Push and Pull Factors and Measuring Modal Shift

There’s a longstanding debate among activists and academics about what the best way of effecting modal shift from cars to public transport is. Pull factors concern making public transport better through building more rail lines, running them more frequently, improving service convenience, or reducing fares. Push factors concern making driving harder through speed limits, fuel taxes, congestion pricing, and reallocation of street space from cars to public and non-motorized transport. There’s a tendency on the New Left to favor push factors (but the East Asian developmental states are best characterized as push-before-pull and not pure pull).

This has been refined by researchers at the climate research institute, the Ariadne Project, who published a paper in late 2021 rating various push and pull policies on effectiveness for reducing transport emissions. They conclude that push factors dominate, and pull factors are small, with construction of new public transit almost insignificant, only worth a reduction of around 300,000 tons of CO2 a year Germany-wide, 0.039% of national emissions as of 2021; instituting a 120 km/h speed limit on the Autobahn is said to have about 10 times that effect, while the biggest effects yet would come from carbon taxes. The study laments that pull factors are so much more popular than push factors, which they admit suppress society-wide consumption.

The research suffers from the same problem as other work in this direction, in that it is bad at estimating the impact of public transport on mode shift. It briefly argues that construction of public transport increases overall consumption and therefore doesn’t do much to reduce emissions. This way, it’s like 2020’s carbon critique of U-Bahn expansion, which I criticized two months ago; the carbon critique argues that each kilometer of U-Bahn built only reduces CO2 emissions by 714 tons a year through mode shift, under the assumption that only 20% of public transport riders are diverted from cars.

This doesn’t pass a sanity check. 300,000 divided by 714 is 420 km, which is about comparable to the total route length of the four grade-separated U-Bahn systems in Germany plus the Wuppertal Schwebebahn; I think the two figures, 300,000 and 714/km, come from different sources, and judging by the other elements in the study, I suspect 300,000 assumes less construction than a full doubling of Germany’s rapid transit network length. Nonetheless, even under a more generous assumption, this is far too low compared with macro trends in public transport usage.

The best way to use macro trends as a sanity check is to look at some cases with much more and much less public transport than the present. Do they look like it’s a total difference of 0.039%? No, and that’s even taking into account that transit cities tend to be wealthier, stimulating more consumption and more production. As I pointed out in my post two months ago, while Germany averages 9.15 t-CO2/capita, Berlin only does 5.38, and while Germany averages 580 cars per 1,000 people, Berlin only does 327. The difference is largely about Berlin’s pull factors. Push factors in the city are not extensive, and what exists is implemented only in areas that already have very low car use.

Even lower household emissions in Berlin must be viewed as downstream of the density that is enabled by the presence of a large urban rail network. Cars are a low-capacity mode of transport, so an auto-oriented region, like American metro regions, has to spread out its homes and destinations to limit congestion, and this increases household emissions (single-family houses emit more than apartment buildings) and also encourages people to travel longer distances for their commute and routine non-commute trips.

This is not easy to measure. Public transport projects have gotten fairly good in the last generation at estimating ridership, but estimating the responsibility of one particular project to modal shift is hard. It interacts with the entire city region. For example, building one rail line can be measured to shift modes in the neighborhoods it serves, but it also encourages destinations to locate in city center since people from the neighborhoods the line serves can now access it, and the increase in office, retail, and community development then leads to a small modal shift citywide. Worse, trying to tease out the effect of the rail line on modal shift sufficiently carefully may lead researchers to count this citywide effect negatively, since one econometric technique is to compare the neighborhoods near the line with neighborhoods in the same city not on the line.

In practice, the construction of rail lines tends to co-occur with other policies that improve public transport, which may be pull or push factors. This means that it’s very easy to misattribute the effect of urban rail expansion to those other factors. I am convinced that this is what is happening here; the proper comparison must be at the level of an entire region, looking at the emissions of different regions with different levels of public transport usage.

The upshot is that if it is hard to measure the effect of public transport construction on modal shift and emissions, then the uncertain factors should not be set to zero. Rather, they should be set to sanity-check levels. For example, one can compare New York with the rest of the United States, since it’s a starker difference between a transit city and an auto-oriented country than anywhere in Europe, and correct for non-transport effects like climate and electricity mix, both of which are easy to measure.

Within Germany, Berlin has 42% lower emissions than the rest of the country per capita. Berlin achieves this with an urban rail network that, in 2019, got 1,289 million rail trips, nearly all within the city of 3.7 million, a minority in the suburban region of perhaps 1.3 million. This is around 250 trips/person regionwide, and 320/person citywide assigning around 20% of S-Bahn ridership to suburbs like Potsdam and Oranienburg. What’s more, Germany doesn’t start from zero; this is not the United States, with multiple large cities with around 10 annual rail trips per capita. Netting out buses from VDV’s data (p. 25) gets around 6.3 billion rail trips in Germany in 2019 including trams, or 75 per capita.

The difference between 320 and 75 is around 250 – I know it’s actually 245 but at this point I’m deliberately reducing precision because those are sanity-check estimates and I don’t want people thinking they’re correct to three significant figures (try 1.5). If we attribute the entire Berlin-Germany difference of about 3.8 t-CO2/capita to public transport and downstream changes to the urban layout, then we get 0.015 t saved per annual trip generated. To get from there to 300,000 tons saved, we just need 20 million annual rail riders, or around 65,000 daily ones, which is easy to generate on a single line; the approximately 2 km extension of U8 to Märkisches Viertel that Berlin keeps postponing is estimated to generate 25,000-30,000.

Now, to sanity-check the sanity check, the estimate here is that every trip on urban rail saves 15 kg-CO2. This is an aggressive figure; new cars nowadays average 100 g/km and averaged 180 g/km in 2001 (source, PDF-p. 15), and the average displaced car trip is not 150 km or even 80 km – Americans average around 45 km/day, or somewhat more when only adults are considered. Rather, the issue is a combination of factors:

  • Because the limiting factor to car transport is capacity, in practice what happens in an auto-oriented region is that it fills from the inside outward, and any modal shift ends up displacing the outermost and longest car trips. I proposed a model for that in a blog post from four years ago.
  • Public transport displaces car trips on a more than one-to-one basis (and certainly more than 20% as in the carbon critique of the U-Bahn). This is because public transport users also walk and bike, and transit cities have high modal splits for active transport by the standards of auto-oriented cities, if not by the standards of Dutch cities. Berlin’s all-trip modal split in 2018 was 26% car, 27% public transport, 18% bike, 30% walking – and the high active transport modal split exists not because of road diets, which are few and far between, but because of the presence of a large core fed by the U- and S-Bahn.
  • Public transport reduces household energy usage by encouraging people to live in apartment buildings with shared walls rather than in single-family houses, which have much greater heating requirements; this is also the mechanism through which transit cities have relatively high usage of active transport even without trying very hard.

I don’t think these factors fully explain away the gap between 45 km/day and 150 km per trip (so around 300/day), but they explain a large enough fraction of it that the installation of a system like what Berlin has – or, better, what Tokyo has – should be a climate priority. If your model says it doesn’t, it needs a lot more work than to just talk about the consumption effects of more public transport (if you’re bothered by how Berlin is poor for its size, compare New York with the rest of the United States).

In fact, if estimating modal shift is hard, then it’s best to approximate it by ridership. It’s imperfect because there is the effect of walking and biking; some lines really do just compete with walking, like city-center streetcars, but usually, to first order, it’s a good enough estimate. If it’s hard to estimate the benefits then they should not be set to zero, but rather set proportionally to something easier to measure, in this case ridership. Investment should follow ridership-maximizing strategies, and only deviate from them in corner cases.

Paris-Berlin Trains, But no Infrastructure

Yesterday, Bloomberg reported that Macron and Scholz announced new train service between Paris and Berlin to debut next year, as intercity rail demand in Europe is steadily rising and people want to travel not just within countries but also between them. Currently, there is no direct rail service, and passengers who wish to travel on this city pair have to change trains in Frankfurt or Cologne. There’s just one problem: the train will not have any supportive infrastructure and therefore take the same eight hours that trains take today with a transfer.

This is especially frustrating, since Germany is already investing in improving its intercity rail. Unfortunately, the investments are halting and partial – right now the longest city pair connected entirely by high-speed rail is Cologne-Frankfurt, a distance of 180 km, and ongoing plans are going to close some low-speed gaps elsewhere in the system but still not create any long-range continuous high-speed rail corridor connecting major cities. With ongoing plans, Cologne-Stuttgart is going to be entirely fast, but not that fast – Frankfurt-Mannheim is supposed to be sped up to 29 minutes over about 75 km.

Berlin-Paris is a good axis for such investment. This includes the following sections:

  • Berlin-Halle is currently medium-speed, trains taking 1:08-1:16 to do 162 km, but the flat, low-density terrain is easy for high-speed rail, which could speed this up to 40-45 minutes at fairly low cost since no tunnels and little bridging would be required.
  • Halle-Erfurt is already fast, thanks to investments in the Berlin-Munich axis.
  • Erfurt-Frankfurt is currently slow, but there are plans to build high-speed rail from Erfurt to Fulda and thence Hanau. The trip times leave a lot to be desired, but newer 300 km/h trains like the Velaro Novo, and perhaps a commitment to push the line not just to Hanau but closer to Frankfurt itself, could do this section in an hour.
  • Frankfurt-Saarbrücken is very slow. Saarbrücken is at the western margin of Germany and is not significant enough by itself to merit any high-speed rail investment. Between it and Frankfurt, the terrain is rolling and some tunneling is needed, and the only significant intermediate stops are Mainz (close enough to Frankfurt it’s a mere stop of opportunity) and Kaiserslautern. Nonetheless, fast trains could get from Frankfurt to the border in 45 minutes, whereas today they take two hours.

