Category: Urban Transit

When Reliability Matters Above All Else

This post is about situations in which the most important thing for transportation is reliability, more so than average speed or convenience. It’s inspired by two observations, separated by a number of years: one is my own about flying into or out of Boston, the other is from a New York Times article from yesterday describing a working-class subway rider’s experience.

My observation is that over the years, I’ve used Logan Airport a number of times, sometimes choosing to connect via public transportation, which always involves a bus as the airport is not on the rail network, and other times via taxi or pickup. My choice was always influenced by idiosyncratic factors – for example, which Boston subway line my destination is on, or whether I was visiting someone with a car and free time. However, over the last eight years, a consistent trend is that I am much more likely to use the bus arriving at the airport to the city than departing. I know my own reasoning for this: the bus between South Station and the airport is less reliable than a cab, so when in a crunch, I would take a train to South Station (often from Providence) and then hail a taxi to the airport.

The New York Times article is about a work commute, leading with the following story:

Maribel Burgos barely has time to change into her uniform before she has to clock in at the McDonald’s in Lower Manhattan where she works, even though she gives herself 90 minutes to commute from her home in East Harlem.

It does not take 90 minutes to get between East Harlem and Lower Manhattan on the subway. The subway takes around half an hour between 125th Street and Bowling Green, and passengers getting on at one of the local stations farther south can expect only a few minutes longer to commute with a cross-platform change at Grand Central. Taking walking and waiting time into account, the worst case is around an hour – on average. But the subway is not particularly reliable, and people who work somewhere where being five minutes late is a firing offense have to take generous margins of error.

When is reliability the most important?

What examples can we think of in which being late even by a little bit is unacceptable? Let us list some, starting with the two motivating examples above:

  • Trips to the airport
  • Work trips for highly regimented shift work
  • Trips to school or to an external exam
  • Work trips for safety-critical work such as surgery
  • Trips to an intercity train station

In some of these cases, typically when the riders are of presumed higher social class, the system itself encourages flexibility by arranging matters so that a short delay is not catastrophic. At the airport, this involves recommendations for very early arrival, which seasoned travelers know how to ignore. At external exams, there are prior instructions of how to fill in test forms, de facto creating a margin of tolerance; schools generally do not do this and do mark down students who show up late. Doctors as far as I understand have shifts that do not begin immediately with a life-critical surgery.

But with that aside, we can come up with the following commonalities to these kinds of trips:

  • They are trips to a destination, not back home from it
  • They are trips to a fairly centralized and often relatively transit-oriented destination, such as a big workplace, with the exception of regimented shift work for retail (the original NY Times example), which pays so little nobody can afford to drive
  • They are disproportionately not peak trips, either because they are not work trips at all, or because they are work trips for work that is explicitly not 9-to-5 office work
  • They are disproportionately not CBD-bound trips

The first point means that it’s easy to miss this effect in mode choice, because people can definitely split choice between taxis and transit or between different transit modes, but usually not between cars and transit. The second means that driving is itself often unreliable, except for people who cannot afford to drive. The third means that these trips occur at a point in time in which frequency may not be very high, and the fourth means that these trips usually require transfers.

What does reliability mean?

Reliability overall means having low variance in door-to-door trip time. But for the purposes of this discussion, I want to stress again that trips to destinations that require unusual punctuality are likely to occur outside rush hour. Alas, “outside rush hour” does not mean low traffic, because midday and evening traffic in big cities is still quite bad – to take one New York example with shared lanes, the B35 steadily slows down in the first half of the day even after the morning peak is over and only speeds up to the 6 am timetable past 7 pm. Thus, there are twin problems: frequency, and traffic.

Traffic means the vagaries of surface traffic. Buses are generally inappropriate for travel that requires any measure of reliability, or else passengers have to use a large cushion. Everything about the mixed traffic bus is unreliable, from surface traffic to wait times, and bunching is endemic. Dedicated lanes improve things, but not by enough, and unreliable frequency remains a problem even on mostly segregated buses like the Silver Line to the airport in Boston.

Frequency is the harsher problem. The worker commuting from Harlem to Lower Manhattan is if anything lucky to have a straight-short one-seat ride on the 4 and 5 trains; most people who need to be on time or else are not traveling to city center and thus have to transfer. The value of an untimed transfer increases with frequency, and if every leg of the trip has routine 10-minute waits due to bunching or just low off-peak frequency guidelines, the trip gets intolerable, fast.

What’s the solution?

Bus redesigns are a big topic in the US right now, often pushed by Jarrett Walker; the latest news from Indianapolis is a resounding success, boasting 30% increase in ridership as a result of a redesign as well as other changes, including a rapid bus line. However, they only affect the issue of reliability on the margins, because they are not about reliability, but about making base frequency slightly better. New York is replete with buses and trains that run every 10-15 minutes all day, but with transfers, this is not enough. Remember that people who absolutely cannot be late need to assume they will just miss every vehicle on the trip, and maybe even wait a few minutes longer than the maximum advertised headway because of bunching.

Thus, improving reliability means a wider toolkit, including all of the following features:

  • No shared lanes in busy areas, ever – keep the mixed traffic to low-traffic extremities of the city, like Manhattan Beach.
  • Traffic signals should be designed to minimize bus travel time variance through conditional signal priority, focusing on speeding up buses that are running slow; in combination with the above point, the idea of giving a late bus with 40 passengers the same priority at an intersection as a single-occupant car should go the way of the dodo and divine rights of kings.
  • Off-peak frequency on buses and trains needs to be in the 5-8 minute range at worst.
  • Cross-platform transfers on the subway need to be timed at key transfer points, as Berlin manages routinely at Mehringdamm when it’s late and trains run every 10 minutes (not so much when they run every 5); in New York it should be a priority to deinterline and schedule a 4-way timed pulse at 53rd/7th.
  • Branch scheduling should be designed around regular gaps, rather than crowding guidelines – variation between 100% and 130% of seats occupied is less important to the worker who will be fired if late than variation between waiting 4 and waiting 8 minutes for a train.
  • Suburban transit should run on regular clockface schedules every 30, 20, or 15 minutes, with all transfers timed, including with fare-integrated commuter trains.

New York City Zoning and Subway Capacity

I got a bunch of accolades and a bunch of flaming replies over a tweetstorm imagining a bigger, better New York. Some people complained about my claim that subway trains in Brooklyn are underfull; I urge everyone to read my analysis of data from 2016 – it’s still relevant today, as the only big change is that Second Avenue Subway has reduced Upper East Side crowding. The point of this post is to demonstrate where zoning should definitely focus on adding more apartments, to fill trains that are not yet full.

The map

Instead of using the current subway map, let us start with a deinterlined map:

The reason for using this map is that it’s cleaner than the real map, since there is no track-sharing between routes of different colors, and not much route-sharing (one color local, one express). Getting from here to this map is cheap but not free, as it requires certain junction rebuilds, especially on the 2/3. I ask that my commenters resist the temptation to argue over the details of this map, since the point about zoned capacity does not really depend on questions like whether the E runs local in Queens and the F runs express or the reverse.

Where there is capacity

In 2016, three directions on the subway were truly at capacity, surpassing 4 standees per square meter: the 2/3 and 4/5 coming into Midtown from Uptown, and the L. The analysis looks at crowding on trains entering the Manhattan core, so it lumps lines from Queens based on which tunnel they enter from, which underestimates crowding on the E, since it shares tracks with the under-capacity M. Counted properly, the express Queens Boulevard trains should be viewed as near or at capacity as well, the F having 3.33 standees per square meter and the E having somewhat more.

Additional lines with capacity crunches, with about 3 standees per square meter or more, include the A/D coming in from Uptown, the 6, and the Astoria Line (then the N/Q, now the N/W). The 1 and 7 trains have capacity crunches as well in outlying areas: the 7 is overcrowded until it hits the transfer points to the E/F and N/W but has plenty of space in Long Island City, and the 1 is fairly crowded north of the junction with the express trains and then unloads passengers onto the overcrowded 2/3. These areas should not be deemed to have much spare capacity until such time as operations on the subway improve, permitting higher frequency and eventually more lines.

In contrast, the remaining lines have space, often plenty of space. Everything in Brooklyn except the L and to some extent the J/M/Z is underfull: these trains have high frequency as determined by crowding guidelines at the Uptown or Queens end, but in Brooklyn there are fewer people today so the ridership is weaker. The local lines on the Upper West Side both have plenty of space on the trains as well as space on the tracks for more trains if need be. The 7 downstream of Queensboro Plaza has plenty of space, and the local Queens Boulevard trains downstream of Jackson Heights have nowhere for passengers to transfer to an overcrowded express service.

Since I’m relying on data from 2016, there’s no accounting for Second Avenue Subway. Even then, the 4/5 was only the third most overcrowded trunk line entering the Manhattan core, and it’s likely that there’s additional capacity coming from the new line. There’s certainly space on the tracks for more trains on Second Avenue, and one of the goals of deinterlining specifically is to make it feasible to run more service on this line, which currently only runs a train every 6-8 minutes at rush hour.

The map of where New York could add housing

The map excludes parts of Lower and Midtown Manhattan where the highest and best use is commercial rather than residential. But the boundaries there are deliberately crude: Downtown Brooklyn, NYU, and the Meatpacking District are drawn, to avoid excessive fragmentation of the drawn area, while Chelsea and Hell’s Kitchen are excluded as too close to Midtown.

The map also does not look at considerations other than capacity. Some of the highlit areas on the Upper East and West Sides and Lower East Side are already built to very high density, at least on the avenues and major streets; these areas should be the template of how the rest of the city should look. At the other end, East New York has too weak demand for massive construction, especially if everything to its west is upzoned.

However, large swaths of desirable, close-in areas with relatively short buildings are highlit. Rich inner Brooklyn neighborhoods like Park Slope and South Brooklyn are currently built to missing middle density, with a floor area ratio of about 1.5 away from corner lots. A more appropriate floor area ratio in these neighborhoods is 12, corresponding to tapering buildings in the 20-30 story range, as on the avenues on the Upper East and West Sides. Park Slope is half an hour from Midtown by subway, and less than that from Lower Manhattan. The population of these neighborhoods is perhaps 150,000, and should be more than a million given their proximity to job centers.

