost. HS2 was the project where the budget kept exploding in the 2010s, sure, but Crossrail’s high cost is older than that.
Crossrail was good value compared to the Paris projects newer than the M14 given its traffic levels and the complexities around central London. HS1 was good value given the need for new city center approaches.
The projected cost per rider for Crossrail and GPE is around the same, on the order of US$25,000/daily rider.
And okay, Crossrail is a high-benefit project, but the cost was very high, and New Labour maintained the privatization of the state that began under and right after Thatcher.
Crossrail was also very complex. It had to weave around a huge number of existing tunnels and interchange with other lines.
Paddington, Bond Street, Tottenham Court Road, Farringdon, Liverpool Street, Stratford and Canary Wharf are all massive stops with the potential for very high passenger numbers.
And also to be fair the Jubilee Line Extension was built as a 150-200m passenger a year railway with suitably large stations so that unlike the Victoria line you don’t have to spend billions retrofitting extra capacity.
We have no idea how well the Paris Metro line 14 would handle such passenger numbers as it only has 80m passengers a year vs 150m on the Jubillee line.
92 million and M14 is shorter, to the point that it has higher ridership/km.
With the Jubilee line the 40% of it that is north of West Hampstead sees much lower usage. The 60% south of there by length probably sees around 145 million passengers a year.
And actually on line 14 in Paris they have had to add extra capacity at Gare de Lyon already because the original station was too small.
Apologies for the multi post. Here is the source – https://www.london.gov.uk/sites/default/files/mqt/Appendix%20P_49.pdf. Adds up to 156m in 2015/16.
Note that Paris métro rideship per line doesn’t count transfers.
In exemple, a Passenger transfering from Line 1 to Line 14 would only be counted in Line 1 numbers.
This gives undervalued rideship yearly data.
Line 14 is planned to have a daily rideship of 1 million when completed.
That seems pretty unlikely when there’s only 3 big stops for entries (and presumably exits) on line 14. Gare de Lyon, Biblioteque and Saint Lazare. These have maybe 40 million entries and exits for the first two and 20-25 million entries and exits at Saint Lazare. Now sure there’s probably a lot of transfers at Chatlet, but still.
And the extensions they are building are into the suburbs so you’d expect less than 10 million riders a stop, likely less than 5 million a stop, unless there’s a massive Canary Wharf style destination – so how the ridership would increase from 100m a year to 400m a year I am not sure.
Data from:
https://data.ratp.fr/explore/dataset/trafic-annuel-entrant-par-station-du-reseau-ferre-2019/table/?
It’s only the number of entries, Paris métro doesn’t have stats for exits.
It’s only entries coming from outside the RATP network.
It means that transfer from RER A and B isn’t included as well as transfer from other metro lines.
That’s why Châtelet has lower rideship number than other less important stations.
Then you understand that Line 14 rideship isn’t 100 million but much more.
It’s just that most of the passengers aren’t counted in the stats (transfering passengers).
The yearly rideship per lines is just the number of entries in Paris network divided by lines but not the actual ridership of those lines.
The tube entry/exit figures also exclude transfers between each line. So that isn’t Paris only.
I can believe the line 14 figures exclude double transfers – i.e people who do RER A, then line 14 then line 5 or something – and potentially given the RER A/B are pretty extensive and miss stops in the city centre there are quite a few of those.
It’s true that London Underground doesn’t count transfers on station data (very few network does) but it includes exits what Paris doesn’t.
My point on transfer was that you can’t have a decent estimation of line ridership while using Paris métro station stats.
Especially for Line 14 that relies on transfer.
In the link you’ve posted, it’s written.
“Les « entrants directs » sont exclusivement les voyageurs provenant de la voie publique ou du réseau SNCF entrant sur le réseau de transport RATP en validant un titre de transport valide.
Les voyageurs en correspondance (y compris correspondances métro/RER) sur le réseau RATP ne sont pas comptabilisés.”
““Direct entrants” are exclusively travelers coming from public roads or the SNCF network entering the RATP transport network by validating a valid transport ticket.
Passengers connecting (including metro/RER connections) on the RATP network are not counted.”
The planned 1 million daily ridership means 1 million passengers during weekday outside August and July.
It would not be a yearly ridership of 400 million.
Osaka has way more ridership than the nyc subway.
