The MTA’s Bus Redesign Plan

Two weeks ago I wrote about the Brooklyn bus redesign I’m working on with Eric Goldwyn. The MTA, which is aware of our efforts, came up with its own plan. So far details are scant; there is a presentation available online, which talks about goals (“network redesign,” “higher frequency”) but no specifics (“a more gridded network,” “6-minute off-peak frequency on the main routes”).

At least so far, the goals seem solid. The MTA has the following list of improvements:

  • Redesign the network from top to bottom based on customer input, demographic changes, and travel demand analysis. Provide better connectivity and more direct service in every neighborhood
  • Optimize the existing network with community consultation by removing closely-spaced and underutilized stops and making street design changes on select corridors in coordination with NYC DOT
  • Expand off-peak service on strategic routes using a toolbox of service strategies including increased frequency and demand based service adjustments
  • Expand Traffic Signal Priority (TSP) to allow an approaching bus to hold a green light or shorten a red light
  • Seek exclusive busways on priority corridors to give buses full access in major congested areas
  • Identify opportunities for new bus lanes and queue jumps in 2018
  • Advocate for strengthened NYPD enforcement of bus lanes to keep bus stops and travel lanes clear throughout the system
  • Recommend dedicated transit-priority traffic teams to focus enforcement in key areas to ensure buses move quickly through trouble spots
  • Use Bus Lane Enforcement Cameras mounted on buses to automatically identify violations and issue tickets. Advocate for legislation to expand beyond the existing 16 authorized routes
  • Install tap readers to speed up the boarding process so buses spend less time waiting at stops
  • Introduce all-door boarding to allow riders to get on through any door of the bus
  • Explore options for a future cashless system to maximize reductions in boarding time
  • Expand fare enforcement on regular bus service to reduce evasion and restore fare revenue

The main problems only appear toward the end, with the implementation of off-board fare collection and all-door boarding. The insistence on “fare enforcement,” which could mean regular proof of payment (POP) inspections but could also mean worse, such as armed cops (not practiced in New York on SBS but practiced on some other US systems, like BART) or holding the bus during inspection (which New York does practice, unlike Berlin and other German-speaking cities). Overall I’m relatively sanguine about Andy Byford specifically – he’s not American and is not used to American levels of police militarization.

However, another aspect of the POP proposal is troubling: the connection with tap readers. The plan’s full text (which is barely more detailed than what I quote above) mentions that POP should come with the so-called New Fare Payment System, or NFPS, which New York is currently planning to roll out starting in 2019, continuing until 2023. The NFPS is based on worst industry practices cobbled from American and British ideas. Here is my second post ever, discussing the plans for smartcards in New York in 2011. New York ignored (and keeps ignoring) the smartcard implementations in a number of East Asian cities in its zeal to make people treat their credit card as a transit fare card, ensuring the agency can surveil all passengers; perhaps Americans lack the values of freedom and individual privacy of Japan and Singapore.

New York also ignored (and keeps ignoring) the POP implementations in cities with paper tickets, such as most of Central Europe. Smartcards are not required for POP: the German-speaking world has POP with paper tickets, as did Vancouver on SkyTrain and some bus lines even before the Compass Card debacle. In Singapore I saw a ticket inspection on a bus even before EZ-Link; I had a magnetic farecard at the time. Given the enormous waste coming from making passengers line up and pay the driver, it’s imperative to move toward POP as soon as possible, even if it means equipping inspectors with MetroCard readers rather than smartcard readers. MetroCard may only last for five years if the NFPS schedule doesn’t keep slipping, but handheld magnetic card readers are a cheap technology whereas making buses idle while passengers dip cards one at a time is not so cheap.

The zeal to go cash-free is the final troubling aspect of New York’s ideas about fare payment, especially when bundled with the idea that the bank card is the fare card. Not everyone has a local bank account. Tourists don’t (and even cards that are supposed to work abroad don’t always). Low-income city residents don’t, either: 11.7% of New York households have no bank account, and they disproportionately appear to be in poverty, judging by which neighborhoods they are most concentrated in. The MTA has always treated anonymous smartcards as an afterthought, and going cash-free means there is no recourse for the unbanked or even for many tourists.

Nor is cash-free operation even necessary. An on-board farebox is compatible with POP. In this system, riders can board from any door, and the driver will begin moving as soon as all passengers have boarded, even if not all passengers have paid yet. Riders with valid transfers or season passes need not do anything. Riders with a pay-per-ride smartcard and no transfer should tap their card at a validator at any bus door or bus stop (validators are cheap that blanketing the system with them is practically free). The remaining passengers should walk to the farebox and pay there; perhaps some busy stations should get fareboxes, as all SBS stops do in New York today, but if the MTA only expects a few smartcard-free, non-transfer passengers at a stop, then having them pay on board a moving bus should not be a problem.

I’d like to stress that other than the ongoing hiccups of the English-speaking world with fare payment systems (hiccups that it seems to export to Paris), the plan appears good, from what few details the MTA has released. There are plans for increasing the average distance between bus stops, adding more bus lanes, getting serious about signal preemption, raising off-peak frequency, and letting passengers board from all doors. The MTA really is noticing that its bus system is collapsing and really is making serious plans to avert a death spiral.

When are Express Trains Warranted? Part 1: Regional Rail

The Regional Plan Association has a detailed regional rail proposal out. It’s the same one from the Fourth Plan that I’ve criticized here, on Streetsblog, and on Curbed, but with more explanation for how the service should run, with stopping patterns and frequency.

There are some good aspects there, like a section about the importance of electrification and multiple-units, though it stops short of calling for full electrification and replacement of locomotives with EMUs; the focus on off-peak frequency is also welcome. There are also bad ones, like the claim on p. 32 that it’s difficult to impossible to provide through-running using the existing Penn Station tracks used by New Jersey Transit. Foster Nichols told me that there are some difficulties with grades but they should be doable if NJT commits to an all-EMU fleet, and reminded me that the ARC studies judged through-running using these station tracks and new tunnels feasible. What he expressed to me as a difficulty turned into a near-impossibility in the report, in order to justify the $7 billion Penn Station South project.

But I want to focus on one particularly bad aspect of the proposal: the stopping patterns. The RPA is proposing three distinct stopping patterns on pp. 32-45, with three separate brands: Metro, in the city and some inner suburbs; Regional Express (RX), in the suburbs; and Trans-Regional Limited (TRL), providing intercity service to New Haven, Ronkonkoma, Philadelphia, and other major stations outside the built-up area. Even as the plan talks about the importance of making sure suburban trains serve urban stations in order to give them frequent service through overlay, the stopping patterns suggest the opposite.

The proposal involves trains from the suburbs expressing through most city stations (including the infill) even on two-track lines, like the Port Washington Branch. Metro trains would make all current stops plus additional infill to Bayside, and RX trains would only serve Willets Point, Flushing, and Bayside, and then run from Bayside to Port Washington. A similar pattern happens from Jamaica to Valley Stream, resulting in the Babylon, Long Beach, and Far Rockaway Branches all having to share a track pair. Moreover, the RX trains may themselves be divided into local and express trains, for example on the New Haven Line.