Subway deserts and future additions

The map is designed to work with more or less the same service as today, maybe with slightly more frequency on lines that could handle it easily (that is, Second Avenue Subway). But what about future service? The L train is overcrowded, but only runs 19 trains per hour at the peak due to electrical limitations, and could go up to 26 with better electrical capacity, or for that matter lighter trains drawing less power during acceleration. Further extensions of Second Avenue Subway could more effectively relieve pressure off the 4/5, to the point of creating more capacity in the Bronx, which remains well below peak population. Commuter rail modernization opens up large swaths of Queens. Decades in the making extensions on Nostrand and Utica fill in the transit desert in southeast Brooklyn, currently served by buses that nominally come every 2 minutes and in practice comes in platoons of 4 every 8 minutes.

As with the map above, a hypothetical map of development sites assuming reasonable subway expansion includes areas that would be unlikely to actually see new development. Williamsburg and Greenpoint may turn into forests of towers given the opportunity, but in neighborhoods like Sheepshead Bay and East Flatbush developers might well stick to the occasional 6-to-10-story mid-rise building that would not look out of place in Paris. In Eastern Queens, the desired density is probably spiky, with clusters of tall buildings around LIRR stations surrounded by single-family houses and missing middle, much like the structure of density in Toronto and Vancouver.

The Hazards of Federal Subsidies for Operations

There’s an interesting discussion on Twitter, courtesy of Adam Batlan, about federal subsidies for capital funding versus operations. It’s become a popular reform proposal among American public transport advocates, who are frustrated with the status quo of federal funding for capital but not for operations. Unfortunately, the proposed change to the status quo – federal funding of operations and maintenance – is even worse than the status quo. The hazards of outside funding sources for operations are considerable and unavoidable, whereas those of outside funding for capital expansion can be mitigated by defining expansion appropriately, to the exclusion of ongoing maintenance.

Why federal funding should only go to expansion

Public transportation has ongoing operating expenses, and capital funding. Ongoing expenses can only change gradually – rail service in particular is dominated by fixed costs, like maintenance, and service changes have little effect on operating costs. This argues in favor of steady funding for operations.

Can federal funding be this steady? The answer is no. The federal government is where politics is. People with serious differences in opinion over issues including the overall size of federal spending, spending priorities, and how sensitive spending should be to economic conditions contest elections, and if one side has a majority, that side will get its legislative way. Nor is this some artifact of two-party majoritarianism. In consensus democracies the salience of a majority is if anything higher – there are big differences in policy, including transportation policy, between the various parties of Switzerland or the Netherlands, as the parties have to deliver results to attract voters rather than relying on polarization and partisan identity.

This kind of politics is very good when it comes to debating one-time capital projects. A center-right government committed to austerity with little attention to climate change, for example Germany since 2005, will not spend much money on rail expansion, and railroads will formulate their plans accordingly. The key here is that planning around maintaining current operations without expansion is not difficult, whereas planning around sudden cuts in operating funding is.

The issue of ongoing capital expenses

Current US policy is for the federal government to fund capital expenses, but not necessarily expansion. Normal replacement of equipment and long-term maintenance both receive federal funding. This is bad policy, because the way agencies respond to changes in funding levels is to defer maintenance when the federal government is stingy and then cry poverty when the federal government is generous.

The most extreme case of this is the state of good repair (SOGR) scam. The origins of SOGR are honest: New York City Transit deferred maintenance for decades, until the system collapsed in the 1970s, leading to a shift in priorities away from expansion and toward SOGR in the 1980s and 90s. There were tangible improvements in the last era, raising the mean distance between failures on the subway from about 10,000 km in 1980 to 250,000 in the 2000s. But this process led to a trend in which agencies would deliberately defer maintenance, knowing they could ask for SOGR funding letting them spend money without having anything to show for it.

By the 21st century, New York’s SOGR program turned into such a scam. The MTA capital plans keep having line items for achieving SOGR, but there are no improvements, nor does the backlog appear to shrink. If anything, throughout the 2010s service deteriorated due to slowdowns, until Andy Byford began the Saving Precious Seconds campaign. The same scam appears elsewhere, too: Amtrak deferred maintenance in the 2000s under political pressure to look profitable for privatization, a Bush administration priority, and when Obama was elected and announced the stimulus, Bush-installed CEO Joe Boardman began to talk about SOGR on the Northeast Corridor as a way of hogging billions of dollars without having to show increases in speed.

The forward solution to this problem is to credibly commit not to fund maintenance, ever. The fix-it-first maxim is for local governments only. The maxim for outside funding should be that any request for funding for maintenance or replacement is a tacit admission the agency cannot govern itself and requires an outside takeover as well.

The issue of frequency

The problem the thread linked to at the beginning of this post sets to solve is that some cities get money to build a light rail line but then only run it every 20 minutes. This, however, is a problem of incompetence rather than one inherent to the incentives.

A long-term revenue-maximizing agency, confronted with an urban rail line that runs every 20 minutes, will increase its frequency to at worst every 10 minutes, secure in the knowledge that the long run elasticity of ridership with respect to frequency in this range is high enough that it will make more money this way. This remains true even for a dishonest agency, which has no trouble maximizing long-term revenue by deferring maintenance and then asking for SOGR money when funding is available.

This fact regarding frequency is doubly true if the trains already run frequently at rush hour and only drop to 20-minute frequency off-peak. Fleet costs are determined by the peak, and large peak-to-base service ratios require expensive split shifts for crews. Therefore, a bump in off-peak frequency, especially from such a low base as 20 minutes, will increase ridership for very little increase in operating cost.

The thread does not mention the issue of connecting bus service much – I got yelled at for proposing half-hourly local buses timed with commuter trains – but there, too, the rule of only subsidizing expansion rather than maintenance or operation leads to good enough incentives. In Seattle, light rail expansion has led to bus service changes designed to feed the trains, increasing bus ridership even as rail service replaces the most crowded corridors.

The bus cuts of (for example) San Mateo County in response to rail expansion should then be put in the same basket of pure incompetence with the light rail line that runs every 20 minutes off-peak. The incentives line up in one direction, but due to such factors as unfamiliarity with best practices and managers who do not ride the trains they run, management goes in the other direction.

The forward solution here is to stick to funding by expected ridership. If the service plan involves low frequency, this should show up in the ridership screen and penalize the project in question, while urban rail lines that run every 5 minutes get funded.

Paint the Trains in Themes

Most urban rail networks in the world use color to distinguish lines, either alone or in combination with line names or numbers. Moreover, most of these networks have different train fleets for different metro lines – for examples, the trains on the Northern line are used only on the Northern line, and the trains on Paris Metro Line 1 are used only on Line 1. The interiors of these trains have static line maps dedicated to the lines they serve. Occasionally, the trains are also painted in their thematic colors, as in Boston. So, why not extend this and not only paint trains in their thematic colors, but also have different art on each trainset, using the thematic color?

A blue line, like the Piccadilly line or the RER B, would use drawings that incorporate the color blue in some essential way. For example, one trainset could depict an endless ocean, one could depict the sky, one could depict glass-clad skyscrapers that appear blue, and so on.


The key here is to make each trainset visually distinct and recognizable. Part of the reason is pure art: it introduces more interesting variability to a mundane activity, serving the same purpose as street sculptures. This exists in Japan to some extent, with public mascots and Hello Kitty trainsets, but this could generalize to every trainset. In a large city, this would require finding several dozen different paint schemes per color, ideally each by a different artist using a variety of styles.

But there’s another reason for this scheme: it makes it easier for passengers to remember which train they were on if they lost something or wish to report a crime. Right now, trains are tracked by model number, which passengers have no reason to remember after getting off the train. In contrast, a heraldic system is easier for passengers to retain, especially if the art covers both the exterior and the interior of the vehicle.

For the latter reason, it’s fine to be repetitive and paint every car in a trainset with the same scheme: passengers can roughly remember if they were near the front or back of the train, so if they lost something on the train, they can give enough information to reduce the search space to maybe two cars. Trainsets on modern urban rail systems are almost always permanently coupled, often in open gangways – even New York permanently couples cars into half-trains and joins two sets at a time to form a train, making it feasible to associate paint schemes with entire sets rather than individual cars.


The choice of art should rely on local history, geography, mythology, and culture whenever possible. For example, in the Eastern United States, one red trainset could depict brilliant fall foliage, but in Europe, trees do not turn red in the autumn so the reference would not be easily understood. In Japan, trees turn red in the spring and not the fall, so a red trainset could be painted with the cherry blossom. While Paris does not associate red with the color of leaves in any season, it was historically a center for impressionist art, so one blue trainset could have an impressionistic painting of foliage depicting it in blue.

Iconic food may be another intensely local element to paint in some cities. Everyone in New York knows what a bagel, a New York-style pizza, and a hero sandwich are, and New Yorkers of all ethnic and social groups eat them. At the deli, the professor and the security guard may well order the same pastrami hero. The same is true of döner and currywurst in Berlin, and bento boxes and yakitori in Tokyo.

Mythology and history add more recognizable symbols that are specific to the region or country. London and Paris may each find famous battles to commemorate, just as London names one of its intercity train stations after Waterloo and Paris names one of its after Austerlitz. An American city, especially Washington, may depict Union troops in the Civil War or the raising of the flag at Iwo Jima. Every major city can find an episode of its labor history to paint on one of its trainsets, in red of course. Mythology can add recognizable elements, such as fire-breathing dragons in red, Poseidon in blue, and pots of gold in yellow. Those elements would naturally look differently in a non-Western city like Tel Aviv or Singapore, but the principle is the same.

Diverse cities especially benefit from being able to depict their various cultural backgrounds, making different trainsets more visually distinct. Paris can paint some of its green and black trains with Arabic calligraphy, New York and Chicago can depict black Union troops with blue uniforms, Washington can depict the March on Washington with a blue sky or green lawn background, London can depict the Windrush and lotus art and Muslim South Asian architecture. These cities are all predominantly Western, but have large and growing minorities from non-Western backgrounds or from backgrounds with different takes on Western cultural production (such as black and Hispanic Americans), and should reflect the majority culture as well as the minorities, treating the transit network as a microcosm of the entire population.

