Introduction — The Puzzle of Tokyo's Mobility

A Morning at Shinjuku

At 8:47 on a Tuesday in October, the south concourse of Shinjuku Station registers a flow rate that would, in almost any other context, be treated as a crowd-control emergency. From the Ōedo Line platforms beneath the ground to the elevated JR Yamanote Line platforms seventeen metres above, passengers are moving in streams whose direction is unmistakable yet whose density, at the level of the individual body, is unremarkable. A woman in a charcoal suit walks at the same pace as the man to her left; both pass a schoolboy carrying a cello without either having to alter course. The streams braid and diverge around a concrete pillar with the ease of water moving around a stone in a current that has been running for decades. Above the streams, suspended from the ceiling, six digital departure boards refresh at twelve-second intervals, each showing a departure time in seconds-to-the-minute, a destination in Japanese and English, and a platform number in green or amber. None of the trains shown is delayed. None of the passengers seems to expect any of them to be. At 8:49, a Yamanote Line train arrives at platform 14, dwells for thirty-eight seconds, and departs. The platform clears within ninety seconds. A Marunouchi Line train surfaces simultaneously at platform 5 in a different terminal block two hundred metres north. Aboard these trains, which by then are accelerating at the standard 0.92 metres per second squared, the same choreography plays out at every one of the roughly nine hundred stations that constitute the metropolitan rail network, three and a half million of whose riders are simultaneously in motion. Outside, on the pavement of the West Exit, a controlled turbulence of cigarette smoke, taxi queues, and schoolchildren in identical navy caps marks the boundary where the rail system ends and the surface city begins. None of this is exceptional. None of it is the product of a particularly clever morning. It is what an ordinary Tuesday in Shinjuku looks like, and what it has looked like, on essentially every Tuesday that the station has existed, since 1885.

The scene is striking not because anything fails. It is striking because nothing does. The station handles, on a typical weekday, approximately 3.5 million boardings across the six railway operators that share its concourses—more than the combined daily boardings of every terminal in Chicago, Philadelphia, and Boston. The metropolitan rail network as a whole handles approximately 28 million boardings per day. The Greater Tokyo region, of which Shinjuku is one node among many, houses roughly 37 to 38 million residents, depending on which of the several definitions one uses to draw the boundary. By any reasonable accounting, this is the largest and most intensively used urban rail system on earth. Yet the system does not visibly fail. Trains arrive when the boards say they will. Queues form without instructions. The escalator on the left has people standing on the right, and on the right has people standing on the left, and at any given escalator the convention is consistent across all riders for its entire length, because the convention is the same at every other escalator in the system. Fare evasion is so rare that transit police do not routinely patrol for it. There is, in other words, a kind of order that is not the absence of disorder but the presence of an order so thoroughly internalized that it becomes invisible to the people enacting it. This book begins by asking what that order is, how it is produced, what holds it together, and what it can teach readers who live in cities where the equivalent scene would not survive a single morning.

The question is not, in the first instance, about trains. Trains are the most visible instrument of Tokyo's mobility, and they will occupy most of this book's attention, but they are not the substance of the question. The substance is this: how does a region of roughly 38 million people, occupying a built-up area of about 8,200 square kilometres, move itself to and from work, school, and commerce twice a day, every day, year after year, with a reliability, a regularity, and a passenger throughput that no other metropolitan region on earth approaches, and do so at a cost per rider that is, in real terms, comparable to or lower than that of cities a tenth its size? What is the engineering, the institutional, and the cultural machinery that makes this possible? And which parts of that machinery are specific to Tokyo and which, if any, are transferable to other places where mobility is failing or has never reached this scale?

This chapter opens the inquiry with three moves. First, it quantifies the phenomenon so that the reader has a defensible sense of the scale of movement being discussed. Second, it identifies the paradox that the scale presents: a system that, by ordinary engineering intuition, should be brittle, congested, or unreliable behaves as if brittleness, congestion, and unreliability have been engineered out of it. Third, it proposes a framework—a three-axis model of efficiency (time, throughput, and reach) and a three-pillar model of causes (rail engineering, land use, and time discipline)—that the rest of the book will populate with evidence, history, geography, and case study. The framework is provisional. It will be tested, refined, and at points partially dismantled in the chapters that follow. But it is the working hypothesis on which the rest of the book proceeds, and the reader should hold it loosely enough to revise.

Quantifying the Tokyo Commute

The phrase "Tokyo commute" is shorthand for a movement of people that is, in its aggregate, larger than the entire population of most sovereign states and that occurs in two sharp daily peaks of approximately three hours each, separated by a long mid-day trough that is itself substantial, and surrounded by an evening peak that mirrors the morning in reverse. To describe it accurately, one has to be careful about which number one is reporting, because several different quantities are routinely confused in popular discussion.

The most commonly cited figure, approximately 28 million, refers to daily rail boardings within the Greater Tokyo region. A "boarding" is a single act of entering a vehicle that is part of the rail system, regardless of whether the rider will later transfer to another vehicle; a passenger who travels from Kawagoe to central Tokyo on the Seibu Shinjuku Line, transfers to the Marunouchi Line, and exits at Tokyo Station will, on most counts, register three or four boardings depending on how transfers through faregates are treated. The 28 million figure therefore counts the number of times people cross faregates, not the number of unique individuals. The number of unique individuals who use rail on a given weekday is substantially lower—estimates range from 16 to 20 million—because a substantial fraction of riders commute both directions and a smaller fraction make multiple trips.

This 28 million figure is distributed across roughly a dozen operators. The dominant carriers are the East Japan Railway Company (JR East), which alone handles about 16 million boardings per weekday on its metropolitan-area services; the Tokyo Metro, with about 7 million; the Toei Subway (operated by the Tokyo Metropolitan Government), with about 2.5 million; and a collection of major private railways—Tokyu, Odakyu, Keio, Seibu, Tobu, Keisei, and others—that each handle between 0.5 and 3 million. Smaller private railways, monorails, automated guideway transit (the Yurikamome), and tramways contribute the remainder. The point worth pausing on is that this is not a single system. It is a federation of operators, many of them private, several of them holding legacy franchises from the prewar era, who share track, share platforms, share rolling stock procurement, share timetable coordination, and share a common fare-adjustment culture, but who do not share a single owne