Despite their best efforts Osaka Metro’s network is large and probably the worst designed of any in Japan thanks to their “Monroe doctrine” which lasted till Ishin came to power*. Osaka is saved by an insanely good legacy above ground network of JR West and the privates plus strong TOD. The underconstruction Naniwa-Suji line should have been built decades before most of the subway lines.
It has 1 god tier line the Midosuji which is one it inherited from its private days and the extensions where done not by Osaka metro but by the prefecture. It handles nearly 50% of the metro ridership.
All the other ones have at least 1 major flaw.
Yotsubashi basically just parallels the Midosuji without through-running or as good walkways/station connection.
Tanimachi should have through-running Kintetsu Minami-Osaka narrow guage network, but they had it standard guage plus 3rd rail because.
Chuo and Sakaisuji each have through-running but don’t properly connect with hub stations. Chuo line should have had through-running from the Ikoma line but they 3rd railed it. Sakaisuji line should have linked Hankyu’s network to Kintetsu’s network via Ueno-honmachi.
Sennichimae and Nagahori Tsurumi-ryokuchi Line have far too few connections to the surface private company networks let alone absent through-running (which was especially stupid given the Sennichimae parallels Kintetsu’s two busiest lines, but again wrong electrification).
The Imazatosuji line is the 3rd stupidest line in Japan after Kobe’s Kaigan line and Fukuoka’s Hakozaki line. Its incomplete orbital which duplicates a lot of the function of the two parallel JR lines. It should be a tram not a smol subway.
*Or maybe Nagoya, but Nagoya metro’s problems are more about Meitetsu and JR Tokai being terrible.
You do know that the MTA is a New York STATE Agency not a New York City agency????? Its paymasters are in Albany, not New York and it’s beholden to the Senate in Albany. New York State is a huge State and has other cities in need of money and transit solutions not everyone in the New York State Assembly is a Democrat and even those who are don’t necessarily appreciate New York City taking all the spotlight. So its always a balancing act, the head of the MTA reports to the New York State Governor, not the NY City mayor…..
As of the last census, New York State is now smaller than Florida, and the gap is growing.
California – 39,237,836.
Texas – 29,527,941.
Florida – 21,781,128.
New York – 19,835,913.
Perhaps Washington DC should increase funding of rail service south from DC.
If the US cannot run great service on the DC to Boston route, then we have no business even looking at routes elsewhere. Nobody likes to hear this, but the priority needs to be get great service on DC to Boston as that route is blindingly obviously great and should get much better passenger numbers if it had great high speed service.
Florida does have high population, but it is a long distance from anyplace else. Maybe in Florida trains could work out (Brightline!), but this isn’t interstate travel.
If you trust my gravity model-based formula, then Florida is marginal. It’s viable at HSR costs, but only as the culmination of a network coming from Atlanta.
To put my fingers on the scale a bit, Floridian cities are very decentralized, even by American standards (in Edgeless Cities, Miami turns out to be the most decentralized of the large US metro areas). The two saving graces are that Florida has an atypically large tourism market, driving Orlando’s air traffic volume way above what you’d expect for its size, and that Miami is at metropolitan scale linear, so more people are within driving distance of a station. But the first saving grace doesn’t really work for Orlando-Miami – the train has to connect to somewhere outside of Florida. The second saving grace is useful intra-Florida, but it’s even more useful for interstate trains, whose competition is planes, not cars.
Meh. Orlando – Atlanta is roughly the same distance as Boston – Washington D.C. Or Chicago – Toronto. Jacksonville with a Combined Statistical Area of 1.8 million is along the way. Meh. Atlanta-Miami is …. too far to compete well with flying. Meh.
The median center of population for the U.S. is in southwest Indiana. Right across the border is half of Canada. Metro Chicago straddles the border between east and west. Washington D.C. straddles the border between north and south. Atlanta-Miami makes more sense than Denver-Salt Lake City but there are a lot more places in the northeast quadrant.
“If you trust my gravity model-based formula”
I’m sorry, but I don’t.
“only as the culmination of a network coming from Atlanta.”
This is why, your model absurdly favors tails making very long distance connections to very-large areas, instead of reasonable distance connections between just-large areas. Its how you end up with things like a connection to Memphis because of phantom ridership to Chicago/NY 1000+km away, but somehow people won’t travel 350km from Oklahoma City to Dallas. Adirondacker is completely correct about FL-Atlanta being too far to be a major HSR route.