Commercial culture and advertising

The plan should be to keep each design for a long time, potentially the entire life of the trainset, or at least through a midlife refurbishment. History, mythology, and geography all provide themes that are sufficiently long-run to remain relevant over the long life of a train.

In some cases, commercial properties can both be expected to exist for a long time and have well-known thematic colors. Examples include Star Wars and the iconic light saber colors, the best-known Pokemon, Hello Kitty, many superheroes, and the Smurfs. Transit agencies could enter long-term advertising contracts with Disney, Nintendo, and other long-lived corporations producing popular culture, and paint their properties on trainsets.

Advertising on the subway has a long history, and can coexist with painting the train if the regular ads are contained to the usual posters. It’s already spilling into painting an entire train: the Hello Kitty train is one example, but negative examples exist as well, when New York wraps an entire subway trainset in an ad for a television show that will be forgotten in a few years.

This kind of long-term advertising, in contrast, reinforces the recognizability of individual trainsets as no two trainsets should ever be painted with the same property (though trains of different colors may be painted with different Pokemon, or one with Jedi and one with Sith, etc.). Moreover, the paint scheme should be stable over 20 years – temporary modifications to help advertise a new film, video game, TV series, or book in the franchise should cost extra, and potentially be treated as regular ad posters.

However, there should be a limit to commercialization: the majority of subway paint schemes should not be based on global brands, but on local factors. Pokemon is everywhere, but the cherry blossom, recognizable skylines, picturesque mountains, and historical battles are specific to a country or region.


Just as cities often have art exhibits at subway stations, and just as they sometimes paint the trains on each color line with the color it’s named after, subway and regional rail networks can paint trains individually in thematic colors. In the largest cities, like New York and London, this could well involve more than a thousand distinct paint schemes; this is fine – those cities have enough artists and enough inspiration for a thousand trainsets.

Overall, the combination of some commercial properties with various aspects of history, geography, tourism, food, and mythology, curated from the majority group as well as from various ethnic and religious minorities, is exactly the mosaic that makes the city’s culture. One of the two prime reasons to do this is as a tool to help passengers remember what train they were on. But the other one is art, which simultaneously is aesthetic and sends a message: on the train we are all New Yorkers, or Londoners, or Parisians, or Berliners.

New York Rolling Stock Costs are Skyrocketing

The Metropolitan Transportation Authority has just released its capital plan for 2020-4. The cost is very high and the benefits substantial but limited, and I urge people to look over criticism by Henry Grabar at Slate about elevators and Ben Kabak’s overview at Second Avenue Sagas. Here I am going to focus on one worrying element: the cost of the trains themselves, on both the subway and commuter rail.

I started comparing subway construction costs nearly ten years ago. Here’s an early post on Second Avenue Sagas, hoisting something I wrote in comments. Over here I started writing about this in 2011. Early on, I was asked about the costs of the trains themselves rather than the tunnels, and said that no, there’s no New York premium there. At the time the most recent rolling stock order for the subway was the R160, for which the base order cost was $1.25 billion for 620 cars (source, PDF-p. 34), or about $110,000 per meter of length. Commuter rail was similar, about $2 million per 25-meter-long M7 in the early 2000s and $760 million for 300 M8s of the same length in the mid-2000s. London’s then-current order, the S Stock, cost £1.5 billion for 191 trains and 1,395 cars, around $90,000 per meter of length for narrower trains; Paris’s MP 05, a driverless rubber-tired train, cost €474 million for 49 trainsets, around $140,000 per meter.

But since then, costs have rapidly risen. The gap is still far smaller than that for infrastructure, which New York builds for an order of magnitude higher cost than the rest-of-world median. But it’s no longer a rounding error. Subway rolling stock costs are rising, and commuter rail rolling stock are rising even faster. The latest subway order, the R211, costs $1.45 billion for 535 cars, or $150,000 per meter, for the base order, and $3.69 billion for 1,612 cars, or $130,000 per meter, including options. Commuter rail equipment costs, once about $100,000 per meter of train length, inched up to $2.7 million per car in 2013, or $110,000 per meter, and then rose to $150,000 per meter for the M9 order.

Construction costs: subway trains

The 2020-4 capital plan has showcased even further rolling stock cost escalation. Go to the link for the MTA capital plan again. On PDF-p. 23 there’s a breakdown of different items on the subway, and rolling stock is $6.057 billion for a total of 1,977 cars, of which 900 are 15 meters long and the rest (I believe) 18, for a total of $185,000 per linear meter.

I’ve blogged before about comparative costs of light rail and regional rail rolling stock. In Europe, both still cluster around $100,000 per linear meter for single-level, non-high-speed equipment. There is no apparent premium over early- and mid-2000s cost even without adjusting for inflation, which is not surprising, as the real prices of manufactured goods tend to fall over time. But what about metros? Here, too, we can look at first-world world comparisons.

In London, a recent Piccadilly line order is, in exchange rate terms, $190,000/meter (the trains are 103 m long) – but it includes 40 years of maintenance and spare parts. In Singapore, a recent order is S$2.1 million per car, which is about $70,000 per meter in exchange rate terms. Grand Paris Express’s first tranche of orders costs €1.3 billion for 183 trains totaling 948 cars, each (I believe) 15 meters long, around $120,000 per meter. Metro Report states Busan’s recent order as 55.6 billion for 48 trainsets (replacing 140-meter long trains), which is almost certainly an error; assuming the actual cost is 556 billion, this is $70,000/meter in exchange rate terms and $90,000/meter in PPP terms (PPP is relevant as this is an entirely domestic order).

In Berlin, the situation is the diciest, with the highest costs outside New York (not counting London’s maintenance-heavy contracts). An emergency order of 20 52-meter trains, tendered because cracks were discovered in the existing trains, cost €120 million, around $150,000 per linear meter. A longer-term contract to supply 1,500 cars (some 13 meters long, most 16.5 meters long) for €3 billion by 2035 is on hold due to litigation: Siemens had already sued over the emergency order of Stadler cars, but now Alstom made its own challenge. But even here, costs are well below the levels of New York, even before we adjust for inflation since Berlin’s future contract is in 2020-35 prices and New York’s is in in 2020-24 prices.

Construction costs: New York-area commuter rail

Commuter rail is faring even worse. On PDF-p. 27 the LIRR is listed as spending $242 million on 17 coaches and 12 locomotives, and on PDF-p. 29 Metro-North is listed as spending $853 million on 80 EMU cars and 30 locomotives.

Figuring out exact comparisons is not easy, because locomotives do cost more than multiple-units and unpowered coaches, and there is a range of locomotive costs, with uncertainty due to currency conversions, as most information I can find about European locomotives is in Eastern Europe with its weak currencies, since Western Europe mostly uses multiple-units. Railway Gazette’s pages on the world rolling stock market suggest that a European locomotive is around €5 million (e.g. the PKP Vectron order), or $6.5 million; PKP’s domestic order (including some dual-modes) is around $4.2 million per unit measured in exchange rate terms, but twice as much in PPP terms; Bombardier has a sale to an undisclosed customer for about $4.8 million. Siemens claims the Vectron costs €2.5 million per unit, although all the contracts for which I can find prices are substantially more expensive.

For what it’s worth, in the US dual-mode locomotives for New Jersey Transit cost around $9.5 million apiece, which is still evidently lower than what the LIRR and Metro-North plan on spending. 242 – 9.5*12 = 128, and 128/17 = 7.5, or $300,000 per linear meter of unpowered coach; similarly, 853 – 9.5*30 = 568, and 568/80 = 7.1, or $280,000 per linear meter of new Metro-North EMU. If we take the normal-world cost of a locomotive at $6 million and that of an EMU or coach at $2.5 million per US-length car, then the LIRR has a factor-of-2.1 cost premium and Metro-North a factor-of-2.2 premium.

The equipment is conservative

The FRA recently realigned its regulations to permit lightly-modified European mainline trains to run on American tracks. Nonetheless, no American commuter rail operator has taken advantage of the new rules – the only ones buying European equipment had plans to do so even before the revision, going through costly waiver process that increased costs. At a public meeting last month, Metro-North’s vice president of engineering did not even know FRA rules had changed. The LIRR and Metro-North are buying the same equipment, to the same standards, as they have for decades.

The subway, likewise, is conservative. It is a laggard in adopting open gangways: the R211 order is the first one to include any, but that is just two test trainsets, the rest having doors between cars like all other older New York trainsets. It is not buying any of the modular products of the global vendors, like Bombardier’s Movia platform or the Alstom Metropolis. It is buying largely the same kind of equipment it has bought since the 1990s.

Despite this conservatism, costs are very high, consistent with a factor somewhat higher than 2 on commuter rail and somewhat lower than 2 on the subway.

But perhaps the conservatism is what increases costs in the first place? Perhaps the reason costs are high is that the world market has moved on and the MTA and some other American operators have not noticed. In Chicago, Metra found itself trying to order a type of gallery car that nobody makes any longer, using parts that are no longer available. Perhaps the same kind of outmoded thinking is present at the MTA, and this is why costs have exploded in the last 10 years.

A secular increase in costs of infrastructure construction is nearly universal. No such trend can be seen in rolling stock: nominal costs in Paris are 15% lower than they were 15 years ago, and real costs are about 30% lower, whereas in New York nominal costs are 70% higher than 10 years ago and real costs about 40% higher. Paris keeps innovating – M1 and M14 have the highest frequency of any metro system in the world, a train every 85 seconds at the peak, and M1 is the first driverless line converted from earlier manual operations rather than built from scratch. In contrast, New York is stuck in the 1990s, but far from keeping a lid on costs, it has seen rolling stock cost explosion.

Update 9/24: I just saw a new commuter rail coach order in Boston. These are bilevels so some cost premium is to be expected, but $345 million for 80 unpowered coaches, or $170,000 per meter, is excessive, and TransitMatters tried hard to fight against this order, arguing in favor of EMUs on the already-electrified Providence Line.