Unfortunately, they’re not talking about any pan-European infrastructure here. Building things is too difficult, so instead the plan is to run night trains – this despite the fact that Frankfurt-Saarbrücken with a connection to the LGV Est would make a great joint project.

Bad Public Transit in the Third World

There’s sometimes a stereotype that in poor countries with low car ownership, alternatives to the car are flourishing. I saw a post on Mastodon making this premise, and pointed out already in comments that this is not really true. This is a more detailed version of what I said in 500 characters. In short, in most of the third world, non-car transportation is bad, and nearly all ridership (on jitneys and buses) is out of poverty, as is most walking. While car ownership is low, the elites who do own cars dominate local affairs, and therefore cities are car-dominated and not at all walkable, even as 90%+ of the population does not own a car.

What’s more, the developing countries that do manage to build good public transportation don’t stay developing for long. The same development model of Japan, the East Asian Tigers, and now China has built both rail-oriented cities and high economic growth, to the point that Japan and the Tigers are fully developed, and China is a solidly middle-income economy. The sort of places that stay poor, or get stuck in a middle-income trap, also tend to have stagnant urban rail networks, and so grow more auto-oriented over time.

The situation in Southeast Asia

With the exception of Singapore, nowhere in Southeast Asia is public transit good. What’s more, construction costs have been high for elevated lines and very high for underground ones, slowing down the construction of metro systems.

In Kuala Lumpur and Bangkok, motorization is high and public transit usage is weak. Paul Barter’s thesis details how both cities got this way, in comparison with the more transit-oriented model used in Tokyo, Seoul, Hong Kong, and Singapore. The thesis also predicts that the poorer megacities of Southeast Asia – Jakarta and Manila – will follow the auto-oriented path as they develop, which has indeed happened in the 13 years since it was written.

The situation in those cities is, to be fair, murky. Manila has a large urban rail under construction right now, with average to above average costs for elevated lines and high ones for subways. But the system it has today consists of four lines, two branded light rail, one branded MRT, and one commuter line. In 2019, the six-month ridership on the system was 162 million. A total of 324 million in a metro area the size of Manila is extraordinarily low: the administrative Metro Manila region has 13.5 million people, and the urban or metropolitan area according to both Citypopulation.de and Demographia is 24-26 million. On the strictest definition of Metro Manila, this is 24 trips per person per year; on the wider ones, it is about 13, similar to San Diego or Portland and only somewhat better than Atlanta.

Jakarta is in the same situation of flux. It recently opened a half-underground MRT line at fairly high cost, and is modernizing its commuter rail network along Japanese lines, using second-hand Japanese equipment. Commuter rail ridership was 1.2 million a day last year, or around 360 million a year, already higher than before corona; the MRT had 20 million riders last year, and an airport link had 1.5 million in 2018. This isn’t everything – there’s also a short light metro called LRT for which I can’t find numbers – but it wouldn’t be more than second-order. This is 400 million annual rail trips, in a region of 32 million people.

The future of these cities is larger versions of Bangkok. Thailand is sufficiently middle-income that we can see directly how its transport system evolves as it leaves poverty, and the results are not good. Bus ridership is high, but it’s rapidly falling as anyone who can afford a car gets one; a JICA report about MRT development puts the region’s modal split at 5% MRT, 36% bus, and the rest private (PDF-p. 69) – and the income of bus riders is significantly lower than that of drivers (PDF-p. 229), whereas MRT riders are closer to drivers.

Even wealthier than Bangkok, with the same auto-oriented system, is Kuala Lumpur. There, the modal split is about 8% bus, 7% train, and the rest private. This is worse than San Francisco and the major cities of Canada and Australia, let alone New York or any large European city. The national modal split in England, France, Germany, and Spain is about 16% – the first three countries’ figures predate corona, but in Spain they’re from 2021, with suppressed public transport ridership. Note that rail ridership per capita is healthier in Kuala Lumpur than in Jakarta or Manila – all rail lines combined are 760,000 riders per day, say 228 million per year, in a region of maybe 7 million people. This is better than a no-transit American city like San Diego, but worse than a bad-transit one like Chicago or Washington, where the modal split is about the same but there is no longer the kind of poverty that is common in Malaysia, let alone in Indonesia, and therefore if people ride the trains it’s because they get them to their city center jobs and not because they’re poor.

Even in Singapore, the best example out there of a transit-oriented rich city, it took until very recently for MRT coverage to be good enough that people willingly depend on it; it only reached NUS after I graduated. In the 1990s, the epitome of middle-class Singaporean materialism was described as owning the Five Cs, of which one was a car; traffic suppression, a Paul Barter describes, has centered fees on cars, much more car purchase than car use (despite the world-famous congestion pricing system), and thus to those wealthy enough to afford cars, they’re convenient in ways they are not in Paris, Berlin, or Stockholm.

The situation in Africa

African countries between the Sahara and the Kalahari are all very poor, with low car ownership. However, they are thoroughly car-dominated.

From the outside, it’s fascinating to see how the better-off countries in that region, like Nigeria, are already imitating Southeast Asia. Malaysia overregulated its jitneys out of existence because they were messy and this bothered elites, and because it wanted to create an internal market for its state-owned automakers. Nigeria is doing the same, on the former grounds; to the extent it hasn’t happened despite years of trying, it’s because the state is too weak to do more than harass the drivers and users of the system.

It’s notable that the Lagos discourse about the evils of the danfo – they are noisy, they are polluting, they drive like maniacs – there is little attention to how cars create all the same problems, except at larger scale per passenger served. The local notables drive (or are driven); the people who they scorn as unwashed, overly fecund, criminal masses ride the danfo. Thanks to aggressive domination by cars and inattention to the needs of the non-driving majority, Lagos’s car ownership is high for how poor it is – one source from 2017 says 5 million cars in the state, another from 2021 says 6.5 million vehicles between the state and Kano State. The denominator population in the latter source is 27 million officially, but unofficially likely more; 200 vehicles per 1,000 people is plausible for Lagos, which to be clear is not much less than New York or Paris, on an order of magnitude lower GDP per capita. Tokyo took until about 1970 to reach 100 vehicles per 1,000 people, at which point Japan had almost fully converged with American GDP per capita.

This is not specific to Lagos. A cousin who spent some time in Kampala told me of the hierarchy on the roads: pedestrians fear motorcycles, motorcycles fear cars, cars fear trucks. There is no pedestrian infrastructure to speak of; a rapid transit system is still a dream, to the point that a crayon proposal that spread on Twitter made local media. That the vast majority of Ugandans don’t own cars doesn’t matter; Kampala remains dominated by the few who do.

Transit and development

I don’t think it’s a coincidence that the sort of developing countries that build successful urban rail systems don’t stay poor for long. Part of it is that public transportation is good for economic development, but that’s not most of it – the United States manages to be rich without it except in a handful of cities. Rather, I suspect the reason has to do with state capacity.

More specifically, the reason cities with 100-200 cars per 1,000 people are thoroughly dominated by cars is that those 10-20% drivers (or people who are driven) are the elites. Their elite status can come from any source – passive business income, landlordism, active business income, skilled professional work – but usually it tilts toward the traditional, i.e. passive. These groups tend to be incredibly anti-developmental: they own small businesses, sometimes actively and sometimes passively, and resent being made redundant through economies of scale. India has problems with economic dwarfism and informality, and this is typical of poor countries; if anything, India is better than most at developing a handful of big businesses in high-value added industries.

The upshot is that the sort of people who drive, and especially the sort of drivers who are powerful enough to effect local changes to get incremental upgrades to roads at the expense of non-drivers, are usually anti-developmental classes. The East Asian developmental states (and Singapore and Hong Kong, which share many characteristics with them) clamped down on such classes hard, on either nationalist or socialist grounds; Japan, both Koreas, and both Chinas engaged in land reform, with characteristic violence in the two socialist states and without it but still with forcible purchase in the three capitalist states. The same sort of state that can eliminate landlordism can also, as a matter of capital formation, clamp down on consumption and encourage personal savings, producing atypically low levels of motorization well into middle-income status. Singapore, whose elite consumption centers vacations out of the country, has managed to do so even as a high-income country – and even more normal Tokyo and Seoul have much higher rail usage and lower car usage than their closest Western analog, New York.

India is in many ways anti-developmental, but it does manage to grow. Its anti-developmentalism is anti-urban and NIMBY, but it is capable of building infrastructure. Its metro program has problems with high construction costs (but Southeast Asia’s are generally worse) and lack of integration with other modes such as commuter rail, which the middle class denigrates as only befitting poor people; but the Delhi Metro had 5.5 million daily riders just before corona, slightly behind New York in a slightly larger metro area, perhaps a better comparison than Jakarta and Manila’s San Diego.