@henrymiller
“Maybe in Florida trains could work out (Brightline!), but this isn’t interstate travel.
@Alon
“The second saving grace is useful intra-Florida”
It’s important to remember that the US is huge, and in many cases intra-state US travel is completely comparable to intercity and international travel elsewhere. Orlando to Miami via the coast is basically the same distance as Paris-Lyon, yet no one said that LGV Sud-Est was too short to make a good HSR line. Yes, Paris is larger than Miami, but Orlando is larger than Lyon and Tampa is larger than Orlando. The linearity of Floridian population is a huge bonus to rail travel (and not just within the Miami MSA, Palm Bay-Melbourne has greater population than Saone-et-Loire, a department that gets two stops on LGV Sud-Est). Tampa-Orlando-Miami is just 542km across dead flat terrain. Even going the wrong way moderate speed HSR travel from Tampa-Miami would easily beat car travel by over an hour and a half even with cars cutting the corner with a shorter route.
@Onux decentralized gets you though in Florida. Sure HSR is faster once you get on the train, but on both ends you potentially have a long way to go after you get off the train. Worse there isn’t great transit so you are looking at getting a car to get around. Many people will be time and money ahead just driving the whole way, and even those who are not will still find driving worth doing just because then you have your car with you.
Remember a car is one price for as many people as you can fit. For families HSR quickly becomes unaffordable because you pay for each person separately. A car often means you can bring more “stuff” because you don’t have to haul your bags around. These are very compelling things in favor of the car. So if you have to have a car once you get there, then driving the whole trip becomes compelling. My point is Florida (unlike Boston, NYC and DC) doesn’t cover many people with great local transit system, so odds are much higher someone there will be looking at a car anyway. (of course Florida does get a lot of tourists flying in, but they wouldn’t use HSR in general)
In Europe there are lots of group and family discounts for the train making it better value in those cases.
@henrymiller74
I disagree with Florida being decentralized. The Florida coast is an extremely linear urban area, no where wider (at its absolute maximum extent) than 30km before you hit the Everglades, mostly 15-20km wide, and the CBDs and most of the people are within a belt about 5-10km wide, which has two intact, in-use rail routes running along it (the Brightline route right through each downtown, and the Amtrack route a few km farther inland). Compare to Philadelphia, which is 45km “wide” from Voorhies to King of Prussia. The central part of Miami (downtown to the airport, including little Havana, etc.) is about 7km wide; this is the same distance as South Station to Dorchester or Tufts – is Boston “decentralized?” The decentralized/sprawl issue is definitely a concern with a lot of US urban areas, but unlike Atlanta or DFW, which are blobs extending 50-100km in all directions, the Florida Coast has the geography of a Swiss valley. This is even more pronounced north of W. Palm Beach and south of Miami, where the urban zones are only 5-10km wide.
You are correct that for many trips cars will be better, depending on origin, destination, and use at the end. I have written before criticizing those who say cars are a “failed” transportation solution – for most people cars are the best transportation choice. But the same is true along the NEC, it is faster to drive from Danbury to Trenton, even before accounting for the fact that a train leaves Danbury only every hour at best. A family of five needs five tickets to ride Acela, but only pays one toll crossing a bridge. So I-95 is crowded, but Amtrak is busy too, because not all trips are suburb to suburb (in fact it is the opposite, trips and destinations are preferentially to CBDs, for work, tourism, etc.) and not all trips are by families.
I am not suggesting that rail will replace cars entirely in south Florida (or anywhere else; France is considered to have an excellent HSR network, plus great intra-urban rail (RER, metro, trams) yet 85% or trips are by car, only 9.5% by rail) yet to look at the perfect linear corridor right along the coast and not think that rail can have an impact is crazy.
The other advantage of a rail project Florida is that on a vote counting basis any rail plan needs to benefit a large number of states. It is going to struggle to pass the senate if it only benefits the north east.
I mean to be honest intermodal freight can manage an average speed of 75mph. So that is compatible with passenger service with a top speed of 110-125mph – and that isn’t shit.