This is bad practice. On a two-track line, there’s no real reason to skip a handful of inner stations just to guarantee the outer ones express service. If anything, the need to schedule trains on the same tracks would lead to more fragile timetables, requiring more schedule padding. My analysis from 2.5 years ago found that the LIRR Main Line is padded 32% and the Babylon Branch is padded 19%: that is, the scheduled travel time on the Main Line (up to Ronkonkoma) is 32% more than the travel time imputed from line speed limits and current fleet acceleration performance. Patrick O’Hara, who the RPA study even quotes as a source elsewhere, investigated this issue separately, looking at best-case timetables, and found that some runs are padded 40-50%.

In Switzerland, trains are padded 7%, and I’m told that in Japan, after the Amagasaki accident showcased the safety problems of overly precise schedules, pads are about 5%. Express trains and locals mixed on the same line make it harder to maintain tight enough reliability for low schedule padding; this way, on an all-local line, trip times may match those of express trains on mixed lines, as they do in my analysis above. The best analogy is the RER B going to the north: the express trains are 4 minutes faster than the local trains, skipping 9 stops. The stop penalty on the RER B is higher, closer to 7 minutes over 9 stops, but the shared tracks with local trains (and with the RER D between Gare du Nord and Chatelet-Les Halles) means that there’s a fudge factor in the schedule, so it’s not possible to reliably do better than 4 minutes, and the trains end up visibly crawling on the mainline.

The reader familiar with technical transit advocacy in the Bay Area may interject, what about Caltrain? Clem Tillier has no trouble proposing timetables mixing local trains, express trains, and high-speed rail on the same track pair with timed overtakes, and a 7% pad. So why am I down on this concept in New York? The answer is line complexity. Caltrain is a simple two-track back-and-forth, and HSR is generally more punctual than legacy trains because it runs for long stretches on high-quality dedicated tracks, so it’s unlikely to introduce new variability to the line. In contrast, the RPA plan for regional rail in New York involves extensive branching, so that train schedules depend on trains elsewhere on the line. In this case, introducing more complexity through local/express sharing is likely to require more schedule padding, erasing the speed advantage.

In general, my questions to establish guidelines for where express trains are warranted are,

  • How long is the line, measured in the number of stations? More stations encourage more express trains, because more stations can be skipped. In higher speed zones, stop penalties are higher, but at equal line length measured in km, higher speeds and fewer local stations reduce the benefit of express trains.
  • How frequent are trains? At low frequency, local stations need more frequency, so express runs are less useful. At very high frequency, there may not be capacity for different stopping patterns unless the line has four tracks. On a two-track line, the optimum frequency for a local/express alternation is about 6-12 trains per hour, 3-6 local and 3-6 express, with a single mid-line overtake. Multiple overtakes on a single line are possible, but more fragile, so they are a bad idea except in special circumstances.
  • What is the demand for travel? Express trains work best if there are a few distinguished stations at regular intervals, or else if the line is long and there is strong demand at the far end; if the inner stations are very strong then it’s more important to give them higher local frequency. When performing this analysis, it’s important to make sure station ridership levels reflect genuine demand rather than service. For example, Caltrain express stops have high ridership in large part because of their better service, not nearby density, as shown in Clem Tillier’s analysis. The LIRR Main Line has far more ridership at Mineola and Hicksville than the other stations on the trunk and also far more service, but Patrick explains that this is due to better highway access, so it’s genuine demand and not just a reflection of better service.

Caltrain needs express service because it has about 20 stations between San Francisco and San Jose, depending on the amount of infill and anti-infill desired; a target frequency of 8-10 peak trains per hour; and strong demand on the outer stations, especially for reverse-peak trips. In New York, none of the two-track lines meets the same standard. Some are too short, such as the Port Washington Branch. Others are too busy, such as the Harlem Line, Babylon Branch, and LIRR Main Line. Yet others have too much demand clustering in the inner stations, such as the Erie lines and the North Jersey Coast Line.

On four-track lines, it’s always easier to run express service. This doesn’t mean it should always be run: the upper New Haven Line is a strong candidate for relegating all commuter trains to the local tracks, making all stops, giving the express tracks to intercity trains. The Northeast Corridor Line in New Jersey is a dicey example: past Rahway there are four tracks, but intercity trains could run at very high speeds, making track sharing on the express tracks difficult. My service pattern map has express trains skipping Edison and Metuchen, but it’s just two stations, making it better to just run local beyond Rahway to clear the express tracks for high-speed rail.

It’s tempting to draw proposals involving intense metro-style regional rail service only serving the urban and inner-suburban stations; I’ve had to argue against such plans on some MBTA lines. The problem is that trains from the outer suburbs are still necessary and still going to pass through the inner suburbs, and in most cases they might as well stop at those stations, which need the frequency more than the outer suburbs need the few minutes of speedup.

Buses in Brooklyn: Frequency is Freedom, but 15 Minutes isn’t Frequency

I’ve recently started working part-time on a project for the Marron Institute at NYU about bus restructuring in Brooklyn; at the end of this summer, I expect to release a proposal for service upgrades and a new map. I’m working on this with Marron scholar Eric Goldwyn, who is funded by the TWU, which is worried that ridership collapse may lead to service cuts and job losses, but I’m funded directly by Marron and not by the union.

Some of what’s likely to appear in the final report should be familiar to regular readers of this blog, or of Human Transit, or of the work TransitCenter has been doing. As I wrote in Curbed earlier this year, bus operating costs in New York are unusually high because the buses are slow, as the main operating costs of buses scale with service-hours and not service-km. Thus, it’s important to speed up the buses, which allows either providing higher frequency at the same cost or the same frequency at lower cost. A bus speedup should include systemwide off-board fare collection and all-door boarding (common in the German-speaking world but also in San Francisco), wider stop spacing, dedicated lanes wherever there is room, and signal priority at intersections; the TWU is an enthusiastic proponent of off-board fare collection, for reasons of driver safety rather than bus speed.

While bus speedups are critical, their impact is not as Earth-shattering as it might appear on paper. New York’s SBS routes have all of the above features except signal priority, and do save considerable time, but are still slow city buses at the end of the day. Brooklyn has two SBS routes: the B44 on Nostrand, averaging 15 km/h (local B44: 11.3), and the B46 on Utica, averaging 13.7 (local B46 on the shared stretch: 10.8). Their speed premiums over the local routes are toward the high end citywide, but are still 30%, not the 200% speed premium the subway enjoys. Moreover, the speed premium over non-SBS limited routes is 15-20%; put another way, between a third and a half of the speed premium comes purely from skipping some stops.

I mentioned in my last post that I met Carlos Daganzo at Berkeley. Daganzo was responsible for the Barcelona bus redesign, Nova Xarxa; you can read some details on Human Transit and follow links to the papers from there. The guiding principles, based on my conversation with Daganzo and on reading his papers on the subject, are,

  • Barcelona has high, relatively uniform density of people and jobs, so there’s no need for buses to hit one CBD. Brooklyn has about the same average residential density as Barcelona, but has a prominent CBD at one corner, but as this CBD is amply served by the subway, it’s fine for buses to form a mesh within the subway’s gaps.
  • Nova Xarxa involved widening the stop spacing in Barcelona from less than 200 meters to three stops per km, or a stop every 300-350 meters; Daganzo recommends even wider stop spacing.
  • While Barcelona’s street network is strictly gridded, the buses don’t run straight along the grid, but rather detour to serve key destinations such as metro stops. This is an important consideration for Brooklyn, where there are several distinct grids, and where subway stops don’t always serve the same cross street, unlike in Manhattan, where crosstown routes on most two-way streets are assured to intersect every north-south subway line.
  • The percentage of transfers skyrocketed after the network was implemented, standing at 26% at the end of 2015, with the model predicting eventual growth to 44%, up from 11% before the redesign.
  • The network was simplified to have 28 trunk routes, the least frequent running every 8 minutes off-peak.