It’s the slower-growing developing countries that are not managing to even build the systems India has, let alone East Asia. They don’t have high car use, but only because they are poor, and in practice, they are thoroughly car-dominated, and everyone who doesn’t have a car wants one. A rich country really is not one where even the poor have cars but where even the rich use public transportation – and those countries aren’t rich and don’t grow at rates that will make them rich.

High Costs are not About Precarity

I’ve seen people who I think highly of argue that high construction costs in the United States are an artifact of precarity. The argument goes that the political support for public transportation there is so flimsy that agencies are forced to buy political support by spending more money than they need. This may include giving in to NIMBY pressure to use costlier but less impactful (or apparently less impactful) techniques, to spread money around with other government agencies and avoid fighting back, to build extravagant and fancier-looking but less standardized stations, and so on. The solution, per this theory, is to politically support public transportation construction more so that transit agencies will have more backing.

This argument also happens to be completely false, and the solution suggested is counterproductive. In fact, the worst cost blowouts are for the politically most certain projects; Second Avenue Subway enjoyed unanimous support in New York politics.

Cost-effectiveness under precarity

Three projects relevant to our work at the Transit Costs Project have been done exceptionally cost-effectively in an environment of political uncertainty: the T-bana, the LGV Sud-Est, and Bahn 2000.

T-bana

The original construction of the T-bana was done at exceptionally low cost. We go over this in the Sweden report to some extent, but, in short, between the 1950s and 70s, the total cost of the system’s construction was 5 billion kronor in 1975 prices, which built around 100 km, of which 57% are underground. In PPP 2022 dollars, this is $3.6 billion, or $35 million/km, not entirely but mostly underground. This was low for the time: for example, in London, the Victoria line was $122 million/km and the Jubilee line was $172 million/km (source, p. 78), and Italian costs in the 1960s and 70s were similar, averaging $129 million/km before 1970.

The era of Social Democrat dominance in Swedish politics on hindsight looks like one of consensus in favor of big public projects. But the T-bana itself was controversial. When the decision was made to build it in the 1940s, Stockholm County had about 1 million people; at the time, metros were present in much larger cities, like New York, London, Paris, Berlin, and Tokyo, and it was uncertain that a city the size of Stockholm would need such a system. Its closest analog, Copenhagen, did not build such a system until the 1990s, when it was a metro region of 2 million. It was uncertain that Stockholm should need rapid transit, and there were arguments for and against it in the city. Nor was there any transit-first policy in postwar Sweden: urban planning was the same modernist combination of urban renewal, automobile scale, and tower-in-a-park housing, and outside Stockholm County, the Million Program projects were thoroughly car-oriented.

Construction costs in Sweden are a lot higher now than they were in the 1950s, 60s, and 70s. Nya Tunnelbanan is $230 million/km, compared with a post-1990s Italian average of $220 million; British costs have exploded in tandem, so that now the Underground extensions clock at $600 million/km. Our best explanation is that the UK adopted what we call the globalized system of procurement, privatizing planning functions to consultants and privatizing risk to contractors, which creates more conflict; the UK also has an unusually high soft cost factor. From American data (and not just New York) and some British data, I believe that the roughly 2.5 cost premium of the UK over Italy is entirely reducible to such soft costs, procurement conflict, risk compensation, and excessive contingency. And yet, Sweden itself, with some elements of the same globalized system, maintains a roughly Italian cost level, albeit trending the wrong way.

And today, too, the politics of rail expansion in Sweden are uncertain. There was controversy over both Citybanan and Nya Tunnelbanan, neither of which passed a cost-benefit analysis (for reasons that I believe impugn the cost-benefit analysis more than those projects); it was uncertain that either would be funded. Controversy remained over plans to build high-speed rail connecting Stockholm with Gothenburg and Malmö, and the newly-elected right-wing government just canceled them in order to prioritize investment in roads. Swedish rail projects today remain precarious, and have to justify themselves on cost and efficiency grounds.

LGV Sud-Est

Like nearly all other rich countries, France was hit hard by the 1973 oil crisis; economic growth there and in the US, Japan, and most of the rest of Western Europe would never be as high as it was between the end of WW2 and the 1970s (“Trente Glorieuses“). On hindsight, France’s response to the crisis models can-go governance, with an energy saving ad declaring “in France we don’t have oil, but we have ideas.” The French state built nuclear power plants with gusto, peaking around 90% of national electricity use – and even today’s reduced share, around 70%, is by a large margin the highest in the world. At the same time, it built a high-speed rail network, connecting Paris with most other provincial cities at some of the highest average speeds outside China between major cities, reaching about 230 km/h between Paris and Marseille and 245 km/h between Paris and Bordeaux; usage per capita is one of the highest in Europe and, measured in passenger-km, not too bad by East Asian standards.

But in fact, the first LGV, the LGV Sud-Est, was deeply controversial. At the time, the only high-speed rail network in operation was the Shinkansen, and while France learned more from Japan than any other European country (for example, the RER was influenced by Tokyo rail operations), the circumstances for intercity were completely different. SNCF had benefited from having done many of its own experiments with high-speed technology, but the business case was murky. SNCF had to innovate in running an open system, with extensive through-running to cities off the line, which Japan would only introduce in the 1990s with the Mini-Shinkansen.

Within the French state, the project was controversial. Anthony Perl’s New Departures details how there were people within the government who wanted to cancel it entirely as it was unaffordable. At the end, the French state didn’t finance the line, and required SNCF to find private loans on the international market, though it did guarantee those loans. It also delayed the line’s opening: instead of opening the entire line from Paris to Lyon in one go, it opened two-thirds of it on the Lyon side in 1981 and the last third into Paris in 1983, requiring trains to run on the classical line at low speed between Paris and Saint-Florentin for two years; in that era, phased opening was uncommon, and lines generally opened to the end at once, such as between Tokyo and Shin-Osaka.

Construction was extraordinarily inexpensive. In PPP 2022 dollars, it cost $8.4 million/km. This is, by a margin, the lowest-cost high-speed rail line ever built that I know about. The Tokaido Shinkansen cost 380 billion yen, or in PPP 2022 dollars $40 million/km, representing a factor of two cost overrun that forced JNR’s head to resign. Spain has unusually low construction costs, and even there, Madrid-Seville was $15.7 million/km. SNCF innovated in every way possible to save money. Realizing that high-speed trains could climb steeper grades, it built the LGV Sud-Est with a ruling grade of 3.5%, which has since become a norm in and around Europe, compared with the Shinkansen’s 1.5-2%; the line has no tunnels, unlike the classical Paris-Lyon line. It built the line on the ground rather than on viaducts, and balanced cut and fill locally so that material cut to grade the line could be used for nearby fill. Thanks to the line’s low costs and high ridership, the financial return on investment for SNCF has been 15%, and social return on investment has been 30% (source, pp. 11-12).

This cost-effectiveness would never recur. The line’s success ensured that LGV construction would enjoy total political backing. The core features of LGV construction are still there – earthworks rather than viaducts, 3.5% grades, limited tunneling, overcompensation of landowners by about 30% with land swap deals to defuse the possibility of farmer riots. But the next few lines cost about $20 million/km or slightly less, and this cost has since crept up to about $30 million/km or even more. This remains low by international standards (but not by Spanish ones), but the trend is negative.

SNCF is coasting on its success from a generation ago, secure that funding for LGVs and state support in political contention is forthcoming, and the routing decisions have grown worse. In response to NIMBYism in Provence, the French state assented to a tunnel-heavy route, including a conversion of Marseille from an at-grade terminal to an underground through-station, akin to Stuttgart 21, which has not been done before in France, and the resulting high costs have led to delays on the project. Operations have grown ever more airline-style, experimenting with low-cost airline imitation to the point of reducing fare receipts without any increase in ridership. One of the French consultants we’ve spoken with said that their company’s third-party design costs are 7-8% of the hard costs, which figure is similar to what we’ve seen in Italy and to the in-house rate in Spain – but the same consultant told us that there is so much bloat at SNCF that when it designs its own projects, the costs are not 7% but 25%, a figure in line with American rates.

Bahn 2000

Switzerland has Europe’s strongest passenger rail network by all measures: highest traffic measured by passenger-km per capita, highest modal split for passenger-km, highest traffic density. Its success is well-known in surrounding countries, which are gradually either imitate its methods or, in the case of Germany, pretending to do so. It has achieved its success through continuous improvement over the generations, but the most notable element of this system was implemented in the 1990s as part of the Bahn 2000 project.

The current system is based on a national-scale clockface system (“Takt”) with trains repeating hourly, with the strongest links, like the Zurich-Bern-Basel triangle, running every half hour. Connections are timed in those three cities and several others, called knots, so that trains enter each station a few minutes before the connection time (usually the hour) and depart a few minutes after, permitting passengers to get between most pairs of Swiss cities with short transfers. Reliability is high, thanks to targeted investments designed to ensure that trains could make those connections in practice and not just in theory. Further planning centers adding more knots and expanding this system to the periphery of Switzerland.