It seems railfans have difficulty with the concepts represented in maps. It seems they think all the dots are equidistant and equipotent. Fantasies of Chicago to Denver by high speed rail because Chicago-St Louis-Kansas City-Denver is only four dots. It’s too far and there aren’t a whole lot of people out there in the wheat fields. Atlanta to Orlando is 439 road miles or 706 km. There’s a lot more people between Atlanta and Orlando than there are between Kansas City and Denver but there aren’t a lot of them. It’s even longer if it swoops east to connect Jacksonville and all the population clinging to the Atlantic coast. Keeping population in mind when having railfan fantasies is good. Denver is 600-ish miles from Kansas City and there aren’t a whole lot of people out there. Everybody in metro Binghamton makes a high speed rail round trip every year it works out to a bus to Syracuse once an hour 16 hours of the day.
Families are a fraction of the travel market. Airlines don’t let you check your steamer truck or automobile. The families from the Midwest and Northeast who want to do the theme parks in Orlando then visit grandma and grandpa on the coast might find the train useful. They don’t need a car because one or both of the grandparents still drives and can pick them up at the station just like they pick them up at the airport when they aren’t doing Disney. The rest of the travel market isn’t hauling around brats and diaper bags.
One interesting, special Amtrak offering is AutoTrain (from Virginia to Florida, ~850 miles). Yes, you can pack your car and take it with you for ~$100. I think that extending the northern terminal to the New York City area would be popular/useful.
Wikipedia says “Lorton was selected as site of the northern terminal because the 20-foot-2-inch (6.15 m) high autoracks were too tall to pass through the First Street Tunnel into Washington, D.C.[15]”
Besides this tunnel, the closest Potomac mainline rail crossing is way up in Harpers Ferry.
So it doesn’t seem that going further north is possible.
Lots and lots of people drive all the way from the Midwest and Northeast too. That doesn’t change that millions of people use airplanes to get from the Midwest and Northeast to get to Florida.
Wikipedia says Amtrak had 279,000 Auto Train passengers in 2022. It also says that Sanford International had 340,000 airline passengers from the Northeast and Midwest. And that Orlando had 339,000 from Bradley outside of Hartford Connecticut and MacArthur in Islip, on Long Island in New York had 175,590 and Columbus Ohio had 259,000. Lots and lots of people leave their cars at an airport in the Northeast or Midwest. Or at home.
No it’s not a waste of money. The space is for all the invisible to assist in walking, carrying. Not to mention the bathrooms need. Europe has bathrooms at each station. Lack of parks and bathrooms.
People don’t want to leave the home.
People working are into fraud to justify their work.
Janno Lieber did amazing work at the World Trade Center, don’t sell him short. He’s no MTA insider. But the bureaucracy is entrenched and politically powerful in Albany. Lieber needs our support, a replacement would be far worse.
Have you ever been to 125th & Lexington
Very dangerous
Yes, a bunch of times. I didn’t notice any danger.
125/Lex definitely _looks_ sketchy as hell — it’s a poor neighborhood and there are some nearby homeless shelters and service orgs — but nowhere in the entire island of Manhattan will you find anything that actually qualifies as a dangerous neighborhood. My daughter went to middle school a stone’s throw from there. It’s fine.
The information this post is based on is false. It makes the basic mistake of adopting a fundamentally inaccurate assertion from the New York Post without any effort to verify it. The Post article is based on an obsolete 20-year-old schematic drawing (from the 2004 Second Avenue Subway EIS) that has long since been substantially modified to reflect modern technology, cost efficiency and reduced size.
Contrary to the obsolete EIS diagram, Phase 2 stations have been substantially reduced from the 2004 EIS assumptions and from the Phase 1 station sizes. For instance, the 125 Street “station box” is 900′, compared to 1,200′ for the 96th Street Station – and this 900′ length includes roughly 650′ to accommodate the platform plus only those basic engineering functions required to safely run a subway line such as electrical distribution, signal maintenance, and dispatch that must be located adjacent to the train line. Based on the experience of Phase 1, other back-of-house functions have been value engineered out entirely or located in lower-cost and above-grade ancillary areas where necessary. Similar measures have been taken elsewhere – for instance, since they carry significantly fewer passengers than 125 Street and do not serve as the terminal of the line, the mezzanine for the 106 St station has been reduced by 50% and the 116 St mezzanine has been eliminated. The 125 St mezzanine is only large enough to meet Fire Code given the number of expected riders.