The high off-peak frequency in Barcelona is a notable departure from Jarrett Walker’s American network redesigns; the evidence in Houston appears mixed – ridership is about flat, compared with declines elsewhere in the country – but the percentage of transfers does not seem to have risen. Jarrett says in his book that having a bus come every 10 minutes means “almost show-up-and-go frequency” with no need to look at schedules, but his work in Houston and more recently in San Jose involves routes running every 15 minutes.

Moreover, unreliable traffic in these car-dominated cities, in which giving buses dedicated lanes is politically too difficult, means that the buses can’t reliably run on a schedule, so the buses do not run on a clockface schedule, instead aiming to maintain relatively even headways. (In contrast, in Vancouver, a less congested street network, with priority for all traffic on the east-west main streets on the West Side, ensures that the buses on Broadway and 4th Avenue do run on a fixed schedule, and the 4th Avenue buses have a 12-minute takt that I still remember four years after having left the city.)

I’ve talked about the importance of radial networks in my posts about scale-variant transit. I specifically mentioned the problem with the 15-minute standard as too loose; given a choice between an untimed 15-minute network and a timed 30-minute network, the latter may well be more flexible. However, if the buses come every 5 minutes, the situation changes profoundly. Daganzo’s ridership models have no transfer penalty or waiting penalty, since the buses come so frequently. The models the MTA uses in New York have a linear penalty, with passengers perceiving waiting or transferring time on the subway as equivalent to 1.75 times in-motion time; bus waiting is likely to be worse, since bus stops are exposed to the elements, but if the average wait time is 2.5 minutes then even with a hefty penalty it’s secondary to in-vehicle travel time (about 18 minutes on the average unlinked bus trip in New York).

Unfortunately, that high frequently does not exist on even a single bus line in Brooklyn. Here is a table I created from NYCT ridership figures and timetables, listing peak, reverse-peak, and midday frequencies. Five routes have better than 10-minute midday frequency: the B12, the B6 and B35 limited buses, and the B44 and B46 SBSes, running every 7, 7.5, 8, 8, and 6 minutes respectively. In addition, the B41 limited and B103 have a bus every 9 and 7.5 minutes respectively on their trunks, but the B41 branches on its outer end with 18-minute frequencies per branch and the B103 short-turns half the buses. Another 14 routes run every 10 minutes off-peak, counting locals and limiteds separately.

The problem comes from the split into local and limited runs on the busiest buses. The mixture of stopping patterns makes it impossible to have even headways; at the limited stops, the expected headway in the worst case is that of the more frequent of the two routes, often about 10 minutes. The average ridership-weighted speed of Brooklyn buses is 10.75 km/h. An able-bodied passenger walking at 6 km/h with a 10-minute head start over a bus can walk 2.25 km before being overtaken, which can easily grow to 3 km taking into account walking time to and from bus stops. To prevent such situations, it’s important to run buses much more frequently than every 10 minutes, with consistent stopping patterns.

This does not mean that NYCT should stop running limited buses. On the contrary: it should stop running locals. The SBS stop spacing, every 800 meters on the B44 and B46, is too wide, missing some crossing buses such as the B100 (see map). However, the spacing on the B35 limited is every 400 meters, enough to hit crossing buses even when they run on one-way pairs on widely-spaced avenues. The question of how much time is saved by skipping a stop is difficult – not only do different Brooklyn buses give different answers, all lower than in Manhattan, but also the B35 gives different answers in different directions. A time cost of 30 seconds per stop appears like a good placeholder, but is at the higher end for Brooklyn.

The question of how many stops to add to SBS on the B44 and B46 has several potential answers, at the tight end going down to 400 meters between stops. At 400 meters between stops, the B44 would average 13 km/h and the B46 12 km/h. At the wide end, the B44 and B46 would gain stops at major intersections: on the B44 this means Avenue Z, R, J, Beverly, Eastern Parkway, Dean, Halsey, and Myrtle, for an average interstation of 570 meters and an average speed of 14 km/h, and on the B46 this means Avenue U, Fillmore, St. John’s, and Dean, for an average interstation of 610 meters an average speed of 13 km/h. Consolidating all buses into the same stopping pattern permits about a bus every 2.5 minutes peak and every 4 minutes off-peak on both routes.

On the other routes, consolidating local and limited routes required tradeoffs and cannibalizing some peak frequency to serve the off-peak. While it may seem dangerous to limit peak capacity, there are two big banks that can be used to boost off-peak frequency: time savings from faster trips, and greater regularity from consolidating stop patterns. The B82 is an extremely peaky route, running 7 limited and 10 local buses at the peak and just 6 buses (all local) for a four-hour midday period; but there is a prolonged afternoon shoulder starting shortly after noon with another 6 limited buses. Some peak buses have to be more crowded than others just because of schedule irregularity coming from having two distinct stop patterns. Consolidating to about 15 buses per hour peak and 10 off-peak, cannibalizing some frequency from the peak and some from the shoulders, should be about neutral on service-hours without any additional increase in speed.

The sixth post on this blog, in 2011, linked to frequent maps of Brooklyn, Manhattan, and Bronx buses, using a 10-minute standard. There has been some movement in the top buses since 2011 – for one, the B41 route, once in the top 10 citywide, has crashed and is now 16th – but not so much that the old map is obsolete. A good place to start would be to get the top routes from a 10-minute standard to a 6-minute standard or better, using speed increases, rationalization of the edges of the network, and cannibalization of weak or subway-duplicating buses to boost frequency.

I Gave a Pair of Talks in the Bay Area About Regional Rail

Thanks to the invitation of Adina Levin’s Friends of Caltrain, I came to the Bay Area for a few days, giving two talks about regional rail best industry practices in front of Friends of Caltrain, Seamless Bay Area, and SF Transit Riders. My schedule was packed, including a meeting with a world-leading expert in transportation systems in which I learned about the Barcelona bus service changes; a Q&A about both general and Bay Area-specific issues for Bay City Beacon subscribers; and several other meetings.

My two talks were in Mountain View and San Francisco, and used the same slides, with minor corrections. Here are the slides I used in San Francisco; I consolidated the pauses so that each page is a slide rather than a line in a slide. This was also covered in Streetsblog, which gives more background and gives some quotes from what I said during the presentation, not printed in the slides.

Unlike in my NYU presentation last year, I did not include a proposed map of service improvements. The reason is that in the Bay Area, there are more questions than answers about which service should use which piece of infrastructure. This makes fantasy maps dangerous, as they tend to fix people onto one particular service pattern, which may prove suboptimal based on decisions made elsewhere in the system. This is not a huge problem in New York or Boston, where the alignments naturally follow where the stub-end commuter lines are, with only a few questions; but in the Bay Area, the situation is more delicate, because big questions like “can/should Caltrain get a trans-Bay connection from San Francisco to Oakland and the East Bay?” can go either way.