Cars-and-Trains Urbanism

For all of the rhetoric about banning cars and the inherent conflict between public transportation and private automobiles, the dominant political view of urbanism in large chunks of the world is the cars-and-trains approach. Under this approach, cities build extensive infrastructure for cars, such as parking, wide arterials, and some motorways, as well as for trains, which are as a rule always rapid transit, never streetcars. In the midcentury developed world this was the unanimous view of urban development, and this remains the preference of mainline center-right parties like CDU, the French Republicans, and the British and Canadian Tories; various 1960s urbanist movements with roots in the New Left arose in specific opposition to much of that mentality, which is why those movements are usually NIMBY in general.

In the post-consensus environment of political conflict in most issues, in this case between auto- and transit-oriented urbanism, it’s tempting to go back to the midcentury elite consensus as a compromise, and call for making cities friendly to both transit users and drivers. This is attractive especially to people who hope to defuse culture war issues, either because they identify as political moderates or because they identify as socialists and have some nostalgia for the Old Left. However, this kind of urbanism does not really work. While a destination can sometimes be friendly to both drivers and transit users, the city overall cannot be; the majority of the points of interest in a successful transit city are hostile to cars and vice versa.

Moreover, this cars-and-transit failure is not just historical. It keeps going on today. Middle-income countries waste vast sums of money on building two separate transportation networks that do not work well together. The United States, too, has adopted this mentality in the cities that are building new light rail lines, resulting in large urban rail systems whose ridership is a rounding error since most of the city isn’t oriented around public transportation.

What is cars-and-trains urbanism?

Postwar West Germany built a number of subway networks in its large cities, such as Munich, Frankfurt, Cologne, Dortmund, Essen, and Hanover. With the exception of Munich and Nuremberg, these are subway-surface systems, in which the trains are underground in city center but run in streetcar mode farther out. For the most part, these systems were built with the support of the driver lobby, which wanted the streetcars out of city center in order to be able to drive more easily, and once those systems opened, the cities dismantled the streetcars. Most of West Germany thus eliminated the streetcars that did not feed into the tunnels, just as the US eliminated nearly all of its streetcars except the ones that were part of a subway-surface system in Boston, Philadelphia, and San Francisco.

In the United States, such development only happened in San Francisco, where Muni buried the main streetcar trunk in conjunction with the construction of BART along the same alignment on Market Street. More commonly, cars-and-trains urbanism led to the development of park-and-rides in the suburbs. An early example is the Green Line D branch in Boston, designed for suburban commuters rather than urban residents using the line for all purposes and not just work. Subsequently, light rail lines have been built with park-and-rides, as have full rapid transit systems in the suburb of Atlanta, Washington, and San Francisco. In the same period, American mainline rail networks evolved to be car-oriented, replacing city center stations with park-and-rides for commuter as well as intercity rail uses.

American cars-and-trains development was not without conflict. The auto lobby opposed trains, believing buses were cheaper; top civil servants in what is now the Federal Highway Administration advocated for bus lanes to create more capacity at the peak into city centers such as Washington’s. However, the trains that were built in this era followed the same mentality of creating more peak capacity in areas where widening roads was too expensive because of high city center land prices.

In the US as well as in Europe, and nowadays in developing countries, construction of rapid transit in the biggest cities and high-speed rail between them is paired with large highway systems for everything else. When the Tories won the 2010 election, they proclaimed the end of Labour’s so-called war on motorists, but maintained their support for Crossrail in London and High Speed 2 from London to the major provincial cities. And in Toronto, even Rob and Doug Ford, for all their anti-walkability demagogy, support subways, just not at-grade streetcars that would take lanes away from cars.

How does cars-and-trains transportation fail?

In the United States, public transportation is divided into three groups. There is transit-oriented urbanism, which covers about half to two thirds of New York, and very small segments of Chicago, Boston, San Francisco, Washington, and Philadelphia. There are people riding public transportation out of poverty. And there is cars-and-trains behavior, common in the outer parts and suburbs of cities with urban rail networks. In the major American metropolitan areas with urban rail other than New York, people who commute by public transport actually outearn people who drive alone, because so much transit ridership consists of rich suburban commuters. Because of the weight of those commuters and because American metro areas with public transportation are richer than the rest of the country, the national gap in income between drivers and transit commuters is small and shrinking. And yet, fuel consumption as a proportion of overall consumption is constant around 3.5% in the bottom nine deciles.

In other words: the United States has spent a lot of money on attracting the rich to public transportation, and has succeeded in the sense that transit commuters earn about the same as car commuters, but the rich still drive so much that they consume as much fuel as the poor relative to their total spending. This is not because rich people inherently like driving – rich Manhattanites don’t drive much. This is because the postwar American transportation network does not provide adequate public transportation for non-commute trips. Off-peak frequencies are low, and service to destinations outside city centers is weak.

In Germany, the politics of cars-and-trains infrastructure is still around. A few months ago, when some Berlin Greens proposed congestion pricing, CDU came out in opposition, saying that without park-and-rides, how can people be expected to use the U- and S-Bahn? Walking or biking to the station is apparently not possible in outer Berlin, per CDU.

How does cars-and-trains urbanism fail?

The problem with cars-and-trains urbanism is not just about lack of frequency. The off-peak frequency on some of the American light and heavy rail systems serving park-and-rides is not terrible for regional rail – trains come every 10 or 12 or 15 minutes. But the development repels non-commuter uses of the system. The stations are surrounded by parking rather than high-density office or residential development. People who already own cars will drive them wherever it’s convenient: they’ll shop by car since retail has no reason to cluster in the central business district, and they’ll probably drive to jobs that do not have such agglomeration benefits as to have to be in city center.

That is not just an American problem. Western Europe, too, has built extensive infrastructure to extend auto-oriented postwar suburbia into older city centers, including motorways and parking garages. If the streets are narrow, then the sidewalks may be extremely narrow, down to maybe a meter in Florence. This encourages anyone who can afford to do so to drive rather than walk.

If there is no transit-oriented core to the city, then the result is a standard auto-oriented city. Examples include Los Angeles and Dallas, both of which have large urban rail networks with approximately no ridership. In the three-way division of American transit ridership – New York (and to a small extent a handful of other city cores), suburban commuters, very poor people – Los Angeles’s transit ridership is mostly very poor, averaging half the income of solo drivers. Public transit construction in this case has been a complete waste without policies that create a transit city, which must include both liberalization (namely, zoning liberalization near stations) and coercion (such as higher car and fuel taxes and removal of parking).

If there is a transit-oriented core, then the result cleaves the metro area in two. To people who live in the transit zone, the auto-oriented parts are inaccessible, and vice versa. A few places at the boundary can be crosshatched, but the city itself cannot be entirely crosshatched – the sea of single-family houses in the suburbs is not accessible except by car, and transit-oriented cities have no room for the amount of parking or road capacity required for auto-centric density.

Does rapid transit mean cars-and-trains?

No. In opposition to the postwar elite consensus and the center-right’s support of cars-and-trains urbanism, the New Left tends to be hostile to rapid transit, on the theory that it’s only good for cars and that tramways with dedicated lanes are as good as subways. This theory is hogwash – enough cities built metros before mass motorization in order to avoid streetcar and horsecar traffic jams – but it’s attractive to people who associate subways with the failings of CDU and its equivalents in other countries.

Paris provides a positive example of rejecting cars-and-trains urbanism while building rapid transit. Postwar France was thoroughly cars-and-trains in its mentality, but 21st-century Paris is the opposite. Mayor Anne Hidalgo has narrowed roadways and removed freeways in order to make the city pedestrian-friendlier. Ile-de-France is expanding its tramway network, but it’s at the same time investing enormous amounts of money in expanding the Metro and RER. I do not think there is any city outside China with more underground route-km built than Paris in 2000-30 – Indian metros are mostly above-ground. In my under-construction database, which largely omits China and Russia due to difficulties of finding information in English, Grand Paris Express is 10% of the total route-length.

Postwar Japan is another example of rapid transit without cars-and-trains typology. Unlike present-day Paris, which is ideologically leftist and green, Japanese development has been in an ideological environment similar to the center-right elite consensus, called dirigism in France. Nonetheless, Tokyo’s motorway network is not large relative to the city’s population, and suburban development has been quite dense and rail-oriented. The private rail operators have preferred to build high-density housing at their suburban stations to encourage more ridership, rather than park-and-rides.

It’s one or the other

Drivers are most comfortable on high-speed arterial streets with generous shoulders and setbacks, with parking right next to their destinations. This encourages dispersal – just try building parking for all the jobs of Midtown Manhattan or Central Tokyo on-site. Pedestrians would need to walk long distances along noisy, polluted streets and cross them at inconvenient signal times or places or risk being run over. Public transit users fare little better, as they turn into pedestrians at their destination – and what’s more, public transportation requires destinations to cluster at a certain density to fill a train at a usable frequency.

This situation works in reverse in a transit city. On a robust public transportation network, the most desirable locations are in the very center of the city, or at key interchanges. Usually the density at those nodes grows so high that drivers have to contend with heavy traffic. Widening roads is not possible at reasonable cost in dense centers of economic production; the very reason for cars-and-trains urbanism as opposed to just 100% cars is that it was never economic to build 20-lane highways in city centers.

On the street, too, conflict is inevitable. A lane can be shared, which means dominated by cars so long as a car with one person inside it gets the same priority as a bus or tram with 40; or it can be dedicated to buses and trams, which means cars have less space. And then there are pedestrians, who need adequate sidewalks even in historic city centers where the street width from building to building is 10 meters rather than the more modern 30.

Defusing conflict is attractive, but this is not possible. A city cannot be friendly to drivers and to non-drivers at the same time. The urban designs for the two groups are too different, and for the most part what most appeals to one repels the other. Trying to build two redundant transportation networks may be attractive to people who just like the idea of visible development with its construction jobs, but both will end up underused and overly costly. Good transit has to convert drivers into non-drivers – sometimes-drivers are too expensive to serve, because the urbanism for them is too peaky and expensive.