Switzerland is famous for its consensus governance system – its plural executive is drawn from the four largest parties in proportion to their votes, with no coalition vs. opposition politics. But the process that led to the decision to adopt Bahn 2000 was not at all one of unanimity. There had been plans to build high-speed rail, as there were nearly everywhere else in Western Europe. But they were criticized for their high costs, and there was extensive center-right pressure to cut the budget. Bahn 2000 was thus conceived in an environment of austerity. Many of its features were explicitly about saving money:

  • The knot system is connected with running trains as fast as necessary, not as fast as possible. Investments in speed are pursued only insofar as they permit trains to make their connections; higher speeds are considered gratuitous.
  • Bilevel trains are an alternative to lengthening the platforms.
  • Timed overtakes and meets are an alternative to more extensive multi-tracking of lines.
  • Investment in better timetabling and systems (the electronics side of the electronics-before-concrete slogan) is cheaper than adding tunnels and viaducts.

Swiss megaprojects have to go to referendum, and sometimes the referendums return a no; this happened with the Zurich U-Bahn twice, leading to the construction of the S-Bahn instead. All Swiss planners know in a country this small and this fiscally conservative, any extravagance will lead to rejection. The result is that they’ve instead optimized construction at all levels, and even their unit costs of tunneling are low; thanks to such optimization, Switzerland has been able to build a fairly extensive medium-speed rail system, with more tunneling per capita than Germany (let alone France), and with two S-Bahn trunk tunnels in Zurich, where no German city today has more than one.

The American situation

The worst offenders in the United States are not at all politically precarious. There is practically unanimous consensus in New York about the necessity of Second Avenue Subway. At no point was the project under any threat. There is an ideological right in the city, rooted less in party politics and more in the New York Post and the Manhattan Institute, with a law-and-order agenda and hostility to unions and to large government programs, but at no point did they call for cancellation; the Manhattan Institute’s Nicole Gelinas has proposed pension cuts for workers and rule changes reducing certain benefits, but not canceling Second Avenue Subway.

At intercity scale, the same is true of Northeast Corridor investment. The libertarian and conservative pundits who say passenger rail is a waste of money tend to except the Northeast Corridor, or at least its southern half. When the Republicans won the 2010 midterm election, the new chair of the House of Representatives Transportation Committee, John Mica (R-FL), proposed a bill to seek private concessionaires to run intercity rail on the corridor. He did not propose canceling train service, even though in the wake of the same election, multiple conservative governors canceled intercity rail investments in their state, both high-speed (Florida) and low-speed (Wisconsin, Ohio).

In fact, both programs – New York subway expansion and the Northeast Corridor – are characterized by continuity across partisan shifts, as in more established consensus governance systems. The Northeast Corridor is especially notable for how little role ideological or partisan politics has played so far. New York has micromanagement by politicians – Andrew Cuomo had his pet projects in Penn Station Access and the backward-facing LaGuardia air train, and now Kathy Hochul has hers in the Interborough Express – but Second Avenue Subway was internal, and besides, political micromanagement is a different problem from political precarity.

And neither of these programs has engaged in any cost control. To the contrary, both are run as if money is infinite. The MTA would surrender to NIMBYs (“good neighbor policy”) and to city agencies looking to extract money from it. It built oversize stations. It spent money protecting buildings from excessive settlement that have been subsequently demolished for redevelopment at higher density.

The various agencies involved in the Northeast Corridor, likewise, are profligate, and not for lack of political support. Connecticut is full of NIMBYs; one of the consultants working on the plan a few years ago told me there was informal pressure not to ruffle feathers and not to touch anything in the wealthiest suburbs in the state, those of Fairfield County. In fact, high-speed rail construction would require significant house demolitions in the state’s second wealthiest town, Darien – but Darien is so infamously exclusive (“Darien rhymes with Aryan,” say other suburbanites in the area) that the rest of the region feels little solidarity with it.

NIMBYs aside, there has not been any effort at coordinating the different agencies in the Northeast along anything resembling the Swiss Takt system. This is not about precarity, because this is not a precarious project; this is about total ignorance and incuriosity about best practices, which emanate from a place that doesn’t natively speak English and doesn’t trade in American political references.

The Green Line Extension

Boston’s GLX is a fascinating example of cost blowouts without precarity. The history of the project is that its first iteration was pushed by Governor Deval Patrick (D), with the support of groups that sued the state for its delays in planning the project in the 1990s, as a court-mandated mitigation for the extra car traffic induced by the construction of the Big Dig. Patrick instituted a good neighbor policy, in which everything a community group wanted, it would get. Thus, stations were to become neighborhood signatures, and the project was laden with unrelated investments, called betterments, like a $100 million 3 km bike lane called the Somerville Community Path.

At no point in the eight years Patrick was in power was there a political threat to the project. It was court-mandated, and the extractive local groups that live off of suing the government favored it. The Obama administration was generous with federal stimulus funding, and the designs were rushed in order to use stimulus funding to pay for the project’s design, which would be done by consultants rather than with an expansion of in-house hiring. It’s in this atmosphere of profligacy that the project’s cost exploded to, in today’s money, around $3.5 billion, for 7 km of light rail in existing rights-of-way (albeit ones requiring overpass reconstruction).

The project did fall under political threat, after Charlie Baker (R) won the 2014 election. Baker’s impact on Massachusetts governance is fascinating, in that he unambiguously cut its budget significantly in the short run, but also had both before (as budget director in the 1990s) and during (through his actions as governor) wrecked the state’s long-term ability to execute infrastructure, setting up a machine intended to privatize the state and avoiding any in-house hiring. Nonetheless, the direct impact of precarity on GLX was to reduce scope: the betterments were removed and the stations were changed from too big to too small. The final cost was $2.3 billion, or around $2.8 billion in 2022 dollars, and half of that was sunk cot from the previous iteration.

There was no expectation that the project would be canceled – indeed, it was not. A Republican victory was unexpected in a state this left-wing. Then, as Baker was taking office, past governors from both parties expressed optimism that he would get not just GLX done but also the much more complex North-South Rail Link tunnel. Nor did contractors have their contracts yanked unexpectedly, which would get them to bid higher for risk compensation. Baker cold-shouldered not jut NSRL but also much smaller investments in commuter rail, but at no point was anyone stiffed in the process. No: the Patrick-era project was just poorly managed.

Project selection

As one caveat, I want to point out a place where precarity does lead to poor project cost-effectiveness: the project selection stage, in the context of tax referendums, especially in Southern California. None of this leads to high per-kilometer costs – Los Angeles’s are exactly as bad as in the rest of the United States – but it does lead to poor project selection, as an unintended consequence of anti-tax New Right politics.

The California-specific issue is that raising taxes requires a referendum, with a two-thirds vote. Swiss referendums are by simple majority, or at most double majority – but in practice most referendums that have a popular vote majority, even a small one, also have a double majority when needed, and in the last 15 years I believe the only two times they didn’t had the double majority veto overturning a 1.5% margin and a 9% margin.

California’s requirement of a two-thirds majority was intended to stop wasteful spending and taxation, but has had the opposite effect. In and around San Francisco, the voters are sufficiently left-wing that two-thirds majorities for social policy are not hard to obtain. But in Southern California, they are not; to build public transportation infrastructure, Los Angeles and San Diego County governments cannot rely on ideology, but instead cut deals with local, non-ideological actors through promising them a piece of the package. Los Angeles measures for transit expansion are, by share of money committed, largely not about transit expansion but operations, new buses, or leakage to roads and bridges; then, what does go to transit expansion is divvied by region, with each region getting something, no matter how cost-ineffective, while core improvements are neglected and so are cross-regional connections (since the local extractive actors aren’t going to ride the trains and can’t tell a circumferential project is useful for them).

This is not a US-wide phenomenon. It’s not even California-wide: this problem is absent from the Bay Area, where BART decided against an expansion to Livermore that was unpopular among technically-oriented advocates, and would like to build more core capacity if it could do so for much less than a billion dollars per km. New York does not have it at all (for one, it doesn’t require two-thirds majorities for budgeting).

At intercity scale, this precarity does cause Amtrak to maximize how many states and congressional districts it runs money-losing daily trains in. But wasting money on night trains and on peripheral regions is hardly a US-only problem – Japan National Railways did that until well past privatization, when its successors spun off money-losing branch lines to prefecture-subsidized third sector railways. This is not at all why there is no plan for Northeastern intercity rail that is worth its weight in dirt: Northeastern rail improvements have been amply funded relative to objective need (if not relative to American costs), and solid investments in the core coexist with wastes of money on the periphery of the network in many countries.

The issue of politicization

The precarious, low-cost examples all had to cut costs because of fiscal pressure. However, in all three, the pressure did not include any politicization of engineering questions. Sweden was setting up a civil service modeled on the American Bureau of Public Roads, currently the Federal Highway Administration, which in the middle of the 20th century was a model of depoliticized governance. France and Switzerland have strong civil service bureaucracies – if anything SNCF is too self-contained and needs reorganization, just not if it’s led by the usual French elites or by people from the airline industry.

Importantly, low-cost countries with more clientelism and politicization of the state tend to be more deferential to the expertise of engineers. Greece has a far worse problem of overreliance on political appointees than the United States, let alone the other European democracies; but engineering is somewhat of an exception. Hispanic and Portuguese-speaking cultures put great prestige on engineering, reducing the extent of political micromanagement, even in countries without strong apolitical civil service bureaucracies. Even in Turkey, the politicization of public transportation is entirely at macro level: AKP promises to prioritize investment in areas that vote for it and has denied financing to the Istanbul Metro since the city flipped to the opposition (the city instead borrows money from the European Investment Bank), but below that level there is no micromanagement.