The role of redevelopment in the area is especially important. Unlike New York, Paris, or other big transit cities, San Francisco does not have much density outside the city proper, Oakland, or Berkeley. Moreover, there is extensive job sprawl in Silicon Valley, which contributes to a last-mile problem for public transit; usually first-mile access is a bigger problem and high-end jobs tend to cluster near train stations. But conversely, the high incomes in the Bay Area and the growth of the tech industry mean that everywhere TOD is permitted in the core and the suburbs, it will be built. This impacts decisions about the total size of transit investment.

At present-day development patterns, San Francisco proper really doesn’t need more rail construction except Geary and the Downtown Extension, depending on construction costs. But general upzoning makes Geary worth it even at $1 billion per km and opens up the possibilities of four-tracking Caltrain within the city (which means expanding some tunnels), giving the N-Judah dedicated tracks and a dedicated tunnel under Mission, and extending the Central Subway to the north and northwest. The land use in the Richmond and Sunset Districts today supports a fair amount of transit, but not new subways if there’s no cost control.

Joe Lhota Admits Defrauding the Public About Costs

The Wall Street Journal is reporting a bombshell story about New York’s subway station renovation program. The MTA had a budget of $936 million for renovating 32 subway stations, but nearly the entire budget is exhausted after the MTA has spent it on only 19 stations. These renovations do not include accessibility, which New York is lagging on. I’m interviewing people in the disability rights community about New York’s problems in this area, but the smoking gun about Lhota is not that issue, on which he is no worse than anyone else. Rather, it’s that Lhota hid the fact of the cost overrun from the MTA board. Per the Journal:

On Monday, Carl Weisbrod, a commissioner who represents New York City, said the program was “ill-conceived,” and that he is glad it has come to an end.

“I don’t know when the MTA management realized that the program had run out of money but it would’ve been helpful to have informed the board when this matter was under discussion,” he said.

Mr. Lhota said he was aware of the increased costs last year, but he chose not to mention it until now. “I didn’t think it was relevant to the debate,” he said.

An alternative way to phrase Lhota’s own words is that he is concealing critical information from the public relevant to public spending priorities. In other words, he is defrauding the public when it comes to costs. Previously he had been merely making excuses for high construction costs (e.g. saying New York, founded in the 17th century, is old, and thus naturally has higher costs than cities founded in the Middle Ages or even in Antiquity). But now it turns out that he’s not only trying to deflect criticism, but is actively putting obstacles in front of board members, journalists, and ordinary citizens who want to discuss MTA capital expansion.

Absent democratic mechanisms for oversight of the state, the state will not engage in cost-effective projects. We know this, because the part of public policy most insulated from public criticism, the military and security in general, is the most bloated. The US is wasting a trillion and a half dollars on the F-35, and allies like Israel are wasting money buying this jet from the US military industry. It’s hard to question the costs when overconfident military commanders say “this is necessary for national security.” The intelligence community is even worse, with self-serving slogans like “our successes are private and our failures are public.”

Evidently, facing criticism over costs, domestic agencies portray their projects as necessary rather than useful, hence the weak claims that Gateway is required to avoid shutting down rail service across the Hudson. My specific criticism of the argument that Gateway is required is that the study recommending long-term shutdowns of the existing tunnels did not even attempt to provide a comparative cost of maintaining the tunnels on nights and weekends as is done today. An informed public can more easily demand an end to bad investments, and specific interest groups can highlight how they are harmed by bad spending: the Journal article mentions disability rights advocates demanding that the MTA instead spend money on putting elevators at stations to make them accessible to people in wheelchairs.

The station renovations are especially at risk of being canceled if an informed public finds their costs offensive. The benefits include better maintenance standards, but those are almost self-evidently useful but not necessary. Activists can complain about costs or demand that the money be spent elsewhere.

In Astoria, activists complain that the MTA is renovating stations at a cost of $40 million per station without even installing elevators for accessibility. In London, the cost of the Step-Free Access program is £200 million for 13 stations, or about $20 million per station, and in Paris, where only Metro Line 14 and the RER A and B are accessible, disability rights activists estimate the cost of making the remaining 300 stations accessible at €4-6 billion. This is profoundly different from the situation with tunneling costs, where London has a large premium over Paris and New York has a large premium over London. It is likely that New York can install elevators at the same cost of its top two European peers if it puts its mind to it.

However, such investments are not possible under the current leadership. If a hack like Lhota stays in charge of the MTA, there is not going to be transparency about contracting and about costs, which means that small overruns can blow out of proportion before anyone notices. In such an environment, high costs are not surprising. If New York State is interested in good, cost-effective transit, it will get rid of Lhota and replace him with an experienced transit manager with a good history regarding cost control and respect for the democratic process.

New York-Area Governments to Form a Coordinated Transportation Planning Agency

Governors Andrew Cuomo (D-NY) and Phil Murphy (D-NJ) announced that they would cooperate to form a Lower Hudson Transportation Association, or LHTA, to supersede what they described as antiquated 20th-century thinking and bring the region’s transportation into the 21st century. LHTA would absorb the transportation functions of Port Authority, which senior New York state officials speaking on condition of anonymity called “irredeemably corrupt,” and coordinate planning across the region. Negotiations with the state of Connecticut are ongoing; according to planners in New Jersey, the timing for the announcement was intended to reassure people that despite the lack of federal funding for Gateway, a lower-cost modified version of the project would go forward.

But the first order of business for LHTA is not Gateway. The governors’ announcement mentioned that LHTA would begin by integrating the schedules and fares throughout the region. By 2019, passengers will be able to transfer between the New York City Subway and PATH for free, and connect from the subway to the AirTrain JFK paying only incremental fare. Engineering studies for removing the false walls between PATH and F and M subway platforms are about to begin.

Commuter rail fare integration is also on the table. Currently, the fare on the subway is a flat $2.75. On commuter rail, it is higher even within New York City: a trip between Jamaica Station and New York Penn Station, both served by the E line of the subway, is $10.25 peak or $7.50 off-peak on the Long Island Railroad (LIRR). The governors announced that they would follow the lead of European transportation associations, such as Ile-de-France Mobilités in Paris, and eliminate this discrepancy under LHTA governance, which also includes revenue sharing across agencies. Detailed studies are ongoing, but in 2019, LHTA will cut commuter rail fares within New York City and several inner cities within New Jersey, including Newark, to be the same as the subway fare, with free transfers.

Simultaneously, LHTA will develop a plan for schedule integration, coordinating New Jersey Transit, the LIRR, Metro-North, and suburban bus agencies. In order to make it easier for suburban passengers to reach commuter rail stations, the suburban buses will be timed to just meet the commuter trains, with a single ticket valid for the entire journey. Today most passengers in the suburbs drive to the commuter rail stations, but the most desirable park-and-rides are full. Moreover, the states would like to redevelop some of the park-and-rides as transit-oriented development, building dense housing and retail right next to the stations in order to encourage more ridership.

Moreover, the LIRR and New Jersey Transit’s commuter trains currently stub-end at Penn Station in opposite directions. LHTA is studying French and German models for through-running, in which trains from one suburb run through to the other instead of terminating at city center. Planners within several agencies explain that the systems on the Long Island and New Jersey side are currently incompatible – for example, LIRR trains are electrified with a third rail whereas New Jersey Transit trains are electrified with high-voltage catenary – but reorganizing these systems for compatibility can be done in a few years, well before the Gateway project opens.