As a corollary of this, political structures that have to give something to drivers too have to be eliminated if public transportation is to succeed. For example, infrastructure funding formulas that give set amounts of money to the two modes, like the 80% cars, 20% transit split of American federal funding, are bad and should ideally be reduced to 0 if the formula itself cannot be changed; the investment in highways is making public transportation less useful, both through direct competition and through incentives for auto-oriented development. The same is true of schemes that are really fronts for highway widening, like some bus rapid transit in the US and India. Good transit activists have to oppose these, even if it means less money in overall spending, even if it means less money in spending specific for some public transit programs. The cost of highways is just too high to try to maintain a culture truce.

Circumferential Lines and Express Service

In a number of large cities with both radial and circumferential urban rail service, there is a curious observation: there is express service on the radial lines, but not the circumferential ones. These cities include New York, Paris, and Berlin, and to some extent London and Seoul. Understanding why this is the case is useful in general: it highlights guidelines for urban public transport design that have implications even outside the distinction between radial and circumferential service. In brief, circumferential lines are used for shorter trips than radial lines, and in large cities connect many different spokes so that an express trip would either skip important stations or not save much time.

The situation

Berlin has three S-Bahn trunk lines: the Ringbahn, the east-west Stadtbahn, and the North-South Tunnel. The first two have four tracks. The last is a two-track tunnel, but has recently been supplemented with a parallel four-track North-South Main Line tunnel, used by regional and intercity trains.

The Stadtbahn has a straightforward local-express arrangement: the S-Bahn uses the local tracks at very high frequency, whereas the express tracks host less frequent regional trains making about half as many stops as well as a few intercity trains only making two stops. The north-south system likewise features very frequent local trains on the S-Bahn, and a combination of somewhat less frequent regional trains making a few stops on the main line and many intercity trains making fewer stops. In contrast, the Ringbahn has no systemic express service: the S-Bahn includes trains running on the entire Ring frequently as well as trains running along segments of it stopping at every station on the way, but the only express services are regional trains that only serve small slivers on their way somewhere else and only come once or twice an hour.

This arrangement is mirrored in other cities. In Paris, the entire Metro network except Line 14 is very local, with the shortest interstations and lowest average speeds among major world metro systems. For faster service, there is Line 14 as well as the RER system, tying the suburbs together with the city. Those lines are exclusively radial. The busiest single RER line, the RER A, was from the start designed as an express line parallel to Line 1, the Metro’s busiest, and the second busiest, the RER B, is to a large extent an express version of the Metro’s second busiest line, Line 4. However, there is no RER version of the next busiest local lines, the ring formed by Lines 2 and 6. For non-Metro circumferential service, the region went down the speed/cost tradeoff and built tramways, which have been a total success and have high ridership even though they’re slow.

In New York, the subway was built with four-track main lines from the start to enable express service. Five four-track lines run north-south in Manhattan, providing local and express service. Outside the Manhattan core, they branch and recombine into a number of three- and four-track lines in Brooklyn, Queens, and the Bronx. Not every radial line in New York has express service, but most do. In contrast, the circumferential Crosstown Line, carrying the G train, is entirely local.

In Seoul, most lines have no express service. However, Lines 1, 3, and 4 interline with longer-range commuter rail services, and Lines 1 and 4 have express trains on the commuter rail segments. They are all radial; the circumferential Line 2 has no express trains.

Finally, in London, the Underground has few express segments (all radial), but in addition to the Underground the city has or will soon have express commuter lines, including Thameslink and Crossrail. There are no plans for express service parallel to the Overground.

Is Tokyo really an exception?

Tokyo has express trains on many lines. On the JR East network, there are lines with four or six tracks all the way to Central Tokyo, with local and express service. The private railroads usually have local and express services on their own lines, which feed into the local Tokyo subway. But not all express services go through the primary city center: the Ikebukuro-Shibuya corridor has the four-track JR Yamanote Line, with both local services (called the Yamanote Line too, running as a ring to Tokyo Station) and express services (called the Saikyo or Shonan-Shinjuku Line, continuing north and south of the city); Tokyo Metro’s Fukutoshin Line, serving the same corridor, has a timed passing segment for express trains as well.

However, in three ways, the area around Ikebukuro, Shinjuku, and Shibuya behaves as a secondary city center rather than a circumferential corridor. The job density around all three stations is very high, for one. They have extensive retail as well, as the private railroads that terminated there before they interlined with the subway developed the areas to encourage more people to use their trains. This situation is also true of some secondary clusters elsewhere in Tokyo, like Tobu’s Asakusa terminal, but Asakusa is in a historically working-class area, whereas the Yamanote area was historically and still is wealthier, making it easier for it to attract corporate jobs.

Second, from the perspective of the transportation network, they are central enough that railroads that have the option to serve them do so, even at the expense of service to Central Tokyo. When the Fukutoshin Line opened, Tokyu shifted one of its two mainlines, the Toyoko Line, to connect to it and serve this secondary center, where it previously interlined with the Hibiya Line to Central Tokyo; Tokyu serves Central Tokyo via its other line, the Den-en-Toshi Line, which connects to the Hanzomon Line of the subway. JR East, too, prioritizes serving Shinjuku from the northern and southern suburbs: the Shonan-Shinjuku Line is a reverse-branch of core commuter rail lines both north and south, as direct fast service from the suburbs to Shibuya, Shinjuku, and Ikebukuro is important enough to JR East that it will sacrifice some reliability and capacity to Tokyo Station for it.

Third, as we will discuss below, the Yamanote Line has a special feature missing from circumferential corridors in Berlin and Paris: it has distinguished stations. A foreigner looking at satellite photos of land use and at a map of the region’s rail network without the stations labeled would have an easy time deciding where an express train on the line should stop: Ikebukuro, Shinjuku, and Shibuya eclipse other stations along the line, like Yoyogi and Takadanobaba. Moreover, since these three centers were established to some extent before the subway was built, the subway lines were routed to serve them; there are 11 subway lines coming from the east as well as the east-west Chuo Line, and of these, all but the Tozai and Chiyoda Lines intersect it at one of the three main stations.

Interstations and trip length

The optimal stop spacing depends on how long passenger trips are on the line: keeping all else equal, it is proportional to the square root of the average unlinked trip. The best formula is somewhat more delicate: widening the stop spacing encourages people to take longer trips as they become faster with fewer intermediate stops and discourages people from taking shorter ones as they become slower with longer walk distances to the station. However, to a first-order approximation, the square root rule remains valid.

The relevance is that not all lines have the same average trip length. Longer lines have longer trips than short lines. Moreover, circular lines have shorter average trips than straight lines of the same length, because people have no reason to ride the entire way. The Ringbahn is a 37-kilometer line on which trains take an hour to complete the circuit. But nobody has a reason to ride more than half the circle – they can just as well ride the shorter way in the other direction. Nor do passengers really have a reason to ride over exactly half the circle, because they can often take the Stadtbahn, North-South Tunnel, or U-Bahn and be at their destinations faster.

Circumferential lines are frequently used to connect to radial lines if the radial-radial connection in city center is inconvenient – maybe it’s missing entirely, maybe it’s congested, maybe it involves too much walking between platforms, maybe happens to be on the far side of city center. In all such cases, people are more likely to use the circumferential line for shorter trips than for longer ones: the more acute the angle, the more direct and thus more valuable the circle is for travel.

The relevance of this discussion to express service is that there’s more demand for express service in situations with longer optimum stop spacing. For example, the optimum stop spacing for the subway in New York based on current travel patterns is the same as that proposed for Second Avenue Subway, to within measurement error of parameters like walking speed; on the other trunk lines, the local trains have denser stop spacing and the express trains have wider stop spacing. On a line with very short optimum spacing, there is not much of a case for express service at all.

Distinguished stops versus isotropy

The formula for optimal stop spacing depends on the isotropy of travel demand. If origins and destinations are distributed uniformly along the line, then the optimal stop spacing is minimized: passengers are equally likely to live and work right on top of a station, which eliminates walk time, as they are to live and work exactly in the middle between two stations, which maximizes walk time. If the densities of origins and destinations are spiky around distinguished nodes, then the optimal stop spacing widens, because planners can place stations at key locations to minimize the number of passengers who have to walk longer. If origins are assumed to be perfectly isotropic but destinations are assumed to be perfectly clustered at such distinguished locations as city center, the optimum stop spacing is larger than if both are perfectly isotropic by a factor of \sqrt{2}.

Circumferential lines in large cities do not have isotropic demand. However, they have a great many distinguished stops, one at every intersection with a radial rail service. Out of 27 Ringbahn stops, 21 have a connection to the U-Bahn, a tramway, or a radial S-Bahn line. Express service would be pointless – the money would be better spent increasing local frequency, as ridership on short-hop trips like the Ringbahn’s is especially sensitive to wait time.

On the M2/M6 ring in Paris, there are 49 stops, of which 21 have connections to other Metro lines or the RER, one more doesn’t but really should (Rome, with a missed connection to an M14 extension), and one may connect to a future extension of M10. Express service is not completely pointless parallel to M2/M6, but still not too valuable. Even farther out, where the Paris region is building the M15 ring of Grand Paris Express, there are 35 stops in 69 kilometers of the main ring, practically all connecting to a radial line or located at a dense suburban city center.

The situation in New York is dicier, because the G train does have a distinguished stop location between Long Island City and Downtown Brooklyn, namely the connection to the L train at Bedford Avenue. However, the average trip length remains very short – the G misses so many transfers at both ends that end-to-end riders mostly stay on the radials and go through Manhattan, so the main use case is taking it a few stops to the connection to the L or to the Long Island City end.


A large urban rail network should be predominantly radial, with circumferential lines in dense areas providing additional connectivity between inner neighborhoods and decongesting the central transfer points. However, that the radial and circumferential lines are depicted together on the same metro or regional rail map does not mean that people use them in the same way. City center lies ideally on all radials but not on the circumferentials, so the tidal wave of morning commuters going from far away to the center is relevant only to the radials.

This difference between radials and circumferentials is not just about service planning, but also about infrastructure planning. Passengers make longer trips on radial lines, and disproportionately travel to one of not many distinguished central locations; this encourages longer stop spacing, which may include express service in the largest cities. On circumferential lines, they make shorter trips to one of many different connection points; this encourages shorter stop spacing and no express service, but rather higher local frequency whenever possible.