The MTA is committed to attacking costs on all projects in a realistic, safe, and responsible manner, and on the next phase of SAS that includes dramatic reductions of station size and price that have resulted in $1 billion of projected savings to-date with more to come. Jumping to conclusions based on a 20-year-old document does the cause no favors – nor does drawing superficial price comparisons to projects that carry far fewer than the 300,000 daily riders of Phase 1 and Phase 2 combined.
Just last month New York celebrated the Federal government’s agreement to fund the Second Avenue extension to 125 St because it’s a great project, that will dramatically improve the lives of New Yorkers, and we will never turn our back on the residents of East Harlem who depend on mass transit.
Re fire code: what I have to keep explaining to advocates is that NFPA 130 is also used in China, Turkey, and (with slight modifications) Spain. It’s not used in the UK, but the UK has its own problem with massive stations; the only difference I’ve seen between the US and UK code has to do with interpretation of cross-passages anyway. The rule that motivates full-length mezzanines – that the station have enough capacity to evacuate all passengers to a place of safety within four minutes under emergency conditions – does not actually require full-length mezzanines (again, they don’t have them in Spain and the rules there are the same).
Re “far fewer than the 300,000 daily riders”: Citybanan carried 408,000 in 2019 (and was projected to grow further); Odenplan had 42,700 weekday boardings, Stockholm C 65,400, cf. pp. 52-53 of the link in the post. Odenplan is a 250 m long dig for 214 m long trains, and puts the staff and mechanical spaces above the platform, where the MTA would put a full-length mezzanine. M14 carried 92 million passengers in 2019, which translates to around 320,000/weekday (France has a low annual-to-weekday passenger ratio, because people get six weeks of paid vacation and go elsewhere in the summer and the weekday counts are taken away from vacation season); the busiest M14-only station, Bibliothèque François Mitterand, had 19,754,334 boardings, so around 70,000/weekday.
And re 900′: are you counting the entire dig? Because if you do, 96th Street is a lot more than 1,200′, more like 1,600′. And the norm pretty much anywhere that doesn’t speak English is that the overage is single-digit percent (I was told 3-5% on Grand Paris Express) or very occasionally the teens (Odenplan is 17%).
You shouldn’t use Paris métro annual rideship per lines because those don’t count transfering passengers.
A Passenger using Line 1 and then Line 14 will be be only counted in Line 1 yearly rideship data.
Paris Line 14 current daily rideship is almost 750,000 and it’s projected to get above 1 million when extension to Orly and Saint-Pleyel will be completed.
https://www.ratp.fr/une-ligne-une-histoire-ligne-14-metro
On the subject of New York passenger rail, do you have a response to this criticism of the Effective Transit Alliance’s Modernizing New York Commuter Rail report? http://www.stevedunham.50megs.com/ModernizingNYCommuterRailComments.htm
Yes, many; Dunham is exactly what is wrong with American commuter rail advocacy. He’s assuming that current service is optimized; it isn’t, and when we point that out with references, he either misreads the references or doesn’t read them at all.
1. The Eisele papers, i.e. the “sources from 1968 and 1978,” describe service today too, with the zonal express trains of peak service. The entire point is that service today is based on how suburban white flighters traveled 50-60 years ago.
2. The line that headway should be no more than half the passenger trip time is based on a lot of literature on the elasticity of travel demand with respect to both. No, we’re not calling for six-minute service to destinations 20-30 minutes away; we literally have a table of proposed frequencies, and the outer-urban ones are a train every 10 minutes.
3. Hourly service on LIRR, Metro-North, and NJT lines is exactly the example of the indifference of commuter rail planners to frequency. Hourly is not good, and neither is half-hourly to Stamford. It only looks good if you only benchmark yourself to other North American failures. More relevantly, the peak-to-base ratio in New York is about 4, cf. 2 in Tokyo, 1.5-2 in Paris depending on the line, and 1-1.3 in Berlin depending on the line. Berlin has more ridership on the S-Bahn than all New York-area commuter rail combined, and almost as much as all American commuter rail combined.
4. Commuter trains are cheaper to run than buses. That’s why competent (i.e. non-North American) commuter rail agencies charge mode-neutral fares; that the S-Bahn is more comfortable than a bus is immaterial, because mode-dependent fares as in New York would incentivize passengers to take worse, higher-cost-to-provide service.