The American examples, in contrast, show much more political micromanagement. This is part of the same package as the privatization of state planning in the globalized system; in the United States often there was never the depoliticization that most of the rest of the developed and middle-income world had, but on top of that, the tendency has been to shut down in-house planning departments or radically shrink them and replace them with consultants. The consultants are then supervised by political appointees with no real qualifications to head capital programs, and the remaining civil servants are browbeaten not to disagree with the political appointees’ proclamations.

Those political appointees rarely measure themselves by any criteria of infrastructure utility. Even in New York they and the managers don’t consistently use the system; in Los Angeles, they use it about as often as the executive director of a well-endowed charity eats at a soup kitchen. To them, the cost is itself a measure of success – and this is true of other agencies, which treat obtaining other people’s money as a mark of achievement and as testament to their power. This behavior then cascades to local advocacy groups, which try to push solutions that maximize outside funding and are at bet indifferent and at worst actively hostile to any attempt at efficiency.

Just giving more support to agencies in their current configuration is not going to help. To the contrary, it only confirms that profligacy gets rewarded. A program of depoliticization of the state is required in tandem with expanding in-house hiring and reversing the globalized system, and the political appointees and the managers and political advocates who are used to dealing with them don’t belong in this program.

Edge Cities With and Without Historic Cores

An edge city is a dense, auto-oriented job center arising from nearby suburban areas, usually without top-down planning. The office parks of Silicon Valley are one such example: the area had a surplus of land and gradually became the core of the American tech industry. In American urbanism, Tysons in Virginia is a common archetype: the area was a minor crossroads until the Capital Beltway made it unusually accessible by car, providing extensive auto-oriented density with little historic core.

But there’s a peculiarity, I think mainly in the suburbs of New York. Unlike archetypal edge cities like Silicon Valley, Tysons, Century City in Los Angeles, or Route 128 north of Boston, some of the edge cities of New York are based on historic cores. Those include White Plains and Stamford, which have had booms in high-end jobs in the last 50 years due to job sprawl, but also Mineola, Tarrytown, and even New Brunswick and Morristown.

The upshot is that it’s much easier to connect these edge cities to public transportation than is typical. In Boston, I’ve spent a lot of time trying to figure out good last mile connections from commuter rail stations. Getting buses to connect outlying residential areas and shopping centers to town center stations is not too hard, but then Route 128 is completely unviable without some major redesign of its road network: the office parks front the freeway in a way that makes it impossible to run buses except dedicated shuttles from one office park to the station, which could never be frequent enough for all-day service. Tysons is investing enormous effort in sprawl repair, which only works because the Washington Metro could be extended there with multiple stations. Far and away, these edge cities are the most difficult case for transit revival for major employment centers.

And in New York, because so much edge city activity is close to historic cores, this is far easier. Stamford and White Plains already have nontrivial if very small transit usage among their workers, usually reverse-commuters who live in New York and take Metro-North. Mineola could too if the LIRR ran reverse-peak service, but it’s about to start doing so. Tarrytown and Sleepy Hollow could be transit-accessible. The New Jersey edge cities are harder – Edison and Woodbridge have lower job density than Downtown Stamford and Downtown White Plains – but there are some office parks that could be made walkable from the train stations.

I don’t know what the history of this peculiar feature is. White Plains and Mineola are both county seats and accreted jobs based on their status as early urban centers in regions that boomed with suburban sprawl in the middle of the 20th century. Tarrytown happened to be the landfall of the Tappan Zee Bridge. Perhaps this is what let them develop into edge cities even while having a much older urban history than Tysons (a decidedly non-urban crossroads until the Beltway was built), Route 128, or Silicon Valley (where San Jose was a latecomer to the tech industry).

What’s true is that all of these edge cities, while fairly close to train stations, are auto-oriented. They’re transit-adjacent but not transit-oriented, in the following ways:

  • The high-rise office buildings are within walking distance to the train station, but not with a neat density gradient in which the highest development intensity is nearest the station.
  • The land use at the stations is parking garages for the use of commuters who drive to the station and use the train as a shuttle from a parking lot to Manhattan, rather than as public transportation the way subway riders do.
  • The streets are fairly hostile to pedestrians, featuring fast car traffic and difficult crossing, without any of the walkability features that city centers have developed in the last 50 years.

The street changes required are fairly subtle. Let us compare White Plains with Metrotown, both image grabs taken from the same altitude:

These are both edge cities featuring a train station, big buildings, and wide roads. But in Metrotown, the big buildings are next to the train station, and the flat-looking building to its north is the third-largest shopping mall in Canada. The parking goes behind the buildings, with some lots adjoining Kingsway, which has a frequent trolleybus (line 19) but is secondary as a transportation artery to SkyTrain. Farther away, the residential density remains high, with many high-rises in the typical thin-and-tall style of Vancouver. In contrast, in White Plains, one side of the station is a freeway with low-density residential development behind it, and the other is parking garages with office buildings behind them instead of the reverse.

The work required to fix this situation is not extensive. Parking must be removed and replaced with tall buildings, which can be commercial or residential depending on demand. This can be done as part of a transit-first strategy at the municipal level, but can also be compelled top-down if the city objects, since the MTA (and other Northeastern state agencies) has preemption power over local zoning on land it owns, including parking lots and garages.

On the transit side, the usual reforms for improvements in suburban trains and buses would automatically make this viable: high local frequency, integrated bus-rail timetables (to replace the lost parking), integrated fares, etc. The primary target for such reforms is completely different – it’s urban and inner-suburban rail riders – but the beauty of the S-Bahn or RER concept is that it scales well for extending the same high quality of service to the suburbs.

The Four Quadrants of Cities for Transit Revival

Cities that wish to improve their public transportation access and usage are in a bind. Unless they’re already very transit-oriented, they have not only an entrenched economic elite that drives (for example, small business owners almost universally drive), but also have a physical layout that isn’t easy to retrofit even if there is political consensus for modal shift. Thus, to shift travel away from cars, new interventions are needed. Here, there is a distinction between old and new cities. Old cities usually have cores that can be made transit-oriented relatively easily; new cities have demand for new growth, which can be channeled into transit-oriented development. Thus, usually, in both kinds of cities, a considerably degree of modal shift is in fact possible.

However, it’s perhaps best to treat the features of old and new cities separately. The features of old cities that make transit revival possible, that is the presence of a historic core, and those of new cities, that is demand for future growth, are not in perfect negative correlation. In fact, I’m not sure they consistently have negative correlation at all. So this is really a two-by-two diagram, producing four quadrants of potential transit cities.

Old cities

The history of public transportation is one of decline in the second half of the 20th century in places that were already rich then; newly-industrialized countries often have different histories. The upshot is that an old auto-oriented place must have been a sizable city before the decline of mass transit, giving it a large core to work from. This core is typically fairly walkable and dense, so transit revival would start from there.

The most successful examples I know of involve the restoration of historic railroads as modern regional lines. Germany is full of small towns that have done so; Hans-Joachim Zierke has some examples of low-cost restoration of regional lines. Overall, Germany writ large must be viewed as such an example: while German economic growth is healthy, population growth is anemic, and the gradual increase in the modal split for public transportation here must be viewed as more intensive reuse of a historic national rail network, anchored by tens of small city cores.

At the level of a metropolitan area, the best candidates for such a revival are similarly old places; in North America, the best I can think of for this are Philadelphia, Boston, and Chicago. Americans don’t perceive any of the three as especially auto-oriented, but their modal splits are comparable to those of small French cities. But in a way, they show one way forward. If there’s a walkable, transit-oriented core, then it may be attractive for people to live near city center; in those three cities it’s also possible to live farther away and commute by subway, but in smaller ones (say, smaller New England cities), the subway is not available but conversely it’s usually affordable to live within walking distance of the historic city center. This creates a New Left-flavored transit revival in that it begins with the dense city center as a locus of consumption, and only then, as a critical mass of people lives there, as a place that it’s worth building new urban rail to.

New cities

Usually, if a city has a lot of recent growth from the era in which it has become taken for granted that mobility is by car, then it should have demand for further growth in the future. This demand can be planned around growth zones with a combination of higher residential density and higher job density near rail corridors. The best time to do transit-oriented development is before auto-oriented development patterns even set in.

There are multiple North American examples of how this works. The best is Vancouver, a metropolitan area that has gone from 560,000 people in the 1951 census to 2.6 million in the 2021 census. Ordinarily, one should expect such a region to be entirely auto-oriented, as most American cities with almost entirely postwar growth are; but in 2016, the last census before corona, it had a 20% work trip modal split, and that was before the Evergreen extension opened.

Vancouver has achieved this by using its strong demand for growth to build a high-rise city center, with office towers in the very center and residential ones ringing it, as well as high-density residential neighborhoods next to the Expo Line stations. The biggest suburbs of Vancouver have followed the same plan: Burnaby built an entirely new city center at Metrotown in conjunction with the Expo Line, and even more auto-oriented Surrey has built up Whalley, at the current outer terminal of the line, as one of its main city centers. Housing growth in the region is rapid; YIMBY advocacy calls for more, but the main focus isn’t on broad development (since this already happens) but on permitting more housing in recalcitrant rich areas, led by the West Side, which will soon have its Broadway extension of the Millennium Line.