Different countries build rapid transit in radically different ways, and yet big cities in a number of different countries have converged on the same pattern: express service on the strongest radial corridors, local-only service on circumferential ones no matter how busy they are. There is a reason. Transportation planners in poorer cities that are just starting to build their rapid transit networks as well in mature cities that are adding to their existing service should take heed and design infrastructure accordingly.

The Future is not Retro

One faction of urbanists that I’ve sometimes found myself clashing with is people who assume that a greener, less auto-centric future will look something like the traditional small towns of the past. Strong Towns is the best example I know of of this tendency, arguing against high-rise urban redevelopment and in favor of urbanism that looks like pre-freeway Midwestern main streets. But this retro attitude to the future happens everywhere, and recently I’ve had to argue about this with the generally pro-modern Cap’n Transit and his take about the future of vacations. Even the push for light rail in a number of cities has connections with nostalgia for old streetcars, to the point that some American cities build mixed-traffic streetcars, such as Portland.

The future was not retro in the 1950s

The best analogy for a zero-emissions future is ironically what it seeks to undo: the history of suburbanization. In retrospect, we can view midcentury suburbanization as a physical expansion of built-up areas at lower density, at automobile scale. But at the time, it was not always viewed this way. Socially, the suburbs were supposed to be a return to rural virtues. The American patrician reformers who advocated for them consciously wanted to get rid of ethnic urban neighborhoods and their alien cultures. The German Christian democratic push for regional road and rail connections has the same social origin, just without the ethnic dimension – cities were dens of iniquity and sin.

At the same time, the suburbs, that future of the middle of the 20th century, were completely different from the mythologized 19th century past, before cities like New York and Berlin had grown so big. Most obviously, they were linked to urban jobs; the social forces that pushed for them were aware of that in real time, and sought transportation links precisely in order to permit access to urban jobs in what they hoped would be rural living.

But a number of other key differences are visible – for one, those suburbs were near the big cities of the early 20th century, and not in areas with demographic decline. In the United States, the Great Plains and Appalachia kept depopulating and the Deep South except Atlanta kept demographically stagnating. The growth in that era of interregional convergence happened in suburbs around New York, Chicago, and other big then-industrial cities, and in parts of what would soon be called the Sunbelt, namely Southern California, Texas, and Florida. In Germany, this history is more complicated, as the stagnating region that traditionalists had hoped to repopulate was Prussia and Posen, which were given to Poland at the end of the war and ethnically cleansed of their German populations. However, we can still see postwar shifts within West Germany toward suburbs of big cities like Munich and Frankfurt, while the Ruhr stagnated.

The future of transit-oriented development is not retro

People who dislike the auto-oriented form of cities can easily romanticize how cities looked before mass motorization. They’d have uniform missing middle built form in most of the US and UK, or uniform mid-rise in New York and Continental Europe. American YIMBYs in particular easily slip into romanticizing missing middle density and asking to replace single-family housing with duplexes and triplexes rather than with anything more substantial.

If you want to see what 21st-century TOD looks like, go to the richer parts of East Asia, especially Tokyo, which builds much more housing than Hong Kong and Singapore. The density in Tokyo is anything but uniform. There are clusters of high-rise buildings next to train stations, and lower density further away, even small single-family houses fronting narrow streets far enough from train stations that it’s not economical to redevelop them. It offends nostalgic Westerners; the future often does.

In the context of a growing city like New York or London, what this means is that the suburbs can expect to look spiky. There’s no point in turning, say, everything within two kilometers of Cockfosters (or the Little Neck LIRR station) into mid-rise apartments or even rowhouses. What’s the point? There’s a lot more demand 100 meters from the station than two kilometers away, enough that people pay the construction cost premium for the 20th floor 100 meters from the stations in preference to the third floor two kilometers away. The same is true for Paris – there’s no solution for its growth needs other than high-rises near RER stations and key Metro stations in the city as well as the suburbs, like the existing social housing complexes but with less space between buildings. It may offend people who associate high-rises with either the poor or recent high-skill immigrants, but again, the future often offends traditionalists.

The future of transportation is not retro

In countries that do not rigidly prevent urban housing growth the way the US does, the trend toward reurbanization is clear. Germany’s big cities are growing while everything else is shrinking save some suburbs in the richest regions, such as around Munich. Rural France keeps depopulating.

In this context, the modes of transportation of the future are rapid transit and high-speed rail. Rapid transit is preferable to buses and surface trains in most cities, because it serves spiky development better – the stations are spaced farther apart, which is fine because population density is not isotropic and neither is job density, and larger cities need the longer range that comes with the higher average speed of the subway or regional train over that of the tramway.

High-speed rail is likewise preferable to an everywhere-to-everywhere low-speed rail network like that of Switzerland. In a country with very large metro areas spaced 500 km or so apart, like the US, France, or Germany, connecting those growing city centers is of crucial importance, while nearby cities of 100,000 are of diminishing importance. Moreover, very big cities can be connected by trains so frequent that untimed transfers are viable. Already under the Deutschlandtakt plan, there will be 2.5 trains between Berlin and Hanover every hour, and if average speeds between Berlin and the Rhine-Ruhr were increased to be in line with those of the TGVs, demand would fill 4-6 trains per hour, enough to facilitate untimed transfers from connecting lines going north and south of Hanover. The Northeast Corridor has even more latent demand, given the huge size of New York.

The future of travel is not retro

The transportation network both follows and shapes travel patterns. Rapid transit is symbiotic with spiky TOD, and high-speed rail is symbiotic with extensive intercity travel.

The implication is that the future of holidays, too, is not retro. Vacation trips between major cities will become easier if countries that are not France and Japan build a dense network of high-speed lines akin to what France has done over the last 40 years and what Japan has done over the last 60. Many of those cities have thriving tourism economies, and these can expect to expand if there are fast trains connecting them to other cities within 300-1,000 kilometers.

Sometimes, these high-speed lines could serve romanticized tourist destinations. Niagara Falls lies between New York and Toronto, and could see expansion of visits, including day trips from Toronto and Buffalo and overnight stays from New York. The Riviera will surely see more travel once the much-delayed LGV PACA puts Nice four hours away from Paris by train rather than five and a half. Even the Black Forest might see an expansion of travel if people connect from high-speed trains from the rest of Germany to regional trains at Freiburg, going from the Rhine Valley up to the mountains; but even then, I expect a future Germany’s domestic tourism to be increasingly urban, probably involving the Rhine waterfront as well as the historic cities along the river.

But for the most part, tourist destinations designed around driving, like most American national parks as well as state parks like the Catskills, will shrink in importance in a zero-carbon future. It does not matter if they used to have rail access, as Glacier National Park did; the tourism of the leisure class of the early 20th century is not the same as that of the middle class of the middle of the 21st. Grand Canyon and Yellowstone are not the only pretty places in the world or even in the United States; the Hudson Valley and the entire Pacific Coast are pretty too, and do not require either driving or taking a hypothetical train line that, on the list of the United States’ top transportation priorities, would not crack the top 100. This will offend people whose idea of environmentalism is based on the priorities of turn-of-the-century patrician conservationists, but environmental science has moved on and the nature of the biggest ecological crisis facing humanity has changed.

The non-retro future is pretty cool

The theme of the future is that, just as the Industrial Revolution involved urbanization and rural depopulation, urban development patterns this century involve growth in the big metro areas and decline elsewhere and in traditional small towns. This is fine. The status anxieties of Basil Fawlty types who either can’t or won’t adapt to a world that has little use for their prejudices are not a serious public concern.

Already, people lead full lives in big global cities like New York and London without any of the trappings of what passed for normality in the middle of the 20th century, like a detached house with a yard and no racial minorities or working-class people within sight. The rest will adapt to this reality, just as early 20th century urbanites adapted to the reality of suburbanization a generation later.

It’s not even an imposition. It’s opportunity. People can live in high-quality housing with access to extensive social as well as job networks, and travel to many different places with different languages, flora and fauna, vistas, architecture, food, and local retail. Even in the same language zone, Northern and Southern Germany look completely different from each other, as do Paris and Southern France, or New England and Washington. Then outside the cities there are enough places walking distance from a commuter rail line or on the way on a high-speed line between two cities that people can if they’d like go somewhere and spend time out of sight of other people. There’s so much to do in a regime of green prosperity; the world merely awaits the enactment of policies that encourage such a future in lieu of one dominated by small-minded local interests who define themselves by how much they can pollute.

What is Light Rail, Anyway?

I’ve been asked on Twitter about the differences between various kinds of urban rail transit. There is a lot of confusion about the term light rail in English, since it can be used for urban public transport typologies that have little to do with one another. The best way to think about urban rail (other than regional rail) is to use the following schema:

Slow in center Fast in center
Slow in outlying areas Tramway Subway-surface
Fast in outlying areas Tram-train Rapid transit


In American parlance, all four have been called light rail: subway-surface and tram-train lines are always called light rail, and officially so are tramways; then one full rapid transit line, the Green Line in Los Angeles, is called light rail as it runs light rail vehicles (LRVs) rather than subways. Nonetheless, in this post I will ignore what things are called and focus on their speed.

In this context, fast and slow refer to right-of-way quality. A tramway in a low-density city with little traffic and widely separated stops may well be faster than a rapid transit line with many stops, such as most Paris Metro lines, but relative to the local urban typology the tramway is still slow while the metro is fast.

The two hybrid forms – subway-surface and tram-train – differ in where they focus higher-speed service. On a subway-surface line, the city center segment is in a subway and then the line branches farther out, for examples the Boston Green Line, San Francisco Muni Metro, Philadelphia Subway-Surface Lines, and Frankfurt and Cologne U-Bahn networks. On a tram-train, the train is fast outside city center, where it runs in a dedicated surface right-of-way, but then in city center it runs in tramway mode on the street at lower speed; the Karlsruhe tram-train is one such example, as are virtually all postwar light rail systems in the United States and Canada.