5. The lines we’re proposing to supercharge frequency on don’t have grade crossings.
6. Commuter trains have more spare capacity than the subway (average crowding on the peak LIRR was less than full seated capacity even pre-corona), so shifting riders from the E and F trains to the LIRR is on net good, even if it annoys Long Island white flighters.
7. Around half of the through-commuters from Jersey live within reasonable range of a commuter rail station that feeds into Manhattan (counting Raritan Valley); the other half live near PATH or the Erie lines. But this is with today’s nearly-unusable service; better service does change how people travel.
8. Clockface timetables are essential for planning timed connections, timed overtakes, timed meets on single track, crew scheduling, and turnaround times – and the line that they lead to long turnarounds rings hollow when American commuter trains take 20 minutes to turn around where German regional trains do so in eight minutes (and regional trains here are still worse than in Switzerland).
Maybe the German trains would be more reliable if they had longer to turn around 😜.
And when we discussed the Leman express in Geneva, yes at one end there is an 8 minute turnaround (and a shit-tonne of padding on the approaches making it more like a 20-25 minute turnaround without it), but the turnaround at the other end is 40 minutes.
The padding on American commuter rail is somehow worse than on the German trains. Here it’s 20-30%, but the LIRR Main Line is around 32%, and I think Metro-North is similar. It’s somewhat better on lines that don’t have zonal expresses, like the Babylon Branch.
4. Why should fares be mode neutral then? If buses are more expensive to run, then charge more for buses. Some people might pay more to walk to a bus stop, but some people might choose to bike to the train station, further improving the cost effectiveness of rail service.
Because buses and trains need to complement each other rather than compete, and this means unified tickets.
I don’t think pricing tickets based on operational costs means they compete rather than cooperate though. Guiding rider behavior to best utilize the available options, is rail and bus cooperating.
A bus trip that just parallels a train trip is more expensive than the train trip to provide, so should be more expensive to use. Maybe some riders will still choose it because the bus stops right in front of their house while the nearest train station is a bike ride away, and they are willing to put their money where their mouth is. Or maybe if you make the bus more expensive than the train, you find almost no one chooses it because it’s not providing people enough value to justify the price over the train, and you can just get rid of the route entirely. Part of bus and rail cooperating instead of competing, is that a lot of parallel bus routes probably shouldn’t exist at all, much less be offered at the same price.
However, buses can often take much more direct routes than trains, e.g., a circumferential suburban trip that is direct by bus, but involves a lot of backtracking to a more central area by train. In this case, it might be cheaper to provide capacity for people taking this trip by bus, so people should be encouraged to use the bus through lower fares. Maybe some people really hate the ride quality of buses, and will choose to take the train anyways (I’m definitely guilty of this), but they (we) should express their preference by paying more. Part of bus and rail cooperating is rail encouraging people to take the bus when it is the better choice.
The preference of a shorter trip to a bus stop vs a train station, or the preference of the ride quality of a train vs a bus, is just a preference, and if it’s more expensive to cater to that preference why shouldn’t it be more expensive for the preferrer?
I would agree that having completely disconnected bus and train tickets can be punitive since base fares are getting double charged, but especially when you have fares based on fairly large zones, why should someone taking a bus to the train station pay exactly the same as someone biking or walking to the train station?
Broadly speaking:
– Outside the peak time×segment×direction, the cost of an extra passenger is approximately zero.
– Transit policy is especially interested in getting individuals’/households’ business in bulk, i.e. getting people to ditch (one of) their car(s). Offering a monthly pass, thus making non-peak “secondary” (i.e. non-commute) trips free is an entirely reasonable way to do this, especially since these trips are approximately free to provide anyway.
– Just as most people don’t care whether on its way from the powerplant their electricity passed over an HVDC line or a synchronous interconnection, transit presents as “the network” so that people don’t have to care which company pays the driver’s wage. “This monthly pass is only valid for the part of the network implemented with vehicles that run on rails” would be an inane mistake. (Also, line 123 is usually a tram line, but this month there’s some track maintenance, so this month it’s implemented by articulated buses (wearing the same livery as the trams) stopping at the same stops (the tram tracks are mostly off-limits to cars, but are filled in with a concrete surface). Is your pass valid for it?)