Less certain but still interesting examples of the same principle are Calgary, Seattle, and Washington. Calgary, a low-density city, planned its growth around the C-Train, and built a high-rise city center, limiting job sprawl even as residential sprawl is extensive; Seattle and the Virginia-side suburbs of Washington have permitted extensive infill housing and this has helped their urban rail systems achieve high ridership by American standards, Seattle even overtaking Philadelphia’s modal split.

The four quadrants

The above contrast of old and new cities misses cities that have positive features of both – or neither. The cities with both positive features have the easiest time improving their public transportation systems, and many have never been truly auto-oriented, such as New York or Berlin, to the point that they’re not the best examples to use for how a more auto-oriented city can redevelop as a transit city.

In North America, the best example of both is San Francisco, which simultaneously is an old city with a high-density core and a place with immense demand for growth fueled by the tech industry. The third-generation tech firms – those founded from the mid-2000s onward (Facebook is in a way the last second-generation firm, which generation began with Apple and Microsoft) – have generally headquartered in the city and not in Silicon Valley. Twitter, Uber, Lyft, Airbnb, Dropbox, and Slack are all in the city, and the traditional central business district has expanded to South of Market to accommodate. This is really a combination of the consumption-oriented old-city model, as growing numbers of employees of older second-generation firms chose to live in the city and reverse-commute to Silicon Valley, and the growth-oriented new-city model. Not for nothing, the narrower metropolitan statistical area of San Francisco (without Silicon Valley) reached a modal split of 17% just before corona, the second highest in the United States, with healthy projections for growth.

But then there is the other quadrant, comprising cities that have neither the positive features of old cities nor those of new cities. To be in this quadrant, a city must not be so old as to have a large historic core or an extensive legacy rail network that can be revived, but also be too poor and stagnant to generate new growth demand. Such a city therefore must have grown in a fairly narrow period of time in the early- to mid-20th century. The best example I can think of is Detroit. The consumption-centric model of old city growth can work even there, but it can’t scale well, since there’s not enough of a core compared with the current extent of the population to build out of.

Schedule Planners as a Resource

The Effective Transit Alliance published its statement on Riders Alliance’s Six-Minute Service campaign, which proposes to run every subway line in New York and the top 100 bus routes every (at worst) six minutes every day from morning to evening. We’re positive on it, even more than Riders Alliance is. We go over how frequency benefits riders, as I wrote here and here, but also over how it makes planning easier. It is the latter benefit I want to go over right now: schedule planning staff is a resource, just as drivers and outside capital are, and it’s important for transit agencies to institute systems that conserve this resource and avoid creating unnecessary work for planners.

The current situation in New York

Uday Schultz writes about how schedule planning is done in New York. There’s an operations planning department, with 350 budgeted positions as of 2021 of which 284 are filled, down from 400 and 377 respectively in 2016. The department is responsible for all aspects of schedule planning: base schedules but also schedules for every service change (“General Order” or GO in short).

Each numbered or lettered route is timetabled on it own. The frequency is set by a guideline coming from peak crowding: at any off-peak period, at the most crowded point of a route, passenger crowding is supposed to be 25% higher than the seated capacity of the train; at rush hour, higher standee crowding levels are tolerated, and in practice vary widely by route. This way, two subway routes that share tracks for a long stretch will typically have different frequencies, and in practice, as perceived by passengers, off-peak crowding levels vary and are usually worse than the 25% standee factor.

Moreover, because planning is done by route, two trains that share tracks will have separate schedule plans, with little regard for integration. Occasionally, as Uday points out, this leads to literally impossible schedules. More commonly, this leads to irregular gaps: for example, the E and F trains run at the same frequency, every 4 minutes peak and every 12 minutes on weekends, but on weekends they are offset by just 2 minutes from each other, so on the long stretch of the Queens Boulevard Line where they share the express tracks, passengers have a 2-minute wait followed by a 10-minute wait.

Why?

The current situation creates more work for schedule planners, in all of the following ways:

  • Each route is run on its own set of frequencies.
  • Routes that share tracks can have different frequencies, requiring special attention to ensure that trains do not conflict.
  • Each period of day (morning peak, midday, afternoon peak, evening) is planned separately, with transitions between peak and off-peak; there are separate schedules for the weekend.
  • There are extensive GOs, each requiring not just its own bespoke timetable but also a plan for ramping down service before the start of the GO and ramping it up after it ends.

This way, a department of 284 operations planners is understaffed and cuts corners, leading to irregular and often excessively long gaps between trains. In effect, managerial rules for how to plan trains have created makework for the planners, so that an objectively enormous department still has too much work to do and cannot write coherent schedules.

Creating less work for planners

Operations planners, like any other group of employees, are a resource. It’s possible to get more of this resource by spending more money, but office staff is not cheap and American public-sector hiring has problems with uncompetitive salaries. Moreover, the makework effect doesn’t dissipate if more people are hire – it’s always possible to create more work for more planners, for example by micromanaging frequency at ever more granular levels.

To conserve this resource, multiple strategies should be used:

Regular frequencies

If all trains run on the same frequency all day, there’s less work to do, freeing up staff resources toward making sure that the timetables work without any conflict. If a distinction between peak and base is required, as on the absolute busiest routes like the E and F, then the base should be the same during all off-peak periods, so that only two schedules (peak and off-peak) are required with a ramp-up and ramp-down at the transition. This is what the six-minute service program does, but it could equally be done with a more austere and worse-for-passengers schedule, such as running trains every eight minutes off-peak.

Deinterlining

Reducing the extent of reverse-branching would enable planning more parts of the system separately from one another without so much conflict. Note that deinterlining for the purposes of good passenger service has somewhat different priorities from deinterlining for the purposes of coherent planning. I wrote about the former here and here. For the latter, it’s most important to reduce the number of connected components in the track-sharing graph, which means breaking apart the system inherited from the BMT from that inherited from the IND.

The two goals share a priority in fixing DeKalb Avenue, so that in both Manhattan and Brooklyn, the B and D share tracks as do the N and Q (today, in Brooklyn, the B shares track with the Q whereas the D shares track with the N): DeKalb Junction is a timetabling mess and trains have to wait two minutes there for a slot. Conversely, the main benefit of reverse-branching, one-seat rides to more places, is reduced since the two Manhattan trunks so fed, on Sixth Avenue and Broadway, are close to each other.

However, to enable more convenient planning, the next goal for deinterlining must be to stop using 11th Street Connection in regular service, which today transitions the R from the BMT Broadway Line and 60th Street Tunnel to the IND Queens Boulevard local tracks. Instead, the R should go where Broadway local trains go, that is Astoria, while the Broadway express N should go to Second Avenue Subway to increase service there. The vacated local service on Queens Boulevard should go to IND trunks in Manhattan, to Eighth or Sixth Avenue depending on what’s available based on changes to the rest of the system; currently, Eighth Avenue is where there is space. Optionally, no new route should be added, and instead local service on Queens Boulevard could run as a single service (currently the M) every 4 minutes all day, to match peak E and F frequencies.

GO reform

New York uses too many GOs, messing up weekend service. This is ostensibly for maintenance and worker safety, but maintenance work gets done elsewhere with fewer changes (as in Paris or Berlin) or almost none (as in Tokyo) – and Berlin and Tokyo barely have nighttime windows for maintenance, Tokyo’s nighttime outages lasting at most 3-4 hours and Berlin’s available only five nights a week. The system should push back against ever more creative service disruptions for work and demand higher maintenance productivity.

Midwestern Urban Geography and High-Speed Rail

I’ve been uploading videos about high-speed rail lately, of which the most recent, from a week ago, is a return to my attempt at producing a high-speed rail map proposal for the eastern half of the United States. I streamed and then blogged a map here with followup here, but having looked at the model more, I’d like to do a refinement – both to introduce a slightly bigger map and explain why it is so, and talk about the issue of connecting low-speed lines. Along the way, I feel like I must talk about an issue mentioned in comments occasionally about the politics of only connecting major metropolitan areas, especially since this map still has fewer lines than various state wishlists stapled by Amtrak; this is especially important because one of the motivations for this post is a criticism of current plans by Matt Yglesias.

The map

In this map, like some old maps but unlike in last year’s map, red denotes high-speed lines and blue denotes low-speed ones, some running through to the high-speed network and some not

A full-size (6 MB) version of the map can be found here. This is not intended as a comprehensive map of all desirable low-speed connections – I made no effort to include the Northeastern ones, which I wrote about in the context of New England and Upstate New York, and which Ben She has done good work on in the context of eastern Pennsylvania and the mid-Atlantic. Rather, I want to focus on the Midwest.

But first, to explain a little more about why this map includes more red (high-speed) lines than previously, the reason has to do with my spreadsheet for computing ridership density based on Metcalfe’s law. The original posts computed everything by hand, which meant that some low-ridership city pairs I just rounded to zero; the spreadsheet does include them, making every line look much stronger. This, in particular, makes St. Louis-Kansas City and Atlanta-Birmingham, omitted last year, and Nashville-Memphis, suggested last year as a maybe, solid propositions.