In response to a question about the cost of this reorganization, one of the planners cited the Swiss slogan, “electronics before concrete.” Per the planner, electronics include systems, electrification, and software, all of which are quite cheap to install, whereas pouring concrete on new tunnels and viaducts is costly. The planner gave the example of resignaling on the subway: the New York Times pegged the cost of modernizing subway signals at $20 billion, and this could increase capacity on most lines by 25 to 50 percent. But the cost of building the entire subway from scratch at today’s costs in New York is likely to run up to $200 billion.

But while the immediate priorities involve fare and schedule integration, LHTA’s main focus is the Gateway project. There are only two commuter rail tracks between New Jersey and Penn Station, and they are full, running a train every 2.5 minutes at rush hour. The Gateway project would add two more tracks, doubling capacity. The currently projected cost for the tunnel is $13 billion, but sources within New York said that this number can be brought down significantly through better coordination between the agencies involved. This way, it could be funded entirely out of local and state contributions, which add up to $5.5 billion. When pressed on this matter, officials and planners refused to say outright whether they expect $5.5 billion to be enough to cover the tunnel, but some made remarks suggesting it would be plausible.

Previous estimates for the costs of Gateway adding up to $30 billion include substantial extra scope that is not necessary. Sources on both sides of the Hudson report that the main impetus for the formation of LHTA was to coordinate schedules in a way that would make this extra scope no longer necessary. “With last decade’s ARC, there was a cavern under Penn Station to let trains reverse direction and go back to New Jersey,” explains one of the planners; ARC was a separate attempt to add tracks to the Hudson rail crossing to Penn Station, which Republican Governor Chris Christie canceled in 2010 shortly after his election. “With Gateway, there was the plan for Penn Station South, condemning an entire Manhattan block for the station for $7 billion. With the plans we’re developing with LHTA we don’t need either Penn South or a cavern to let trains run between stations.”

Moreover, some planners suggested reactivating plans to connect Penn Station and Grand Central as part of Gateway. They refused to name a cost estimate, but suggested that at the low end it could be funded out of already-committed state money. Under this plan, there would be through-running between not just New Jersey Transit and the LIRR but also Metro-North, serving the northern suburbs of New York and Connecticut. Sources at the Connecticut Department of Transportation said they are studying the plan and have reservations but are overall positive about it, matching the reports of sources within New York, who believe that Connecticut DOT will join LHTA within six months.

Officials are optimistic about the effects of LHTA on the region both privately and publicly. The joint press release referred to the metro area as “a single region, in which decisions made in far apart areas of New York nonetheless affect people in New Jersey and vice versa.” Planners in both states cited examples of friends and family in the other state who they would visit more often if transportation options were better. With better regional rail integration, they said, people would take more trips, improving regional connectivity, and take fewer trips by car, reducing traffic congestion and pollution.

Job Sprawl in the US vs. Europe

Both American and European cities have prominent central business districts with high job density. But when jobs sprawl beyond the CBDs, they do so in very different manners on the two sides of the Pond, which is both a cause and an effect of higher US automobile usage. Much of this job sprawl happens in places that people on the other side of the Pond would not recognize as part of The City. Besides the obvious misunderstandings, this can compromise the quality of analysis of urbanism and the transportation required to serve it. In short, the European model, for which my models are Stockholm and Paris, is that jobs sprawl contiguously from the CBD, enlarging its physical area, whereas the North American model, for which my models are New York and Washington, is that jobs leap large swaths of residential neighborhoods into auto-oriented suburbia.

CBDs and job density

Office towers are rare in European CBDs. Paris is largely built up to 6 to 9 stories, and the higher end is more common in residential areas like Nation than in the CBD, which stretches from Les Halles to Saint-Lazare and Etoile. Stockholm has a total of five towers in its center, none especially tall. Contrary to the common European belief that high-rises don’t add density, the mid-rise character of most European CBDs leads to a real limit on their ability to agglomerate. The job-densest arrondissement of Paris, the 2nd, has 60,000 jobs in a square kilometer (look for table EMP T6 here); Midtown Manhattan has about 800,000 jobs in 4 square kilometers.

Not all jobs are in the CBD. Some are local community facilities, such as schools, supermarkets, and hospitals. But even more exportable jobs are not all in the CBD. Some industries cluster in sections of the CBD (such as advertising on Madison Avenue in New York or, traditionally, the media on Fleet Street in London), and similarly some cluster in off-CBD locations, perhaps near one firm that located idiosyncratically. In the other direction, not even the job density of Midtown can contain every workplace that wants a central location, and this pushes out firms that can’t compete for CBD office rents. The difference between the North American and European models is where these firms are likely to locate.

Contiguous sprawl

In both Paris and Stockholm, the solution to the restricted job density of their city centers is, superficially, high-rise clusters in a particular suburban place: La Defense in Paris, Kista in Stockholm. The job density at the center of La Defense is actually higher than in the 2nd arrondissement, though it drops drastically outside the very center, whereas the Paris CBD maintains a density of about 50,000 jobs per km^2 over 4 or 5 square kilometers. In both cases, this leads to spatial inequality: in Paris, the richest suburbs are in the west and southwest, and La Defense is west of the city; in Stockholm, Kista itself is surrounded by working- and middle-class areas, and the favored quarter is separated from it by a lake, but the ill-favored quarter to the south is the farthest away.

However, there is much more to Paris employment than the CBD and suburban office towers. Paris has a total of 1.8 million jobs, with only around one eighth to one sixth of them in the CBD. There are corporate headquarters in La Defense and a number of other suburbs on the RER, but there are job clusters all over the city. My arrondissement, the 12th, has 120,000 jobs in a little more than 6 square kilometers, giving it the same job density as is average for the city. According to OnTheMap, Upper Manhattan, defined to be north of East 96th and West 110th Streets, has 150,000 jobs in 19 square kilometers, and the Upper East and West Sides, defined to be north of 62nd Street so as to exclude Columbus Circle jobs, have a total of 175,000 jobs in 9 square kilometers. While the Upper East and West Sides hold their own, in large part thanks to the hospital cluster around Weill-Cornell, Upper Manhattan does not.

These clusters in Paris are everywhere. In my arrondissement the cluster in question is Bercy, home to the Ministry of Finance; there’s also the university cluster in the Latin Quarter, an under-construction judicial cluster around Clichy-Batignolles, and high-end professional services spillover west of the CBD in the 16th and 17th arrondissements. In effect, office uses are sprawling into otherwise-residential neighborhoods.

In Stockholm, the same situation occurs. Spotify is headquartered two T-Bana stops north of T-Centralen, a short walk from where I used to live in Roslagstull (in fact, one of the people viewing my apartment as I was leaving it worked there). There’s also a prominent peak travel flow of students heading to KTH and the University on the trains from points south. In the south, Södermalm has its own secondary CBD around Slussen, the second busiest T-Bana station after T-Centralen.

Office park sprawl

North American cities do not have high overall job density in the core when one counts both the CBD and surrounding inner neighborhoods, which are typically entirely bedroom communities like Upper Manhattan. Instead, there is discontinuous job sprawl: jobs hop over residential areas into farther-away places, typically suburban office parks. The most famous in American urbanist discourse is Tysons Corner in Northern Virginia, but the Washington metropolitan area is generally replete with edge cities, including Reston, Bethesda, and Silver Spring, all located in the northern and western favored quarters of the region. Kista is really a high-rise version of these edge cities.