– Preference/price segmentation works perfectly fine by travel class. Railway 1st/2nd/3rd, airline 1st/business/economy — crucially, they are within the same vehicle (and same landside station, from 19th c. ticket-separated waiting halls to today’s airport lounges). However, trying to segment by route, within a single ticketing organization — the railway equivalent of “we only sell business-class tickets between these two cities (and have planes with all-biz cabins captive to the route), if you want economy you have to travel to a different destination (and we don’t offer premium seats to there)” — has worked spectacularly poorly. (Even more poorly than for airlines, because trains have many stops with significant seat turnover, while planes basically don’t.)
Alon, you are comparing apples and pomegranates here. This is the 2004 EIS cross-section when it was likely to be an open cut. Now it is a deeper mined station in rock at 125th so I’m pretty sure that the cross-section and footprint will be somewhat smaller. As for mining beneath the Lex Line, East Side Access passed beneath the Lex Line as it swung towards Park Ave. The clearance between the Lex Line and the ESA construction was something like 10 or 12 ft, and ESA passed beneath 7 line on 34th St with limited cover. The Manhattan Schist is an extremely competent rock so mining beneath the Lex Line at 125th should not be a problem.
The comparison that is shown with some phantom European system is meaningless without some form of justification as to what is included in that back-of-house space. So much depends on where you locate the support systems, for example what space do you have above ground to house your substations etc.
What an ignorant article! Asburger’ Syndrome run amok
So help me fix my ignorance here–Why does the whole station have to be mined? The entire block southeast of 125th and Lexington has been a vacant lot (with prior low-rise overbuilding) since Obama left office. Seems like you ought to be able to just do most of the excavation through there, locate the platforms closer to 3rd Ave, knock out the east wall of the top level of the existing 125th St 4/5/6 station to provide a transfer, and be pretty much good except for the actual track.
Obviously there’s a reason it doesn’t work this way, but why is that?
It has to go under the Lex, and once it’s there, might as well have it go a bit farther west to allow for an easier transfer to Metro-North.
Thanks Alon! That makes sense & I can see why you’d want to allow the station to have an exit by the Metro North (though Wikipedia says they’re planning to have a staircase to the south side street level, so it won’t connect directly into the actual station building and you’ll still have to cross 125th St to reach ticketing, or platforms outside commute hours–a weird corner to cut).
But even a block west there’s a similar situation, where you’ve got a vacant-for-a-decade lot covering 60% of the block southwest of Lex & 125, and another covering the nearer 1/3 of the block southeast of the same intersection. Seems like that’s an awful lot of room to allow open digging, and restricting the mined portions to those actually passing under the train lines plus whatever pedestrian tunnel is needed to connect to the mezzanine of the 4/5/6 station at Lex (or the latter could be cut-and-cover under 125th St). (Or, honestly, the other half of that block is mostly low-rise commercial that could probably be bought out for less than paying to mine the whole thing–the derelict bank at 124 E 125th might be landmarked and something would need to be done to replace/protect the nursing facility midblock, but otherwise it’s low-rise commercial, fast food and dollar store and the like.)
Again, I’m sure there’s a reason this isn’t what’s being proposed, I just don’t understand the technology or practices well enough to know what that reason is. Like is there a technical/practical reason that you don’t mix cut-and-cover segments with mined segments? Would there not actually be much savings once the tracks are already so deep? Or is this another case of overspending to avoid anticipated local opposition to cut-and-cover?
I mean to be fair basically we are saying the second avenue subway needs to have 1 station out of 6 with the same complexity as the median station on the Jubilee line extension with the other 5 having the complexity of the simplest tunnelled one – and similar passenger numbers to that station as well.
The only difference that should raise costs on the second avenue subway is that the trains are 180m long vs 120m or so on the Jubilee line.
Park Ave. – where Metro North is, should be Fourth Ave. Lexington Ave is halfway-ish between Third and Park. Park to Lex is shorter distance than the train length. Pick which way you want to measure it to come up with different numbers. West side of Park’s building line to east side of Lex’s building line, center of both avenues,etc. The west end of the platforms could be at Metro North and the east end at the Lexington Ave. lines/subway. That would make sense and this is something that umpteen committees and reports and consultants have spit out.