The 2*2 typology simplifies the situation somewhat. There exist lines that don’t fully obey it, and instead change between metro and streetcar mode haphazardly. Some of the Cologne lines go back and forth. Buffalo has a single light rail line without branches, dubbed the Buffalo Metro Rail, running on the surface in the center and in a greenfield tunnel farther out toward Amherst and the university campus. Frankfurt’s U1/2/3/8 trunk is the opposite of Buffalo, running in a tunnel in the center and on the surface farther out even downstream of the branch point. The Los Angeles Blue Line is underground at Metro Center but then runs on the surface, transitions to a grade-separated right-of-way later, and finally drops back to streetcar mode in Downtown Long Beach.

The most fascinating case is that of the Boston Green Line D branch. It is technically rapid transit, since the trunk line is in a tunnel alongside the other branches whereas the branch itself is a former commuter rail line; it is called light rail because it runs LRVs, like the Los Angeles Green Line, and shared the trunk with the B, C, and E branches, all of which have surface segments. But conceptually, it presages most proper American light rail lines: it was built in the 1950s as suburban-oriented rapid transit, with park-and-rides and downtown-focused service, creating a paradigm that postwar metros like BART and the Washington Metro would sometimes follow and that light rail systems from the 1980s onward (San Diego, Portland, etc.) always would.

Nonetheless, such aberrations are uncommon enough that the 2*2 simplification works when explaining what cities should be building.

Right-of-way availability

Cities are more likely to build fast trains when there is preexisting right-of-way for them. The Karlsruhe Zweisystem is based on using the area’s extensive legacy mainline network, on which LRVs run in train mode, and then diverging toward city center in streetcar mode. Jarrett Walker has a good post about Karlsruhe specifically: there is no good right-of-way with which to drag the Stadtbahn into city center in train mode, and thus the alternative to a tram-train is an expensive tunnel; such a tunnel is under construction now, at the cost of about €1 billion, but as Karlsruhe is a small city, it comes a generation after the tram-train system was put into place.

North American light rail systems often use mainline rail corridors as well, but thanks to federal regulations as well as weak regional rail systems, they almost never use mainline tracks; the Blue Line in San Diego, the first tram-train in the United States, is one of very few exceptions, and even then it shares track with a very lightly-used freight line, rather than with a frequent S-Bahn as in Karlsruhe. It is more common for North American tram-trains to run in disused corridors, on new tracks parallel to the mainline, or even in highway medians.

Reusing legacy rail lines and running in freeway medians are not unique to tram-trains. Rapid transit does both outside city center; the first subway network in the world, the London Underground, makes extensive use of branches of former commuter lines, and even shares track with a still-active one on a portion of the Watford DC Line. New York, likewise, connected former excursion lines in Brooklyn to the subway, forming most of the Coney Island-bound system, and later did the same with the LIRR in the Rockaways, now carrying branches of the A train. It is usually easy to spot whether an urban rail line descends from a legacy branch line – if it does then it is very unlikely to follow a single street (none of the lines serving Coney Island does), whereas if it doesn’t then it is usually a subway or el on a major arterial (such as Fourth Avenue in Brooklyn).

The upshot is that cities are likelier to build tram-trains and rapid transit in preference to tramways and subway-surface lines if they have high-quality right-of-way. New York and London were unlikely to build subway-surface lines in the early 20th century either way, but the high density of their metro networks in Southern Brooklyn and West London respectively can be explained by the extent of preexisting legacy lines in these areas. Comparable areas that did not have such good connections, for example Queens, have much less rapid transit coverage.

While this issue in theory affects tram-trains and rapid transit equally, in practice it is especially relevant to tram-trains. Rapid transit is more expensive, so it is likely to be built in larger and denser cities, where it is more acceptable to just tunnel under difficult segments. Tram-trains are present in smaller cities – Calgary, Edmonton, Karlsruhe, and so on – as well as in American Sunbelt cities that are so auto-oriented that they have the public transport of European cities one third or even one tenth their size. In those cities, tunneling is harder to justify, so the train goes where it can go cheaply. Downtown transit malls like those of Portland and Calgary are the least bad solution for connecting fast lines from the suburbs to provide better city center coverage and connect to lines on the other side of the region.

Subway-surface branching

Subway-surface lines are fast in city center and slow outside of it. Moreover, in city center their right-of-way segregation (in a tunnel in all of the American cases) means there is more capacity than on the surface. This makes branching especially attractive. Indeed, in all three American cases – Boston, Philadelphia, San Francisco – the subway-surface line has four to five branches.

Outside the United States, subway-surface branching is more complicated. In Frankfurt, the U4/5 and U6/7 lines work as in the United States, but with only two branches per trunk rather than four or five; but the U1/2/3 line has a surface segment on the mainline. In Cologne, there is extensive reverse-branching (see map), and while most of the system runs in subway-surface mode, one line runs in tramway mode through city center but then drops to a tunnel in Deutz and splits into two surface branches farther east.

Tel Aviv is building a subway-surface line from scratch, without any branching. The Red Line is to run underground in Central Tel Aviv, Ramat Gan, and Bnei Brak, and on the surface farther east in Petah Tikva as well as at the other end in Jaffa and Bat Yam. At the Petah Tikva end, an underground connection to the depot is to enable half the trains to terminate and go out of service without running on the surface; at the Jaffa/Bat Yam end, a loop near the portal is to enable half the trains to terminate and reverse direction without running on the surface.

The American way works better than the incipient Israeli way. The main advantage of branching is that the greater expanse of land in outer-urban neighborhoods and suburbs means more lines are needed than in the center to guarantee the same coverage. Thus Downtown San Francisco has just one line under Market Street, serving not just Muni Metro but also BART on separate tracks, but in the rest of the city, BART and the five branches of Muni serve an arc of neighborhoods from the Mission to the Sunset. The lack of branching on the Tel Aviv Red Lines means that it will not be able to serve Petah Tikva well: the city is not very dense or very central and has no hope of getting the multiline crisscross pattern eventually planned for Central Tel Aviv.

One implication of the fact that subway-surface lines should branch is that they are more appropriate for cities with natural branching than for cities without. Boston in particular is an excellent place for such branching. Its street network does not form a grid, but instead has arterials that are oriented around the historic city center; the Green Line makes use of two such streets, Commonwealth Avenue hosting the B branch and Beacon Avenue hosting the C branch. A light rail line following Washington Street could likewise branch to Warren Street and Blue Hill Avenue and potentially even branch farther out on Talbot Avenue to Ashmont, effectively railstituting the area’s busiest buses.

In contrast, cities whose street networks don’t lend themselves well to branching should probably not build subway-surface lines. North American cities with gridded street networks have little reason to use this technology. If they are willing to build downtown tunnels and have the odd right-of-way running toward city center diagonally to the grid, they should go ahead and build full rapid transit, as Chicago did on the Blue Line of the L.

Speed and range

Tramways are the cheapest variety of urban rail and metro tunnels are the most expensive. The reason cities don’t just build tramways in lieu of any grade separation is that tramways are slow and therefore have limited range. Berlin’s tramways average around 16 km/h; they run partly in mixed traffic, but I don’t think they can cross 20 km/h even with dedicated lanes and signal priority.

What this means is that tramways are mainly a solution for city centers and near-center neighborhoods. The tramways in Berlin work okay within the Ring, especially in U-Bahn deserts like the segment of East Berlin between U2 and U5. But in suburban Paris, they’re too slow to provide the full trip and instead work as Metro and RER feeders, providing circumferential service whereas the faster modes provide radial rail transport.

Tramway-centric transit cities can work, but only in a constrained set of circumstances:

  • They must be fairly small, like Karlsruhe, Strasbourg, or Geneva.
  • They should have a network of sufficiently wide streets (minimum 20-25 meters including sidewalks, ideally 30-35) through city center as well as radiating out of it.
  • They should have a supplementary regional rail network for longer trips.

Tramway-and-regional-rail is a powerful combination. Zurich is based on it, having rejected a subway network in two separate referendums. However, once the city grows beyond the size class of Strasbourg, the regional rail component begins to dominate, as there are extensive suburbs that are just too far away from city center for streetcars.

Upgrading to rapid transit

It’s common for cities to replace light rail with rapid transit by building new tunnels and burying the tracks. Historically, Boston and San Francisco both built their subway-surface networks by incrementally putting segments in tunnel, which would later protect these lines from replacement by diesel buses. Stockholm and Brussels both incrementally upgraded streetcars to metro standards, calling the intermediate phase pre-metro. Karlsruhe is building a tunnel for its Stadtbahn.

However, in the modern era, not all such tunneling projects are equally useful. Subway-surface lines stay subway-surface indefinitely: they have so much surface branching that the cost of putting everything underground would be prohibitive. San Francisco activists have flirted with a plan to replace one Muni Metro surface line with rapid transit and then reduce the rest to tramways with forced transfers; this plan is both terrible and unlikely to happen. Tel Aviv might eventually come to its senses and bury the entire Red Line, but this is possible only because the current branch-free layout is already more suited for a subway than for a subway-surface system.

Tram-trains are easier to convert to rapid transit. All that’s needed is a short tunnel segment in city center. Thus, in addition to the Karlsruhe tunnel project, there are serious discussions of city center tunneling in a variety of North American cities, including Portland and Calgary (in the near term) as well San Diego (on the 2050 horizon).

Finally, tramways can be upgraded to full rapid transit more easily than to either of the two intermediate forms. A good tramway is rarely a good subway-surface system, because the subway-surface system ideally branches and the pure tramway ideally does not. Moreover, a good tramway is unlikely to go very far into the suburbs because of its low speed, whereas a tram-train’s ability to leverage high speed in train mode allows it to go deep into the suburbs of Karlsruhe, Calgary, or San Diego. The optimal place for a tramway – dense city neighborhoods following a single line – is also the optimal one for a metro line, making the upgrade more attractive than upgrading the tramway to a hybrid.

New Report on Construction Costs Misses the Mark

In the last few years, ever more serious and powerful actors have begun investigating the fact of high American infrastructure construction costs. First it was Brian Rosenthal’s excellent New York Times exposé, and then it was the Regional Plan Association’s flop of a study. At the same time, I was aware that the congressional Government Accountability Office, or GAO, was investigating the same question, planning to talk to sources in the academic world as well as industry in order to make recommendations.