A note of caution is still advised. Those weak city pairs that aggregate to sufficient ridership for significant return on investment are often at long distance, such as Kansas City-New York. The ridership model is trained on Shinkansen data out of Tokyo and sanity-checked with some French, German, and Spanish data, but the same model overpredicts Shinkansen ridership on inter-island trips for which planes are a convenient alternative, like Tokyo-Fukuoka or Tokyo-Hakodate. This makes me reluctant to add a Kansas City-Dallas connection, which the spreadsheet thinks generates a bit more than $1 million in annual operating profit per km of new construction: the extra ridership out of Kansas City-Dallas includes some very long-distance trips like Dallas-Detroit, for which the model is likely an overprediction.

The truth is likely between the spreadsheet and the handmade version of the model; while the Shinkansen is not competitive with planes when trains take five hours, European high-speed trains are, for example Paris-Nice. This leads to the inclusion of the new sections, but the exclusion of Kansas City-Dallas. Note also that I did look at Birmingham-New Orleans and Memphis-Little Rock, and both were weak even in the spreadsheet (though I did not attempt Birmingham-New Orleans-Houston) – the Deep South is too low-density and rural to support as expansive a system.

But the topic of this post is not the South, but the Midwest.

Midwestern urban geography

The United States is usually a country of fewer, bigger metropolitan areas, like rich Asia and unlike Europe; unlike both Europe and Asia, American cities are very decentralized, and the exceptions are in the Northeast and West, not the Midwest. In particular, naive comparisons of Midwestern to French high-speed rail corridors are unwarranted: while Chicago is of the same approximate size as Paris, and secondary Midwestern metropolitan areas like Detroit and St. Louis are substantially larger than French ones like Lyon and Marseille, Lyon and Marseille are ringed by many small metropolitan areas with their own TGV service, whereas at the same radius, St. Louis has only its suburbs.

However, this phenomenon of fewer, bigger metro areas has exceptions. Michigan, in particular, has a slightly more European geography. Using the smaller numbers produced by the metropolitan statistical area (MSA) calculation rather than the broader combined statistical area (CSA), Metro Detroit has 43% of Michigan’s population as of the 2020 census. The median Michigander lives in the Grand Rapids MSA, with 1.1 million people, fewer than the US-wide median of 1.6-1.7 million. Michigan is a fairly urban state, and below Grand Rapids is a succession of six-figure metropolitan areas: Ann Arbor, Lansing, Flint, Kalamazoo, Battle Creek, Saginaw.

Ohio is similar to Michigan. Its three main metro areas, excluding Cincinnati’s out-of-state suburbs, have just a hair less than half the state’s population; the median Buckeye therefore lives in Dayton, MSA population 800,000. Moreover, the southern half of Michigan has fairly high population density, as does Ohio – nothing as dense as the Northeast or Germany, but they’re comparable to France.

This geography lends itself to an expansive intercity rail network: the cities are relatively close to one another, and there are many of them meriting a connection. In Ohio, this happens to take the form of an entirely high-speed network, since Cleveland, Columbus, Dayton, and Cincinnati all lie on one line, and then the most natural east-west route between the Northeast and the Midwest passes through Cleveland and Toledo. Ohio, in this case, is a state with fairly good geography for low-speed intercity rail that just happens to also have good geography for high-speed rail due to its location. Michigan, in contrast, is not on the way between much, and thus should get a low-speed rail network, including both connections to Chicago (such as to Grand Rapids) and an intra-state network.

Wisconsin has many, smaller cities as well: the median resident is in an MSA of around 200,000 people, currently Racine. Fortunately, many of those cities lie on just one line between Chicago and Minneapolis, plus a low-speed branch up to Green Bay. Unfortunately, coverage is lacking by the standards of Ohio, Michigan, or much more big city-dominated Illinois and Minnesota.

Getting low-speed rail right

I am happy to report that in Michigan and Ohio at least, good projects for low-speed rail are pursued. When I streamed my video, I was told in Twitch chat that Michigan is looking into funding a Detroit-Lansing-Grand Rapids intercity train. Ohio likewise has long had ideas for a statewide network, beginning with the Cleveland-Cincinnati spine.

It is unfortunate that these projects are not planned well. In a future post, I should write more about the concept of the wrong project versus the right project done poorly; I obliquely pointed this out when writing about leakage in the context of urban transit, where some American cities have poor project prioritization (such as Los Angeles) whereas others choose more or less the right projects but execute them poorly (such as New York and San Francisco). In this schema, the current plans for low-speed rail are often the right project, done wrong.

What I mean by this is that there’s a set of best industry practices for getting low-speed (that is, legacy) rail right, emanating out of Germany and surrounding countries, especially Switzerland. These include,

  • Integration of timetable planning and infrastructure, to minimize construction costs – if higher costs are acceptable, just build high-speed rail.
  • A clockface all-day schedule with a minimum of one train every two hours, and ideally a train every hour if the distances are shorter or the cities are bigger.
  • Timed connections at major nodes to both other intercity trains within the same network and regional mass transit (such as regional trains or connecting buses).
  • Reliability-centric planning, in which sources of delays are to be proactively eliminated – on a system this tightly coordinated, delays on one line propagate across the entire system.
  • An average speed of around 100-130 km/h – the higher numbers are more appropriate in flat terrain.

Marco Chitti has an excellent post that I must revisit in the future giving more detailed guidelines, mostly at regional level but also with an eye toward national intercity rail planning.

The upshot is that trying to incrementally build up ridership for a few trains per day does not work. The US has a few trains per day on a few corridors now, such as Chicago-Detroit, and daily trains on most others, and this hasn’t built up ridership. The low-speed, low-frequency intercity trains Europe had before the introduction of the TGV in France and the high-frequency, tightly-linked InterCity network in Germany were rapidly losing market share to cars and planes. To build such a network now would be like to build infrastructures wired telephones in a developing country rather than just skipping straight to cellphones as developing countries have.

Politics and Matt Yglesias’s post

Matt likes pointing out that current transportation plans in the United States are deficient, and to link to my posts as a better alternative. There was a lot of dunking on Matt’s post about this (as there is on anything that Matt writes) by left-identified people who, in effect, think daily or twice-daily trains between small cities are a great investment. This dunking usually takes the form of “how dare Matt, a lifelong East Coaster, tell people in [insert Midwestern town] that they don’t deserve trains?”. In a less abrasive form, some people in comments here, like Pharisee, have made the point that these expansive maps proposing daily trains to many places have good geographic coverage whereas what I propose does not. Let me explain why this line of thinking is wrong.

The issue is that the United States is, again, a country that for the most part has fewer, larger metropolitan areas than Europe. The map I made above hits MSAs with a large majority of the country’s population. Of the top 50 as of the 2020 census, the only misses are Denver, Oklahoma City, Salt Lake City, and New Orleans. Denver and Salt Lake City are far from everything else, and the other two are in theory on routes from Texas to the rest of the country but there’s too little population on the way for a connection in the geography of 2022.

Moreover, within the Midwest, coverage is ample. The Midwest is a highly (sub-)urbanized region, much of which has fairly high population density, and the biggest exception to the high density, Minnesota, has a large enough city to justify a line to Chicago by itself (and Milwaukee is on the way, too). The areas that are usually most moralized as Real America – Michigan, Ohio, Pennsylvania – are on the way. This shouldn’t be too surprising: the Real America moralization centers areas with a past industrial history, evoking feelings of nostalgia for midcentury American industrial dominance, and those areas remain major cities today, just relatively poorer than they were in the 1960s. This way, in the Midwest, every state has a large majority of its population in an MSA with a high- or low-speed train station on my map, except Iowa, which is unusually rural.

This is not even out of any consciously political desire to serve these areas. I draw maps out of a ridership model. It just so happens that metropolitan areas of 4 million people situated in dense geographies scream “build high-speed trains to me,” and those include Detroit.

The problem – the reason Matt is so negative on current plans – is that current plans are bad. They are low- and not high-speed rail, which by itself is not horrible, but they’re also bad low-speed rail. Daily trains are just not good. But this does not mean the Midwest can’t or shouldn’t get a good intercity rail network: it should, combining high- and low-speed trains as appropriate.

TransitCenter’s Commuter Rail Proposal

Last week, TransitCenter released a proposal for how to use commuter rail more effectively within New York. The centerpiece of the proposal is to modify service so that the LIRR and Metro-North can run more frequently to stations within the city, where today they serve the suburbs almost exclusively; at the few places near the outer end of the city where they run near the subway, they have far less ridership, often by a full order of magnitude, which pattern repeats itself around North America. There is much to like about what the proposal centers; unfortunately, it falls short by proposing half-hourly frequencies, which, while better than current off-peak service, are far short of what is needed within the city.

Commuter rail and urban ridership

TransitCenter’s proposal centers urban riders. This is a welcome addition to city discourse on commuter rail improvement. The highest-ridership, highest-traffic form of mainline rail is the fundamentally urban S-Bahn or RER concept. Truly regional trains, connecting distinct centers, coexist with them but always get a fraction of the traffic, because public transit ridership is driven by riders in dense urban and inner-suburban neighborhoods.