Washington is the purest example of office job sprawl. However, even there, there is a complication: there are some nearer job clusters like the Pentagon. New York and other large American cities are the same, with even more complications like this. In New York, the in-state side of the metro area has large suburban job clusters such as White Plains and Stamford, but the New Jersey side includes the formerly independent Downtown Newark, contiguous job sprawl in Jersey City directly facing Lower Manhattan, and very decentralized job sprawl in Middlesex County, contrasting with the centralized office sprawl of White Plains.

Robert Lang and Jennifer Lefurgy call Central Jersey edgeless cities and White Plains and Tysons edge cities. While edge cities exist in Europe, edgeless cities do not. Exurban retail in France resembles American exurban retail, with Carrefour inventing the hypermarket at the same time Wal-Mart did, but there is almost no equivalent of the small American office park. The closest I am aware of, Sophia-Antipolis, is an edge city with a large concentration of jobs, just built at automobile scale without any walkability.

New York is large enough to have an intermediate form: namely, a secondary CBD that’s not contiguous with the main city center. Downtown Brooklyn arose as such a CBD, serving Brooklyn, even though it’s contiguous with Lower Manhattan across the water. Jamaica is the best example, as it is quite far from Manhattan. La Defense should be put in this category as well – it is contiguous with the dense built-up area, if not with the CBD itself, and it is closer to Les Halles than Jamaica is to Midtown.

Favored quarters

There are multiple instances of large American firms moving their entire headquarters to be close to where the CEO lives. IBM moved to Armonk and General Electric moved to Fairfield, both leaving New York, to avoid making executives drive in Manhattan traffic. In Europe, too, the edge cities tend to be in rich areas. The corporate headquarters around Paris cluster in La Defense and to a lesser extent northwestern and southern suburbs, and not in Seine-Saint-Denis.

This is a straightforward consequence of the fact that rich Americans left city centers starting in the early 20th century, culminating in middle-class white flight in the 1950s, whereas Paris and Stockholm remain richer than their suburbs. The inhabitants of the 16th arrondissement of Paris are unlikely to be interested in job sprawl. Instead, the Paris CBD is slowly migrating westward, as retail and office rents at the western end (Etoile) are higher than in the center (the Opera) and eastern end (just west of Les Halles).

One would suspect that in American cities that are richer than their suburbs, the phenomenon of job sprawl would not occur. The problem is that there is no clean example today. Boston is still poorer than its suburbs; Cambridge is quite rich, but is functionally one favored-quarter wedge. San Francisco is overall richer than most of its suburbs, but really the entire strip of land from San Francisco down to San Jose is rich, and at any rate Silicon Valley formed in a then-independent metro region, rather than sprawling out of the center the way White Plains and Tysons Corner did.

However, as white flight is giving way to gentrification, and American cities are economically outgrowing their suburbs, this theory would predict that job sprawl should decrease, with more corporate jobs shifting back to the cities. This seems to indeed be happening: General Electric moved from Fairfield to Downtown Boston, and United-Continental moved its headquarters to the Sears Tower in the Chicago Loop; Aaron Renn periodically talks about the resorting of the American economy, in which the highest-end jobs are back to city centers whereas lower-end jobs are in the suburbs and smaller cities.

Is the US Europeanizing?

There is some evidence to suggest that American cities not only are reducing the extent of job sprawl in the highest pay categories, but also adopting the European pattern of contiguous CBD sprawl. This process is haphazard, and many urban boosters overrate the extent to which near-CBD locations like the West Loop in Chicago or the Seaport in Boston are attracting jobs, but these areas are nonetheless growing.

Boston is perhaps the best example of this trend. Locally, urban boosters anxiously talk about transportation connections to the Seaport, but the biggest action is happening in the other direction. Kendall Square is growing as the Cambridge CBD, with a cluster of tech firms, two stops out on the subway from the central transfer points. Boston is unique that Back Bay is a nearly-contiguous secondary CBD as well, based on extensive postwar urban renewal next to a rich residential neighborhood. This situation is especially notable given that both Cambridge and the Seaport are separated from the CBD by water, with unpleasant walking environments on the bridges, making the organic process of CBD extension more difficult.

Outside Boston, several more examples are notable. In San Francisco, tech jobs within the city cluster not in the Financial District but in the adjacent South of Market (“SoMa”) area. In Chicago, in addition to some growth in the West Loop, there is some job growth on the Near North Side. In New York, the tech jobs cluster in the Meatpacking District: the Google building, which I believe is the second largest Google office after the Googleplex, occupies the block between 8th and 9th Avenues and 15th and 16th Streets. But even before Google, there was growth in adjacent secondary CBDs across the water: Jersey City and Long Island City. Lang and Lefurgy’s writeup on edgeless cities classifies Jersey City as a secondary CBD rather than an edge city because “its context is old”: it’s built out of a near-CBD residential and industrial area, rather than developed from scratch near a road or rail intersection.

Starting last decade, urbanist writers in the United States noted that the US was Europeanizing in its pattern of rich cities and poor suburbs. Brookings was writing about suburbanization of poverty in 2010, describing a 2000-8 trend. The growth of near-CBD office clusters in Boston, New York, San Francisco, and Chicago suggests that the US is also Europeanizing in its pattern of how jobs spill over from the center. Instead of the traditional auto-oriented office park near the CEO’s residence, the highest-income, highest-prestige jobs in the US are decamping to the same near-CBD locations where they can be found in Paris or Stockholm, leaving the sprawl for the poor.

More on Weekend Service Changes and Redundancy

I’m visiting New York again, and the subway is noticeably deteriorating in speed and frequency. But it’s not the speed I’d like to discuss (it’s been covered in the media, both by me and by others) – it’s the weekend service changes. With its 24/7 subway service, New York relies on selective closures of line segments on nights and weekends for maintenance. Expressions like “the D is running on the F and the F is running on the D” are well-known to regular subway riders. However, I don’t think the way the MTA is choosing which lines to shut down every weekend is optimal. I wrote about how different subway lines can be used to substitute for each other in Boston last year; this year, I encountered a big change in New York that, while not as destructive as what I saw in Boston in November, was avoidable.

The trip itself was between Central Harlem and Inwood. I’m staying in Harlem near the 2 and 3 trains, and was visiting friends in Inwood near the A. The most direct path would have been to walk to 125th and St. Nicholas to take the A, but this was blocked because the A was not running north of 168th. The second best option would have been to ride the 2 or 3 to 96th and change wrong-way to the 1, but this was also blocked since the 2 and 3 were not running at all on their Harlem trunk line. The route I took was to walk to 125th and St. Nicholas to take the A up to the 1, changing at 168th Street, an inconvenient transfer involving an elevator (the 1 is deep and only elevator-accessible at 168th).

It was not an especially egregious change. The trip took around 45 minutes and was all-rail; without the changes, it would have been about half an hour. In Boston, the subway shutdowns forced me onto multiple buses in mixed traffic, and a trip that should have taken 40 minutes ended up taking maybe an hour and a half. However, the extra transfer penalty and the difficulty of finding a good all-rail route are real.

In the post about Boston, I argued that even in relatively sparse rail networks, like Boston’s, rail lines are redundant with one another. In New York, this needs no argument, judging by conversations with many New Yorkers, even ones who aren’t railfans, who recognize the extent of redundancy in the system. It would be useful to design weekend service changes to take advantage of this redundancy and not shut down multiple lines that could substitute for each other. In the case of last weekend, if a full shutdown of the 2 and 3 between 96th Street and the Bronx was unavoidable, then the other north-south lines through Upper Manhattan should not be shut down.

