The Urban Geography of Tokyo

Sam Bur takes us on a urban geography tour of Tokyo 


Earlier this week Rob Mayo delivered a guest post on Japan’s transit system and how stations are used as destination places rather than ones of just travelling through (as Auckland does) (see: Guest Post: Turning passengers into shoppers – adopting a Japan-style public transport service model for Auckland).

In a ‘just-in-time’ coincidence resident Australian Cities Skylines Player and Planner Sam Bur released a video this morning on the Urban Geography of Tokyo – one of the World’s largest and most dense cities! 

The video from Sam:


If you have been to or seen pictures of Tokyo you notice the grid street patterns connected with non-grid roads as the City continued to expand in a haphazard but also organic fashion. This is compared to North American cities (and Auckland post 1991) which seemed to have grown like (as we call it in Sim City 4 or Cities Skylines) ploppables to get some form of “perfection” or rather control.

As a side: In Sim City and Cities Skylines there were two dominate playing styles (that often triggered off very long winded debates in chat rooms):

  • Organic using just the RCI (residential, commercial and industrial) zone tools. 
  • Plop everything (RCI buildings could be plopped by the player allowing the player to micro detail how the City is developed and control what goes where rather than let “city forces” do the development itself)

With the Organic city building style the basic premise was that you build some basic transport spines (arterial road or transit lines) then build offshoots from that spine and place down your RCI zones. When it was time to expand you basically just added more RCI to the edge of the previous expansion and if need be build a secondary transport spine.  The result was a rather haphazard method of spatial development that would give its organic urban geographic appearance. 

The basic premise of Plop Everything was that everything was minutely controlled from transport, civic infrastructure to what RCI buildings could go where. It would be rare for the City to to develop or expand on its on convolution. The City appearance could be grid or cul-de-sacs everywhere but the urban geography of the area is one of uniformity of the urban fabric rather than the rather interesting heterogeneous “mess” seen with organic urban geographies. 

Tokyo has developed in the organic fashion that seems so haphazard and would look completely chaotic to those use to the North American “ploppable” way of urban geography. Yes this comes down to how each Authority treats their planning laws (Japan is very relaxed while American like New Zealand is very prescriptive) 

Ontakesan shops. Note the urban design layout – and the lack of cars Source: Rob Mayo

Rather ironically and as Sam would point out Tokyo despite its organic urban geography keeps it all together and functions extremely well given its size. While in North America their cities with their “ploppable” urban geographies (apart from maybe the core of their oldest cities) seem to be a magnitude of chaos to the point urban and infrastructure decay/degradation grip them (requiring often fraught urban renewal and transit schemes) .

If you are wondering I prefer the organic urban geography style:


2 thoughts on “The Urban Geography of Tokyo

  1. Matias Echanove and Rahul Srivastava are a great writing team on Tokyo. They call it “an incrementally redeveloped slum”.

    It is fascinating how Tokyo and Japanese cities manage to have good-value Real Estate without the sprawl that is the guarantor of the good value in many US cities (and the sprawl in those cities guarantees the affordability regardless of low density mandates, NIMBYism and other factors alleged to be problematic for affordability).

    Japan did have one severe lesson-teaching RE bubble and bust, so their institutional arrangements did not completely guarantee stable affordability like the USA’s median-multiple-3 cities have. But it is still remarkable that cities in Japan have remained so non-sprawling without there being the relentless trend to land rent inflation and cyclical volatility repeating endlessly every 15-20 years like in every Anglo market that decides to ration its land supply.

    I think there are four factors. Japan’s demographic collapse. The lassez-faire nature of redevelopment rights. Cultural love of hygeine and hence “new” things. Japanese hate sitting in seats that other people have sat in for years, including car drivers seats; and they hate living in spaces previously inhabited by others. So they are constantly tearing houses down and building new ones, cheaply and with minimal regulatory mandates. This weakens the important bubble-pumping factor in other countries, of speculative mania in existing housing units.

    The fourth factor is the institutional arrangements for public transit. Public transit enterprises are landlords as well, owning most of the land around their own stops. When separate owners are involved and the public transit is a subsidized monopoly, what results is guaranteed capital gains and wealth transfers for the property owners regardless of what they do. But in Japan the incentives are aligned: the interprises involved want riders and they want tenants and they want “trip attractors”. They want their network to be like a “master planned community” in network form. And there are multiple, competing enterprises each with their own such networks on offer. Hence downwards competition in rents; all of them are competing for “volume” in ridership and tenants. This is the opposite of the perverse consequences that sabotage “transit oriented planning” in the Anglo world.

    In fact this is the one way that mass transit and redevelopment can be a substitute for the competitive effects of automobile-based sprawl. Japan’s cities also have wonderful “articulated density” as a result; density spikes at the right places. No Anglo-world city with fringe land rationing policies has achieved anything other than inefficient cramming in of structures everywhere including at the least efficient locations, which is where you should have low density. See Alain Bertaud, “The Spatial Distribution of Urban Densities”.

    It might surprise you that the Tokyo-Yokohama urban area density is 4400 people per square km versus London’s 5,900. And seeing Tokyo has a lot better articulated density than London, this means that Tokyo has much less crowded suburbs than London. Another important factor in the funtionality of urban density, is the intensity and connectedness of the street network. Auckland is singled out for ridicule by the authors of the UN report “Streets as Public Spaces and Drivers of Urban Prosperity” for its outlier-low street network intensity and bad connectivity. It has only 1/3 to 1/2 the street-network capacity of famous powerhouse cities like Tokyo. This explains why Auckland is imploding with congestion inefficiencies at 2800 people per square km – and the planners are hell-bent on increasing this!

    The UN report completely blew the cover on the absurd myth that Auckland’s problems involve excessively “car dependent” planning. The opposite is the case – its street network inadequacy has no comparators in the first world, but is close to former USSR cities where no-one was allowed to own cars (and post-totalitarianism, they are now the world’s most congested, actually worse than Auckland). One of the UN report’s main arguments was that the viability of repurposing street space for non-car modes is strongly affected by the amount of street space you have to start with. We are stark, raving bonkers if we think Auckland can have cycle lanes and light rail tracks netted out from the already world-outlier-low street capacity without disastrous consequences for urban efficiency. NZ cities are going to become a historical political object lesson in the consequences of complex issues that no-one has the patience to get knowledgeable about, being left to “experts” who are quasi-religious ideologues and charlatans.

Comments are closed.