A lot of transit and environmental activists are uncomfortable with the idea of urban service. I can’t tell why, but too many proposals by people who should know better keep centering the suburbs. But in reality, any improvement in commuter rail service that does not explicitly forgo good practices in order to discourage urban ridership creates new urban ridership more than anything else. There just aren’t enough people in the suburbs who work in the city (even in the entire city, not just city center) for it to be any other way.

TransitCenter gets it. The proposal doesn’t even talk about inner-suburban anchors of local lines just outside the city, like Yonkers, New Rochelle, and Hempstead (and a future update of this program perhaps should). No: it focuses on the people near LIRR and Metro-North stations within the city, highlighting how they face the choice between paying extra for infrequent but fast trains to Midtown and riding very slow buses to the edge of the subway system. As these neighborhoods are for the most part on the spectrum from poor to lower middle-class, nearly everyone chooses the slow option, and ridership at the city stations is weak, except in higher-income Northeast Queens near the Port Washington Branch (see 2012-4 data here, PDF-pp. 183-207), and even there, Flushing has very little ridership since the subway is available as an alternative.

To that effect, TransitCenter proposes gradually integrating the fares between commuter rail and urban transit. This includes fare equalization and free transfers: if a bus-subway-bus trip between the Bronx and Southern Brooklyn is covered by the $127 monthly pass then so should a shorter bus-commuter rail trip between Eastern Queens or the North Bronx and Manhattan.

Interestingly, the report also shows that regionwide, poorer people have better job access by transit than richer people, even when a fare budget is imposed that excludes commuter rail. The reason is that in New York, suburbanization is a largely middle-class phenomenon, and in the suburbs, the only jobs accessible by mass transit within an hour are in Midtown Manhattan, whereas city residents have access to a greater variety of jobs by the bus and subway system. But this does not mean that the present system is equitable – rich suburbanites have cars and can use them to get to edge city jobs such as those of White Plains and Stamford, and can access the entire transit network without the fare budget whereas poorer people do have a fare budget.

The issue of frequency

Unfortunately, TransitCenter’s proposal on frequency leaves a lot to be desired. Perhaps it’s out of incrementalism, of the same kind that shows up in its intermediate steps toward fare integration. The report suggests to increase frequency to the urban stations to a train every half an hour, which it phrases in the traditional commuter rail way of trains per day: 12 roundtrips in a six-hour midday period.

And this is where the otherwise great study loses me. Forest Hills, Kew Gardens, and Flushing are all right next to subway stations. The LIRR charges higher fares there, but these are fairly middle-class areas – richer than Rosedale in Southeast Queens on the Far Rockaway Branch, which still gets more ridership than all three. No: the problem in these inner areas is frequency, and a train every half hour just doesn’t cut it when the subway is right there and comes every 2-3 minutes at rush hour and every 4-6 off-peak.

In this case, incremental increases from hourly to half-hourly frequency don’t cut it. The in-vehicle trip is so short that a train every half hour might as well not exist, just as nobody runs subway trains every half hour (even late at night, New York runs the subway every 20 minutes). At outer-urban locations like Bayside, Wakefield, and Rosedale, the absolute worst that should be considered is a train every 15 minutes, and even that is suspect and 10 minutes is more secure. Next to the subway, the absolute minimum is a train every 10 minutes.

All three mainlines currently radiating out of Manhattan in regular service – the Harlem Line, the LIRR Main Line, and the Port Washington Branch – closely parallel very busy subway trunk lines. One of the purposes of commuter rail modernization in New York must be decongestion of the subway, moving passengers from overcrowded 4, 5, 7, E, and F trains to underfull commuter trains. The LIRR and Metro-North are considered at capacity when passengers start having to use the middle seats, corresponding to 80% of seated capacity; the subway is considered at capacity when there are so many standees they don’t meet the standard of 3 square feet per person (3.59 people/m^2).

To do this, it’s necessary to not just compete with buses, but also directly compete with the subway. This is fine: Metro-North and the LIRR can act as additional express capacity, filling trains every 5 minutes using a combination of urban ridership and additional ridership at inner suburbs. TransitCenter has an excellent proposal for how to improve service quality at the urban stations but then inexplicably doesn’t go all the way and proposes a frequency that’s too low.

The Official Brooklyn Bus Redesign is Out

The MTA just released a draft of the Brooklyn bus redesign it and its consultant had been working on. It is not good. I’m not completely sure why this is – the Queens redesign was a good deal better, and our take on it at the Effective Transit Alliance was decidedly positive. But in the case of Brooklyn, the things that worked in Queens are absent. Overall, the theme of this is stasis – the changes to the network are minor, and the frequencies are to remain insufficiently low for good service. The only good thing about this is stop consolidation, which does not require spending any money on consultants and is a straightforward fix.

This is especially frustrating to me because my first project for Marron, before the Transit Costs Project, was a redesign proposal. The proposal can be read here, with discussion in blog posts here, here, and here. The official reaction we got was chilly, but the redesign doesn’t look anything like a more politic version, just one produced at much higher consultant cost while doing very little.

The four-color scheme

The Brooklyn project retains the Queens redesign’s four-color scheme of buses, to be divided into local (green), limited (red), Select Bus Service (blue), and rush (purple). The local buses are supposed to stop every 300-400 meters, which is not the best (the optimum for Brooklyn is about 400-500) but is a good deal better than the current spacing of about 200-250. The other three kinds of buses are more express, some running on the same routes as local buses as express overlays and some running on streets without local service.

In Queens, this four-way distinction emerges from the pattern in which in neighborhoods beyond the subway’s reach, bus usage is extremely peaky toward the subway. The purpose of the rush route is to get people to the subway terminal, such as Flushing or Jamaica, with not just longer stop spacing but also long nonstop sections close to the terminal where local service exists as an overlay, imitating the local and express patterns of peaky commuter rail operations in New York. I still think it’s not a good idea and buses should run at a more uniform interstation at higher frequency. But over the long stretches of Eastern Queens, the decision is fairly close and while rush routes are not optimal, they’re not much worse than the optimum. In contrast, Brooklyn is nothing like Queens: people travel shorter distances, and long routes are often used as circumferential subway connectors with ample turnover.

Ironically, this is something the MTA and its consultants understood: the Brooklyn map is largely green, whereas that in Queens has a more even mix of all four colors. Nonetheless, some rush routes are retained and so are some limited-only routes, in a way that subtracts value: if nearly all buses in Brooklyn offer me something, I should expect it on the other buses as well, whereas the rush-only B26 on Halsey Street is different in a way that isn’t clear.

In general, the notable feature of our redesign, unlike the more common American ones, is that there is no distinction among the different routes. Some are more frequent than others, but all have very high base frequency. This is because Brooklyn has unusually isotropic travel: density decreases from the East River south- and eastward, but the subway network also thins out and these effects mostly cancel out, especially with the high density of some housing projects in Coney Island; the busiest buses include some running only within Southern Brooklyn, like the B6 and B82 circumferentials.

In contrast, small-city redesigns tend to occur in a context with a strong core network and a weak peripheral network (“coverage routes,” which exist to reassure loud communities with no transit ridership that they can get buses too), and the redesign process tends to center this distinction and invest in the stronger core network. Queens has elements that look like this, if you squint your eyes sufficiently. Brooklyn has none: the isotropic density of most of the borough ensures that splitting buses into separate classes is counterproductive.

Frequency

The frequency in the proposed system is, frankly, bad. The MTA seems to believe that the appropriate frequency for urban mass transit is a train or bus every 10 minutes. This is acceptable in the suburban neighborhoods of Berlin or the outermost parts of New York, like the Rockaways and the eastern margin of Queens. In denser areas, including all of Brooklyn, it is not acceptable. People travel short distances: citywide, the average trip distance before corona was 3.4 km, which works out to 18 minutes at average New York bus speed (source: NTD 2019). In Brooklyn, the dense mesh of buses going between subway lines rather than to them makes the average even slightly lower. This means that very high frequency is a high priority.

So bad is the MTA’s thinking about frequency that core routes in the borough are split into local and limited variants, each running every 10 minutes off-peak, including some of the busiest corridors in the borough, like the outer circumferential B6 and B82 and the more inner-circumferential B35 on Church (split in the plan into a local B35 and an SBS B55). This is not changed from the current design, even though it’s easy to do so in the context of general consolidation of stops.

To make this even worse, there does not appear to be any increase in service-km, judging by the plan’s lack of net increase in frequency. This is bad planning: bus operating costs come from time (driver’s wage, mainly) and not distance, and the speedup provided by the stop consolidation should fuel an increase in frequency.

The Battery Tunnel

The most annoying aspect, at least to me, is the lack of a bus in the Brooklyn-Battery Tunnel, connecting Manhattan with Red Hook. Red Hook is isolated from the subway and from the rest of Brooklyn thanks to the freeway, and has bus connections only internally to Brooklyn where in fact a short bus route through the tunnel would beat bus-subway connections to Lower Manhattan.

We got the idea for the inclusion of such a bus service from planners that we spoke to when we wrote our own redesign. The service is cheap to provide because of the short length of the route, and would complement the rest of the network. It was also popular in the neighborhood meetings that tee consultants ran, we are told. And yet, it was deleted on a whim.