Thus, the first principle I’d like to propose for weekend shutdowns is, if one line is shut for repairs, then parallel lines should be fully open. This should cover shutdowns as well as one-way running, in which one direction is only served by express trains and passengers wishing to access local stations are told to ride an express trains and change to a local train going in the opposite direction.

In fact, given the extent of redundancy in the core, I would propose that if local service is shut down in one direction, then it should be shut down in the other direction as well. Backtracking is cumbersome and rarely worth it: stop spacing is often close enough that walking from the express stop is faster, and in Manhattan it’s usually easier to just go on a parallel line instead.

Then, to compensate for loss of service on a line that’s being closed for repairs, there should be more service on parallel lines. Subway service changes in New York, called general orders, or GOs, fall under a labor agreement in which train operators and conductors are still paid for the time they would’ve had to work if service had run normally. The MTA should negotiate a change in which those drivers and conductors can be redistributed to nearby subway lines without this counting as overtime. If express trains on a route are closed, then there should be more local service, and if a trunk line is closed, as the 2 and 3 were last weekend, then there should be more service on parallel lines (in this case, the 1, 4, 5, A, and D) to absorb the extra ridership. With reduced service, the trains can get quite busy, more so than the crowding guidelines indicate, if my weekend trips are any indication.

Finally, the MTA should build crossovers and switches to enable more efficient single-tracking. This is unlikely to work on four-track lines, since trains would have to cross three tracks to get around a work zone, but on two-track lines, including the L, the 7, and most tails outside the Manhattan core, it would be useful, especially on lines that use island platforms rather than side platforms. On the newest lines, which use bored tunnel, it’s easy to do maintenance on one track without disturbing the other track, as long as there are crossovers at regular locations. On the older lines the situation is the opposite: trains need to slow near work zones, unless some hard barriers are built between parallel tracks, but installing new crossovers is relatively easy. More shutdowns may be required to install such switches, but the subsequent benefits to making weekend trackwork easier are substantial.

Off-Peak Subway Service in New York

New York City Comptroller Scott Stringer has a new report out about the poor state of off-peak subway service. It’s a topic I’ve talked about a lot here (e.g. here), but there’s a big difference in focus: I normally talk about midday service for efficiency reasons, and as far as I remember this is the bulk of what I discussed with report author Adam Forman, but the report itself highlights non-traditional commutes in the early morning and evening:

Traditional commuters
(depart 7-9 am)		 Early morning commuters
(depart 5-7 am)
Median Income $42,300 $35,000
Bachelor’s Degree or Higher (Age 25+) 52% 31%
Foreign Born 47% 56%
Person of Color 64% 78%
Work in Healthcare, Hospitality, Retail, Food Services, or Cultural industries 36% 40%
Growth in the Last Quarter Century 17% 39%

Citywide, there are 1,888,000 commuters leaving to go to work between 7 and 9 am, and 711,000 leaving between 5 and 7. The latter group has to contend with much worse subway frequencies: the report has a table (chart 8) detailing the reduction in frequency, which is typically about half. The report does not say so, but an additional hurdle facing early-morning commuters is that some express trains run local: for example, the northbound A train only starts running express at 6 in the morning, forcing a substantial minority of early morning commuters to ride what’s effectively the C train.

The one saving grace in the early morning, not mentioned in the report, is that buses aren’t as slow. For example,  the B6 limited takes 1:11 end-to-end at 6 am, compared with 1:26 at rush hour. However, this is a 16 km route, so even the faster speed at 6 am corresponds to an average speed of 13.7 km/h, which is not competitive with a bicycle. Moreover, in practice, slow circumferentials like the B6 are used in situations where transferring between subway lines is not viable or convenient, such as early in the morning, when subway frequencies are low; this means that far from a substitute for slower rush hour buses, early morning buses have to substitute for much faster subway lines.

The report has charts about subway and bus service by the time the route begins operation. As expected, there’s a prominent morning peak, and a slightly less prominent afternoon peak. In the evening there’s a dropoff: 350 subway runs begin around 9 pm compared with just under 600 subway runs in the morning peak, a reduction of 40%. For buses, the dropoff is larger: about 1,700 versus 3,700, a 54% reduction. The most worrying trend is that the buses peak at the same time as the subway in the afternoon, starting at 4:30 or so; in reality, buses are often a first-mile rather than a last-mile connector, which means that people returning from work typically ride the subway and then the bus, so we should expect buses to peak slightly later than trains, and drop off in the evening at a slower rate. Instead, what we see is the same peak time and a faster dropoff.

Some of this can be attributed to operating costs. Buses have lower fixed costs than trains and higher marginal costs, so the economics of running them at less busy times are weaker than those of running trains. However, in reality buses and trains in New York run as a combined system; running just the subway in the evening but not the buses means that people can’t come home from work if they live in neighborhoods not connected to the subway.

Evening frequencies on many routes are low enough that they are almost certainly negatively impacting ridership. Some individual subway routes run every 11-12 minutes in the evening, including the B, C, D, W, and 5; in the every 9-10 minutes category are the 2, 3, A, F, J, N, and R. Other than the J, these are all branches sharing track with other lines, but they branch off the trunks and recombine. A Bronx-bound rider on the 2 and 3 can only ride the 2, and a Flatbush-bound rider can choose between the 2 and a 3-to-5 transfer, both of which are infrequent. Without timed transfers, the effective frequency as experienced by the rider remains low, about every 10 minutes.

This isn’t how other top metro systems work – in Paris the trains on Metro Line 9, not one of the top lines in the system, come every 7 minutes at 10 or 11 at night. The RER is less frequent on individual branches, but the individual branch points are all outside the city except on the RER C, sometimes well outside it. Other than on the RER C heading west, the branch points are at worst 6 km outside the center (at Vincennes), more typically 10 km (such as Nanterre and Bourg-la-Reine), and at best 16-18 km out (Aulnay and Villeneuve-Saint-Georges). In New York, the R and W branch at Lexington and 60th, a little more than 2 km outside Times Square, and the Q and N branch even earlier; the A-B-C-D branch and recombine at Columbus Circle, and branch again at 145th Street, 8.5 km out of Port Authority. This branching affects a majority of bedroom communities in the city, including almost the entire Bronx, much of Upper Manhattan, all of Queens except the 7, and Central and Southern Brooklyn.

To my knowledge, there is no public study of the effect of frequency on ridership. Occasionally there are ridership screens that incorporate it, but the examples I know are designed around the needs of specific project studies. There can be rules of thumb about frequency at different scales (the smaller the scale, the higher the minimum frequency is), but without more careful analysis, I can only bring up some best industry practices. It does not seem common to run metro trains every 10 minutes in the evening. On the Piccadilly line, there are 22 northbound trains departing Leicester Square between 9 and 10 in the evening, of which 19 go all the way to Cockfosters. On the Central line, 24 trains depart Oxford Circus eastbound, 9 going to Epping (in Essex, 31 km from Oxford Circus and 27 km from Bank), and another 13 serving Newbury Park, in outer East London.