The GAO report is out now, and unfortunately it is a total miss, for essentially the same reason the RPA’s report was a miss: it did not go outside the American (and to some extent rest-of-Anglosphere) comfort zone. Its literature review is if anything weaker than the RPA’s. Its interviews with experts are telling: out of nine mentioned on PDF-p. 47, eight live in English-speaking countries. Even when more detailed information about non-English-speaking countries is readily available, even in English, the GAO report makes little use of it. It is a lazy study, and people who ideologically believe the American federal government does not work should feel confident citing this as an example.

Cost comparisons

Brian himself already notes one of the reasons the report is so weak: Congress mandated a comparative study, but the report made no international comparisons at all. Instead, the report offered this excuse (PDF-p. 27):

The complexity of rail transit construction projects and data limitations, among other things, limits the ability to compare the costs of these projects, according to the stakeholders we interviewed. As highlighted above, each project has a unique collection of specific factors that drive its costs. According to FTA officials, each proposed transit project has its own unique characteristics, physical operating environment, and challenges. Some stakeholders said that the wide disparity in the relative effect of different cost factors renders cost comparisons between projects difficult. For example, representatives of an international transit organization said that because of the large number of elements that can affect a project’s costs and the differences in what costs are included in different projects’ data, projects should be compared only at a very granular level and that aggregate cost comparisons, such as between the costs per mile or costs per kilometer of different projects, are likely flawed. Some stakeholders also said that project costs should not be compared without considering the projects’ contexts, such as their complexity. For example, one academic expert contended that project costs cannot be compared without considering the context of each project, and that analysis of projects should focus on leading practices and lessons learned instead.

There is a big problem with the above statement: disaggregated costs for many aspects of urban rail construction do exist. The Manhattan Institute’s Connor Harris has done a lot of legwork comparing tunnel boring machine staffing levels and wages in New York and in Germany, and found that New York pays much higher wages but also has much higher staffing levels, 25-26 workers compared with 12. I have done some work looking at station costs specifically, and at the cost of installing elevators for wheelchair accessibility.

There is a lot of detailed comparative research about the costs of high-speed rail; the report even references one such meta-study undertaken within Europe, but omits the study’s analysis of causes of cost differences and instead asserts that it shows that comparing different projects is hard. In the interim, California contracted Deutsche Bahn to do a post-mortem of its elevated high-speed rail costs, which found that California needlessly built larger structures than necessary, explaining its cost premium over Germany.

Instead of probing these disaggregated estimates, the GAO preferred to say that they are too hard and move on.

Sanity checks

Even without disaggregation, there are some good sanity checks one can make about construction costs. The most important is that big projects – major subway expansions, regional rail tunnels, high-speed rail – cost an appreciable amount of the government’s budget. The budget for the 200-kilometer Grand Paris Express project is €35 billion, plus another €3 billion in contributions for related suburban rail extensions such as that of the RER E. There may be future cost overruns, but they will be reported in the media, just as the current overrun has been; it is extremely difficult to hide cost overruns measured in tens of billions in a Paris-size city, and even in a China-size country it may not be easy.

Is it plausible that GPX is inherently easier to build than New York’s $1+ billion/km subway tunnels? Yes. It’s equally plausible that it is inherently harder. Second Avenue Subway runs under a wide, straight throughfare, a situation that simplifies construction. In Israel, the ministry of transportation has long mentioned the ease of tunneling under wide, straight boulevards in connection with plans to extend the second line of the Tel Aviv subway to North Tel Aviv under Ibn Gabirol, and admitted this even when it opposed the extension on land use grounds.

The most important sanity check is that in a world with several dozens of cities with a wide variety of wealth levels, land use patterns, geologies, and topographies, no city has managed to match or even come close to New York’s construction costs. New York is not special enough to be an edge case in all or even most relevant geographic variables – it is dense but no denser than Seoul or Paris, it is wealthy but no wealthier than London or Paris or Munich and barely wealthier than Stockholm, it has hard rock but less hard than Stockholm (and in Stockholm the gneiss is cited as a cost saver – bored tunnels do not require concrete lining), etc.

Moreover, the cities that have the highest construction costs outside New York are almost without exception in the same set of countries: the US, Canada, Britain, Singapore, Australia. What’s likelier – that there is some special geographic feature common to the entire Anglosphere (including Quebec) but absent from all other developed countries, or that there is a shared set of legal and political traditions that developed in the last 50 years that impede cost-effective construction? Instead of probing this pattern, the GAO preferred to wash its hands and refuse to compare projects across countries.

Internal comparisons

In lieu of making international comparisons, the GAO has engaged in extensive internal comparison. It cites aspects that have raised the costs of Second Avenue Subway above other American subway projects, such as overdesign for stations. Apparently, it’s completely legit to compare two different cities’ construction if they’re in the same country.

Over and over again, it references its own domestic standards. The GAO has 12 design standards, e.g. on PDF-pp. 51-52 and 56-60; the report mentions that existing cost estimation methodologies by the Federal Transit Administration, or FTA, meet 7 of them; thus, it exhorts transit agencies to meet the other 5 standards.

The only problem is that there is no evidence supplied that those design standards are really useful. After all, the United States has very high costs, so why should anyone trust its standards? Even domestically, the report makes no effort to bring up successful examples of low overall costs coming from following prescribed standards. Seattle recently opened a light rail tunnel built for around $400 million per kilometer, a cost that would get most European project managers fired but that is still the lowest for an American urban rail tunnel built in this century. But the report never brings up Seattle at all, never mind that New York would salivate over the prospect of tunneling at Seattle’s cost.

The real internal comparisons then are not between different cities in the United States. Rather, they’re between different stages of cost estimation for the same project. There is published literature on cost overruns, most famously by Bent Flyvbjerg and his research group. The report cites Flyvbjerg. Moreover, one of the nine academic experts it consulted is Don Pickrell, who published a seminal paper on American cost overruns and ridership shortfalls in 1990. Pickrell was influential enough that a 2009 review found that not only had cost and ridership projections improved greatly in the intervening two decades, but also there was an improvement in ridership estimate quality attributable to Pickrell’s paper.

The GAO report is not the best source on cost overruns, but it is not completely useless there. Unfortunately, it remains useless when it comes to discussing absolute costs, a different topic from relative increases. Flyvbjerg’s original paper found that the US did not have higher cost overruns than Europe; but absolute costs in the US are several times as high. Flyvbjerg’s paper found that urban rail has higher cost overruns than road projects; but when a rail tunnel and a road tunnel are built in the same city, the road tunnel is more expensive by a factor of 1.5-2.5, at least in the four-city pilot I reported in 2017, owing to the need to build bigger bores with ventilation to carry heavy car traffic.

Lazy analysis, lazy synthesis

Americans who think of themselves as reformers like to point out real problems to solve, but then propose solutions that they made up without any connection with their analysis. The RPA study is one such example: even though one of its sources (namely, former Madrid Metro CEO Manuel Melis Maynar’s writeup about low Spanish costs) explicitly calls for separation of design and construction, its recommendations include a greater reliance on design-build. The same design-build recommendation appeared in a 2008 report in Toronto comparing the costs of the Sheppard subway, opened in 2002, with those of subways in Madrid; construction costs in Toronto have since tripled, while those of Madrid have barely risen.

To the GAO report’s credit, it does not recommend design-build. It even mentions the biggest drawback of design-build: it shifts cost risk to the private contractor, who compensates by demanding more public money up front. Nonetheless, it does not follow through and does not make the correct recommendation on this subject – namely, that cities and states should cease using this approach. It buries a recommendation for in-house expertise alongside a fad for peer review of projects.

Instead of lazily proposing design-build, the GAO lazily proposes two barely relevant tweaks (PDF-p. 43):

  • The FTA administrator should ensure that FTA’s cost estimating information for project sponsors is consistent with all 12 steps found in GAO’s Cost Estimating and Assessment Guide and needed for developing reliable cost estimates.
  • The FTA Administrator should provide a central, easily accessible source with all of FTA’s cost estimating information to help project sponsors improve the reliability of their cost estimates.

In other words, the report makes no recommendation about how to reduce costs, only about how to tell the public in advance that costs will be unaffordably high.

Why are these reports so bad?

This is not the first time a serious group releases an incurious study of American construction costs. What gives?

I suspect the answer has to be a combination of the following problems:

  • Reform factions often have a lot of internal ideas about how to improve things based on what they already know. They will cite new information if they feel like they must do so to save face, but they will not let new evidence change their conclusions. A little knowledge can be dangerous.
  • Finding information from outside the US, especially outside the English-speaking world, puts Americans (and Canadians) at a disadvantage. They know few to no foreigners, have little experience with cities abroad except as tourists, and do not speak foreign languages. Even when machine translation is decently accurate, which it is in the engineering literature in European languages, they are intimidated by the idea of dealing with non-English material. The process of learning is humbling, and some people prefer to remain proud and ignorant.
  • Open-ended analysis does not always lend itself to easy explanations or easy solutions. Even when solutions do present themselves, they may not flatter the people in power. Ten years ago I did not think senior management at American transit organs should be fired; today I think mass layoffs of the top brass, especially the political appointees, are somewhere between very useful and essential.

All three problems interact. For example, senior management is even less likely to be multilingual than junior staffers, who may be second-generation immigrant heritage speakers of a foreign language; thus, anything relying on foreign material disempowers the high-ups in favor of up-and-comers. The quick-and-easy-and-wrong solutions reformists seize upon if they find a little bit of knowledge let outfits like the GAO feel more powerful without actually challenging any obstructive politician or interest group, and if those solutions fail, they can always keep churning reports about implementation.

Last year, I did not know whether the GAO was capable of providing a blueprint for improving American infrastructure at lower cost. I assumed good faith because I had no reason not to. With this report, it is clear to me as well as to other observers of American public transit that the GAO is not so capable. Instead of doing what was in the country’s best interest, the people who commissioned and wrote the report delivered the minimal product that would get them kudos from superiors who do not know any better. They could have learned, or made a serious effort to learn, but that might challenge their assumptions or those of the high political echelons, and thus they preferred to say nothing and propose to do nothing.