Evening service also has one more complication: it serves several distinct markets. There are commuters working non-traditional hours, themselves split into shift workers and professionals who work late (I spoke to several Manhattan lawyers who told me that they work from 10 in the morning to 8 in the evening). There are tourists and local leisure travelers, some coming late from work after dinner and some coming from a non-work destination. Non-work trips don’t always have the same centers as work trips: in London, non-work trips are dominated by the West End, with little contribution from the City, whereas in New York, presumably Lower Manhattan punches below its weight while Union Square punches above its weight. New York already takes care of non-work trips in the evening, with high frequencies on the 1, L, and 42nd Street Shuttle (“GS” in chart 8), but its frequency guidelines are unfriendly to commuters who are working late.

Construction Costs and Experience

The most persistent criticism I have heard of my writings on construction costs, coming from YIMBY Princeton, is the importance of gradual expertise and experience. Against my claim that Americans build subways for higher costs than the rest of the world due to poor management practices, regulations, and procurement, and scope creep, YIMBY Princeton says that high costs are a result and not a cause of the rarity of American subway investment. I believe that high US costs are endogenous and therefore the US is reluctant to fund rail transit; he believes that disinterest in transit is endogenous and if the US were willing to build more rail lines, then construction costs would naturally go down through economies of scale and steady accumulation of project management expertise. I promised last year that I would go over his argument more carefully, and am going to do so in this post.

The obvious difficulty with this debate is that we agree there is negative correlation between construction costs and the extent of construction, and disagree on causation. Wikipedia lists 55 countries with metro systems, with a handful more with under-construction metros, but this is not enough of a dataset for large-n studies. There are too many control variables – for example, it’s easier to build the first subway line than the tenth, which reduces the proper comparisons for New York to a handful of large cities. Instead, the only real way to figure out what causes what is to rely on a handful of natural experiments.

I can come up with a number of natural experiments. One is ambiguous about causation: the role of poor project management in the high construction costs in Boston and Paris. The others all suggest that high costs are endogenous to the US, rather than unwillingness to build subway tunnels. These include the history of construction costs in New York, London, and Paris in the 1930s; the construction costs in London today; and the history of construction costs in Seoul from the 1970s to the present.

Project management

I know of two overpriced rail extensions blamed explicitly on poor project management: the Green Line Extension in Boston, and Grand Paris Express. As I explained in CityLab a few months ago, the GLX was budgeted at $3 billion for just 7.5 km of light rail trench in preexisting open cuts, but the MBTA cut this to $1.1 billion in actual construction plus $1.2 billion in rolling stock and sunk costs through hiring a more experienced project manager. In Paris, one of the reasons the Cour des Comptes cited in its report about the cost overrun is lack of experience in managing such a large project; as a result, the 200 km system, with 160 km underground, is now up to €35 billion.

The problem is that even with better cost control, Boston’s construction remains pricey. At $150 million per km, GLX is expensive for a line in a preexisting right-of-way, and not far behind GPX’s $220 million per km for an 80% underground network. While both Boston and Paris can expect future construction to be cheaper if they apply the lessons of GLX and GPX cost overruns, their absolute costs remain different, with Boston spending more per unit than Paris. At best this is neutral between my explanation and that of YIMBY Princeton.

Construction costs, dieselpunk edition

New York’s subway construction costs have risen since the start. In 1900-4, the First Subway cost $32 million for 22 km of subway and 1 km of viaduct (namely, the 125th Street viaduct on today’s 1 train), and another $3 million for 9 km. In today’s money, this is $39 million per km underground and $9 million per km elevated. JRTR has some statistics for the Dual Contracts, built in the 1910s and early 20s, and the IND, built in the 1930s. The Dual Contracts cost $366 million, equivalent to around $8 billion today; the total route-length added was about 180 km, of which 70 km was underground, consistent with a cost of about $80 million per underground km. Then the IND cost $815 million, equivalent to about $14 billion today, for 97 route-km, practically all underground, or about $140 million per km.

The projected cost of Second Avenue Subway kept rising. In 1929 the projection was $86 million, or about $1.2 billion in today’s money, or $90 million per km; this was before the IND cost overruns materialized (at a time of general deflation). In 1939 it was up to $249 million, or $4.2 billion today, about $320 million per km; by 1949, it had crept up to $500 million, or $5 billion today and $390 million per km. Put another way, WW2-era America, a country that had just built massive public works in the Depression as well as the war, including the IND and Chicago’s two subway lines through the Loop, was already projecting a higher per km cost than is routine in nearly the entire world today. Moreover, the plan was to build Second Avenue Subway cut-and-cover, a technique that is cheaper today than the deep boring typical of comparable infill subways in the first world.

I have less data than I’d like for other cities’ construction costs in the interwar era, but where they exist, they are a fraction of New York’s. The London Underground extension to Cockfosters in the Depression cost £4m for 12 km, 60% underground, per Wikipedia. In today’s money it’s $45 million per underground km. In Paris, there was little growth in real costs between 1913 and 1930: according to a presentation by Pascal Désabres, construction costs in today’s money in both 1913 and 1930 were about €23 million per km, or about $29 million per km.

London’s mounting costs

In the 1930s, London built an Underground extension for $45 million per km. After the war, it could no longer do so. According to numbers in the Financial Times, the Victoria line cost £4.5 million per km in the 1960s, all underground, which is about $110 million per km in today’s money, while the Jubilee line, built in the 1970s, cost about $250 million per km.

The Victoria and Jubilee lines were more complex projects than the Cockfosters extension, going under older Underground lines. The Jubilee line also included the construction of a transfer station with the Northern and Bakerloo lines at Charing Cross, whereas previously they only connected at the next two stations, Embankment and Waterloo. However, the construction technique, the tunnel-boring machine, is one that is supposed to have a much smaller city-center premium over outlying construction, since there is no surface disruption.

But whereas the Victoria and Jubilee lines had excuses for their high costs, more recent Underground extensions do not. In the 1990s, the Jubilee line extension cost around $500 million per km in today’s money, going under a few older Underground lines and crossing the Thames four times (in an environment with not much underwater premium) but mostly extending the system to the east, to previously underserved areas like Canary Wharf. The under-construction Battersea extension, crossing under one older line and serving a relatively undeveloped area at Battersea Power Station, is about $550 million per km. The next Underground extension under discussion today, that of the Bakerloo line to Lewisham, is budgeted at £3.1 billion over 7 km, or about $620 million per km, crossing under no Underground lines and largely following a wide road.

Under YIMBY Princeton’s theory, London’s construction costs should be decreasing as it obtains more experience tunneling in a constrained urban area with millennium-era sensitivity to environmental impact like noise. But on the contrary, costs keep growing.

Seoul’s low costs

If London is the expensive city that should under YIMBY Princeton’s theory get cheaper but isn’t, Seoul is the cheap city that should have been expensive in the 1980s but wasn’t. JRTR has data from the 1970s to the 1990s: after an increase at the beginning, Seoul’s construction costs stabilized in the 1980s and 90s at about $80 million per km in constant dollars in today’s money. These costs seem to persist today, judging by the Sin-Bundang Line, which cost 1,169 billion won for 18 km, converting to about $90 million per km in PPP dollars.

Seoul is consistent with the theory that costs are endogenous to a city or country. There is high correlation between the construction costs of different lines within the same city: having set non-US records with the Jubilee line extension, London keeps building very expensive Underground extension of the Northern and Bakerloo lines; Paris is spending around $250 million per underground km on a number of Metro extensions; Seoul keeps building subways at just under $100 million per km.